CHESS Andrew Greet - Starting Out - The Accelerated Dragon (Everyman 2008)

starting out: fundamental coverage of a dynamic Sicilian

starting ut: the accelerated 1 drag 0 n ~~~~~::n~~ a

dynamic Sicilian

ANDREW GREET

Gloucester Publishers pic www.everymanchess.com

First published in 2008 by Gloucester Publishers plc (formerly Everyman Publishers plc), Northburgh House, 10 Northburgh Street, London EClV OAT Copyright © 2008 Andrew Greet The right of Andrew Greet to be identified as the author of this work has been asserted in accordance with the Copyrights, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

ISBN: 9781 85744 5305 Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 06437-0480. All other sales enquiries should be directed to Everyman Chess, Northburgh House, 10 Northburgh Street, London EC1V OAT tel: 020 7253 7887; fax: 020 7490 3708 email: info®everymanchess.com: website: www.everymanchess.com Everyman is the registered trade mark of Random House Inc. and is used in this work under licence from Random House Inc. ToAeHa

(formerly Cadogan Chess) Chief Advisor: Byron Jacobs Commissioning editor: John Emms Assistant editor: Richard Palliser EVERYMAN CHESS SERIES

Typeset and edited by First Rank Publishing, Brighton. Cover design by Horatio Monteverde. Printed and bound in the US.

Contents Bibliography

4

Introduction

5

1 The Hyper-Accelerated Dragon

9 2 5 Nc3 -Classical and Other Lines 49 3 Yugoslav Accelerated Dragon: Introduction and 7 ... Qa5!? 72 101 4 Yugoslav Accelerated Dragon: Main Line with 7... 0-0 5 Lines with Nxc6

131

6 Mar6czy Bind: Classical Variation

147

7 Mar6czy Bind: Gurgenidze System

177 226 262

8 Mar6czy Bind: 7 ... Ng4 9 Mar6czy Bind: Avoiding Exchanges with Nc2 10 Mar6czy Bind: Sidelines

289

Index of Variations

317

Index of Complete Games

319

Bibliography Books

3 h4 in the Hyper-Accelerated Dragon, ].Donaldson & ].Silman (pp 33-50 in Secrets Of Opening Surprises, ].Bosch (ed), New In Chess 2007) Accelerated Dragons, ].Donaldson & J.Silman (Everyman 1998) Beating the Sicilian 3, J.Nunn & ].Gallagher (Batsford 1995) Chess for Tigers (2nd edition), S.Webb (Pergamon 1990) Chess Openings for Black, Explained, L.Alburt, R.Dzindzichashvili & E.Perelshteyn (henceforth known as 'ADP') (Chess Information and Research Center 2006)

Experts vs. the Sicilian, J.Aagaard & J.Shaw (Eds) (Quality Chess 2004) Opening According to Kramnik: Volume 3, A.Khalifman (Chess Stars 2001) Play the Sicilian Dragon, E.Dearing (Gambit 2004) Starting Out: The Sicilian Dragon, A.Martin (Everyman 2005) The Seven Deadly Chess Sins, J.Rowson (Gambit 2001) The Sicilian Accelerated Dragon, P.H.Nielsen & C.Hansen (Batsford 1998) Media Accelerated Dragon Assault!, A.Martin (Bad Bishop 2003) The Mar6czy System, S.Tiviakov (ChessBase 2007) Databases Mega Database 2007; Correspondence Base 2006; The Week In Chess 1-670; Britbase (Relevant games were extracted from these sources and copied into a single database, from which any duplicated games were deleted. This became a single reference database used for research plus statistical reports.)

Analysis Engines I used three analysis engines: Fritz 9 for general use; Deep Shredder 10 for positional/strategic situations; and Deep Junior 10 for more dynamic positions or those featuring material imbalances.

Introduction Greetings, dear reader! I hope that you will enjoy reading this, my second book for Everyman Chess. I will always feel proud of Play the Ruy Lopez (PTRL), although of course I remain humble and fully aware of the almost limitless scope for improvement in my writing. The present book has posed a very different set of challenges, but I hope that the lessons learned from PTRL have enabled me to correct a few shortcomings while building on the positive features of a book which, notwithstanding the occasional criticism, was on the whole received very positively by readers and reviewers alike. The aim of this book is to provide the reader with everything he needs to begin playing either side of the Accelerated Dragon with confidence. As implied by the 'Starting Out' title, I have not assumed that the reader will have any experience of this opening, and have endeavoured to explain all of the fundamental principles as lucidly as possible for the benefit of those lacking any prior knowledge. At the same time I hope and believe that more experienced practitioners of the Accelerated Dragon will also benefit from a number of original ideas and analyses relating to several important variations of this opening. This book contains plenty of guidance for both colours, although it seems only natural that the balance should be weighted somewhat towards Black, in order to cater for the majority of readers. This is not intended to be a dense theoretical work; instead I will be focusing on the key ideas and strategies for both sides. At the same time you can still find plenty of original suggestions and deeper analysis of certain critical lines. As is customary for the Starting Out series, plenty of Notes, Tips and Warnings have been included in order to emphasize the most important points. I am not recommending a precise repertoire, but there is a lot of general guidance about certain good or bad lines as well as some specific recommendations in key variations.

History of the Accelerated Dragon According to my database the opening moves 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 were first seen in the game H.Kennedy-M.Wyvill, London 1851. At that time Black's play would have seemed highly irregular and it took another full century before the Accelerated Dragon achieved any real popularity, although a few prominent players (most notably Emanuel Lasker) dabbled with it from time to time. In terms of longevity, the greatest exponent of the Accelerated Dragon must surely be the Danish Grandmaster Bent Larsen,

5

Starting Out: The Accelerated Dragon whose experience in this opening spans five different decades, from the 1950s through to the new millennium. Other long-term devotees of this opening include Margeir Petursson and Dragoljub Velimirovic, while in modern times two world class Grandmasters deserve special recognition as champions of Black's cause: the Russian Vladimir Malakhov, and his compatriot (though now a Dutch citizen) Sergei Tiviakov. Furthermore a host of top GMs have employed the Accelerated Dragon on an occasional basis, including Viswanathan Anand, Kiril Georgiev, Vassily Ivanchuk, Gata Kamsky, Arty Timofeev, Veselin Topalov and Vadim Zvjaginsev.

My Personal Experience of this Opening Like many Accelerated Dragon players, my foundational knowledge of this opening came through extensive experience of the standard Dragon: 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 g6. As a junior this was always my main defence to 1 e4, although I experimented with other openings -most notably the French - from time to time,. After my teenage years I realized that the Dragon, while fascinating, was stylistically quite a poor choice of opening for me, and I began searching for something a little less 'on the edge'. The Accelerated Dragon was a natural choice, leading to positions with an obvious resemblance to those to which I had become accustomed, without allowing the game to erupt into what felt like a random melee of tactics and checkmating combinations. It has been a part of my repertoire since around the turn of the new millennium. For much of that time it has been my main weapon against 1 e4, although nowadays I play l...e5 more frequently. Still, I am happy to say that the process of researching and writing this book has greatly reinvigorated my interest in this opening, as well as improving my understanding and causing me to revise my opinion of certain important variations. So let my future opponents be warned the Accelerated Dragon may yet return as a central part of my repertoire against 1 e4!

Overview The purpose of this short section is to summarize the opening moves of the Accelerated Dragon and identify some of the key features of this opening. Each of the major variations will receive a full examination in its own dedicated chapter, to be found elsewhere in the book. 1 e4c5 Black's first move brings us the Sicilian Defence. It should be noted that the Accelerated Dragon can also arise via transposition from a few different openings, including the English and occasionally even the King's Indian. 2 Nf3 This is the most common response to the Sicilian. White develops a piece and prepares to open the centre with d2-d4, although this is by no means obligatory and there are many players who prefer to deviate with something like 3 Bb5. If Black intends to play the Accelerated Dragon then he must make an important decision on the second move. 2... Nc6

This is the officially recognized move order of the Accelerated Dragon according to the Encyclopaedia of Chess Openings (ECO) system of classification. There are, however, many

6

Introduction players who prefer to shuffle the move order with 2...g6!?, which is known as the HyperAccelerated Dragon. This is a perfectly valid option which has the important advantage of avoiding the Rossolimo variation which might occur after 2... Nc6 3 Bb5. After 2... g6 the most common continuation is 3 d4 cxd4 4 Nxd4, leading to an Accelerated Dragon after 4... Nc6. The sole drawback is that White is presented with a few additional- though not necessarily advantageous- options such as 3 c3!? and 3 d4 cxd4 4 Qxd4!?. For a full discussion of the Hyper Accelerated Dragon please refer to Chapter One. 3d4

With this standard move White brings about the Open Sicilian. 3...cxd4 4 Nxd4

Now Black can choose between a variety of different defensive systems, including 4... e6 and 4 ... Nf6 5 Nc3 followed by 5 ...d6, 5 ... e6 and so on. Instead, our subject of the Accelerated Dragon occurs after ... 4...g6 (Diagram 1)

Diagram 1 (W) The Accelerated Dragon

Diagram 2 (W} The 'standard' Dragon variation

Diagram 1 shows the official tabiya (starting position) of the Accelerated Dragon. With his fourth move Black has prepared the active fianchetto development of his king's bishop on g7. Next will come a speedy mobilization of the kingside forces with ... Nf6 and ... 0-0, followed by active play in the centre and/or on the queenside. Immediately we can identify three healthy features of his opening set-up: 1) Black develops rapidly and castles quickly. 2) His pieces are positioned actively, especially the characteristic 'Dragon bishop' on g7. 3) His pawn structure is sound and reasonably flexible. According to circumstances he may play an early ... d7-d6, ... d7-d5, or occasionally ... e7-e6 and ... d7-d5.

A comparison with the 'normal' Dragon variation Before moving on to our main subject of the Accelerated Dragon, I think it will be beneficial

7

Starting Out: The Accelerated Dragon to compare the standard Dragon variation of the Sicilian. This comparison will become something of a recurring theme in Chapters 2-5 which deal with the different variations occurring after 5 Nc3. The Dragon variation occurs after the opening moves 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 g6 {Diagram 2) Black's fifth move characterizes the Dragon variation, so called because the line of black pawns from h7 to d6 is said to resemble the shape of the mythical fire-breathing lizard. Black hopes that his dark-squared bishop will exert a powerful influence from g7, and the system is renowned as one of Black's most active and combative methods of fighting against 1 e4. An initial comparison between the two related systems reveals a couple of differences: 1) In the Accelerated Dragon, Black has 'substituted' the developing move ... Nc6 in place of ...d6.

In principle this difference ought to favour Black slightly. The reason is that, while both moves are quite useful, ... Nc6 is- with the exception of a few rare Dragon lines in which Black opts for ... Nbd7- practically an indispensable developing move, whereas ... d7-d6 is less than essential for the time being. In particular, there are certain variations in which Black may be able to achieve the freeing break ... d5 in one move instead of two, thereby saving a whole tempo compared with the analogous variation from the Dragon. The early ... d7-d5 thrust is perhaps the single most dominant theme of Chapters 2-4.

2) Both sides are missing a knight move; Black's knight is on g8 instead off6, whilst White's is on b1 instead of c3. Compared with the standard Dragon, this gives White an additional, and very important option of the pawn advance 5 c4 before his queen's knight comes to c3. This space-gaining move sets up the notorious Mar6czy Bind, the permission of which constitutes the sole drawback of the Accelerated as opposed to the normal Dragon system. Chapters 5-10 are dedicated to this branch of the opening.

That concludes our brief introduction to the Accelerated Dragon. We will now begin to examine individual variations in further detail, beginning with the Hyper-Accelerated Dragon: 2...g6. All that remains is for me to thank you once again for reading this book, and to wish you great success playing both sides of this opening. Andrew Greet Cornwall, England February 2008

8

Chapter One

The Hyper-Accelerated Dragon

-----

Black's Move Order Dilemma Basic Theory and Early Deviations 3••• Bg7!? - Advice for White Advice for Black The Charge of the h-pawn A Quiet Bishop Development The Pseudo-Aiapin: 3 c3

Summary and Conclusions

The Main Line: 4 Oxd4

Starting Out: The Accelerated Dragon

Black's Move Order Dilemma If you wish to start playing the Accelerated Dragon, then your first task is to decide which of the possible move orders will best suit your needs. As we saw in the Introduction, the two options are as follows: Option 1 -the 'traditional' move order according to the ECO classification: 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6

Option 2- the Hyper-Accelerated Dragon: 1 e4 c5 2 Nf3 g6 3 d4 cxd4 4 Nxd4 Nc6

Both move orders carry their own unique advantages and disadvantages. The main drawback of 2... Nc6 is that, in addition to studying the positions after 3 d4, you will also need to prepare an antidote to 3 Bb5, a popular line known as the Rossolimo variation. The Rossolimo is regarded as one of the more promising anti-Sicilian systems and accounts for approximately one in six games after the opening moves 1 e4 c5 2 Nf3 Nc6. For some readers, this will be of little or no concern. Those of you with prior experience of one of the 2... Nc6 Sicilian systems (e.g. the Accelerated Dragon, Sveshnikov, Classical, etc) will probably have encountered 3 Bb5 on numerous occasions and may already have a favourite anti-Rossolimo system worked out. If this applies to you, then the 2... Nc6 move order would appear to be the most straightforward option. For other readers, matters may not be so simple. Perhaps you are reading this book with a view to playing the Accelerated Dragon for the very first time? Maybe you are an experiencec Sicilian player who is fed up with meeting 'cowardly' opponents who continually avoid the main lines in favour of anti-Sicilians like the Rossolimo. Or you may simply not enjoy facing ~ Bb5 -after all, it is a highly respectable line and plenty of players struggle against it. If any of the above applies to you, then you may wish to consider bypassing the Rossolimo with the move 2...g6!? (Diagram 1).

10

Diagram 1 (W)

Diagram 2 (W)


A slightly offbeat and risky line

The Hyper-Accelerated Dragon Obviously the move 3 Bb5 would be completely pointless here, so White must decide whether to head for the main lines or opt for a different kind of anti-Sicilian. A lot of games continue 3 d4 cxd4 4 Nxd4 Nc6 (or 4... Bg7), when Black has achieved the desired transposition and successfully bypassed the Rossolimo.

0

NOTE: Black occasionally plays 4... Nf61? 5 Nc3 (5 e5? would be a blunder due to 5...Qa5+ winning thee-pawn) 5... d6, reaching a standard Dragon variation. This move order is sometimes used by Dragon players wishing to avoid other anti-Sicilian lines such as 2... d6 3 Bb5+.

The drawback of 2... g6 is that White does have some additional options, the most important being 4 Qxd4!?. It all boils down to something of a trade-off from Black's point of view. If he wants to bypass the Rossolimo variation, then he must be prepared to deal with any one of a number of deviations which are unique to the Hyper-Accelerated Dragon.

Which Move Order would be Best for me: 2 ... Nc6 or 2 ... g6 - ? There are no right or wrong answers here - the decision is very much a matter of personal taste. Would you be more comfortable facing the Rossolimo, or one of the variations contained in the present chapter? I hope that the material presented in these pages will enable you to make an informed decision.

Basic Theory and Early Deviations Before we begin, I must emphasize that this chapter is not intended to provide complete coverage of every sub-variation of the Hyper-Accelerated Dragon. The subject of this book is the Accelerated Dragon, and I will assume that readers on Black's side will be interested in utilizing 2...g6 merely as a vehicle for reaching this system while bypassing the Rossolimo. Likewise, I will assume that reader interested in the White side of the position will also be content to reach an Accelerated Dragon. Therefore in this chapter I will provide advice for both sides concerning what to do if the opponent avoids playing into the Accelerated with 3 d4 cxd4 4 Nxd4 Nc6. Let us start by familiarizing ourselves with the basic framework of the theory of 2...g6. This will enable us to identify the most important variations. We will also take a look at some of the highly irregular (though not necessarily bad) possibilities available to both sides. 1 e4 c5 2 Nf3 g6

Now a lot of games simply continue 3 d4 cxd4 4 Nxd4, when 4... Nc6 reaches the basic starting position of the Accelerated Dragon.

0

NOTE: The alternative move order 4... Bg7 is also possible, but please note that if you wish to meet the Maroczy Bind (5 c4) with the Gurgenidze System from Chapter Seven, then 4... Nc6 will have to be played in order to facilitate the early knight exchange on d4.

If you do not intend to play this particular system then 4... Bg7 is perfectly fine, and I suppose it has the minor benefit of ruling out the (admittedly rare and not at all dangerous) option of 5 Nxc6 bxc6 6 Qd4 which will be discussed in Chapter Five. Black has occasion-

11

Starting Out: The Accelerated Dragon ally experimented with delaying or omitting ... Nc6 altogether; e.g. 5 c4 Nf6 6 Nc3 0-0 7 Be2 d6 8 0-0 b6, aiming for a 'Hedgehog' formation after 9 Be3 Bb710 f3 Nbd7!? with a reasonable position, though objectively White ought to have a slight edge. Still, if you fancy tryin~ something new with practically no theory to learn then this could be worth a punt. Starting on the third move, there are several ways in which the game may deviate away from the Accelerated Dragon, although it is White who has the greatest variety of such options available. The most popular and critical variation is 3 d4 cxd4 4 Qxd4, while 3 c3 also needs to be taken seriously. Both of these lines will be examined later. For the moment hen are a couple of minor possibilities which will usually just transpose to lines examined elsewhere. a) 3 Nc3 Bg7 also leaves White with nothing better than 4 d4 cxd4 5 Nxd4, leading to norma positions after s... Nc6. b) 3 c4 carries little independent significance from Black's point of view. After 3... Nc6 (or 3... Bg7) White's only sensible plan is to play 4 d4, leading to a Mar6czy Bind after 4... cxd4! Nxd4. ~

..,

WARNING: White players intending to play a Mar6czy Bind should be aware that 3 c4 is by no means a guaranteed route to his intended destination.

FolloQg 3 ... Bg7 4 d4, Black is not forced to exchange pawns, and can instead opt for one of the alternatives mentioned in the note to White's fourth move in the following section (3 ...Bg7). Although White can still fight for the advantage, the positions are not always eas] to handle and on balance I think that the majority of White players will be better advised t< stick with the usual move order of 3 d4. We will now address the principal ways in which both sides may avoid the transition to th Accelerated Dragon, beginning with a deviation from Black on the third move.

3... Bg7!? -Advice for White This book is about the Accelerated Dragon, and thus I am assuming that readers interested in the White side will be content to reach the position after 3 d4 cxd4 4 Nxd4. Most Black players who employ the 2... g6 move order are merely using it as a means of bypassing the Rossolimo variation, and will be quite happy to answer 3 d4 with 3 ... cxd4. However, from time to time you may run into an opponent who tries 3... Bg7!? (Diagram 2). This is normally used by players who like to get 'out of book' at an early stage. It is playable and certainly has the right to exist, though modern analysis has revealed a variety of hidden dangers for the second player. 4 dxcs!

Definitely the critical continuation - now if Black wants to regain his pawn then his queen will have to take up an exposed position on c5. White players sometimes opt for 4 c4 or 4 Nc3 here, hoping for a transposition to their favoured line against the Accelerated Dragon after 4... cxd4 5 Nxd4. The problem is that Black has some ways of keeping the game in independent territory. Thus, after 4 c4, Black can choose between 4... Qa5+, 4 ... Nc6 or 4... d6!?, intending some combination of ... Nc6 and ... Bg4 to control the sensitive d4-square. Similarly, after 4 Nc3 White must consider possibilities such as 4... Qa5!?, e.g. 5 Be2 Nc6 6 d5 Nd4. Methods do exist for White to combat these variations, but the positions are some-

12

The Hyper-Accelerated Dragon what irregular and often end up resembling queen's pawn openings which may not suit the typicall e4 player. My recommended solution offers White excellent chances of an opening advantage while preserving the overall structure and character of the Open Sicilian. 4...Qa5+ Other moves may leave Black struggling to regain the pawn. 5 c3 Qxcs Now 6 Be3 used to be the main line, when 6...Qc7 7 Bd4 Nf6 8 eS Ng4 9 e6 f6 brings Black an acceptable position. However, recent years have seen the development of something altogether more venomous. 6 Na31 (Diagram 3)

Diagram 3 (B)

Diagram 4 (B)

The 'knight on the rim' is not dim!

Black is already busted!

This excellent move forces Black to proceed with the utmost caution. The black queen is somewhat misplaced on cS, and White reasons that an immediate attack with 6 Be3 will only help Black by chasing it to a better square. The idea of the text is to prepare NbS, after which the threat of Be3 will become much more severe thanks to the vulnerability of the a7pawn and the inability of the queen to retreat to her preferred home on c7. Black must proceed with great caution, as evidenced by the following variations: a) To begin with, there is a particularly lethal trap lurking in this position, which continues to catch out even quite strong players. The point is that 6 ... Bxc3+7 7 bxc3 Qxc3+ looks as though Black is winning material under relatively safe conditions. But after 8 Qd2! Qxa1 White unleashes the sting in the tail with 9 NbS! (Diagram 4). Now Nc7+ is threatened, and after 9 ... Na6 (no better is 9... Kd8 10 Nc3!) 10 Nc3! the black queen is trapped on al, and lO... NcS is met by 11 Nd4 when Black will have to make too many concessions to rescue her. Furthermore, a surprising number of Black's other candidate moves can lead to an equally swift catastrophe: b) 6 ...a67 prevents NbS, but the cure proves to be worse than the disease after 7 Nc4! intending Be3, when the weakness of b6 is disastrous for Black. c) 6 ...Qc77! loses too much time after 7 NbS Qd8 (7... QaS 8 QdS! threatens Nd6+, and 8 ... Nc6

13

Starting Out: The Accelerated Dragon 9 Bc4 is terrible for Black) 8 Bf4 d6, and now the energetic continuation 9 e5! a6 10 Qa4! Kf8 11 exd6 e5 12 Bg5 Qd7 13 Qa3 gave White a huge advantage in A.Galkin-A.Gubajdullin, Tomsk2006. d) 6... Nc6 7 Nc4! Nf6 8 Be3 Qh5 9 Be2 Ng4 10 Bf4 Qc5 11 0-0 Nce5 12 Nfxe5 Nxe5 13 Ne3 was clearly better for White in B.Tereick-L.Konietzka, Ruhrgebiet 2005. e) 6... d6 7 Nb5 a6 8 Be3 Qc6 9 Na7! Qc710 Nxc8 Qxc8 11 Qb3 Nf6 12 e5 dxe5 13 Nxe5 0-0 1< Be2 led to a position in which the two bishops and queenside majority gave White a lasting advantage in S.Dolmatov-D.Solak, Istanbul2003. f) 6... Qa5!? is a better attempt, when play may continue 7 Qd5!? (7 Nc4 is a decent alternative) 7...Qxd5 (7 ... Nc6 8 Qxa5 Nxa5 9 Nb5- Shaw, and 7...Qd8 8 Bc4 e6 9 Qd6 Bf8 10 Qd4 fe 11 Nb5 are both dismal for Black) 8 exd5 d6! (8 ... Nf6 9 d6 is evaluated as better for White b) Bruzon) 9 Nb5 Na610 Be3 Nf611 Rd1 b612 Be2 Bb713 c4 0-0 with only a slight advantage for White in A.Goloshchapov-A.Kislinsky, Poltava 2006. Finally, the most natural and probably best move for Black is 6... Nf6, which will be considered in Game 1.

Theoretical? There is not too much that White needs to know here. Most of the attacking moves come fairly naturally, although a precise knowledge of the forcing lines can only improve his chances. It is Black who really needs to know his stuff, as his margin for error is very small and one mistake can lead to a swift disaster.

Statistics 6 Na3! scores a promising 66% for White, although this drops to a more modest 55% after Black's best reply of 6... Nf6.

Illustrative Game Gamel

D E.Shaposhnikov • V.Malakhov Moscow2007 1 e4 cs 2 Nf3 g6 3 d4 Bg7 4 dxcs QaS+ 5 c3 Qxcs 6 Na3 Nf6

This is Black's best attempt at reaching an acceptable position. Vladimir Malakhov is one 01 the world's leading specialists in the Accelerated Dragon, so any variation that he plays deserves to be taken seriously.

7 Nbs! (Diagram 5) White continues with the standard plan. He intends Be3, harassing the black queen while also menacing the a7-pawn. 7... b6!

This looks like the best chance to render the black position playable. A brief perusal of the alternatives will reveal how dangerous this variation can be for an unprepared Black player.

14

The Hyper-Accelerated Dragon a) 7... Nxe4? 8 Be3 Qc6 9 Nfd4 Bxd4 10 Qxd4 Nf611 Bf4! Na6 (or ll ... d6 12 Nxd6+!) 12 Nd6+! Kf8 13 Bc4 gives White a huge initiative. b) 7... Ng4 8 Nfd4! d6 9 h3 Nxf2!? (9 ... Nf6 10 Be3 Nxe4 11 Ne6led to a swift victory for White in R.Swinkels-C.Ghysels, Belgium 2005) 10 Kxf2 eS 11 Be3 exd4 12 cxd4 Qb6 was M.AI Sayed-F.Velikhanli, Abu Dhabi 2003, and now Shaw's recommendation of 13 Rcl Na6 14 eS! looks extremely strong for White. c) 7 ...0-0 8 Be3 Qc6 9 Nfd4! Qxe410 Nc7 b611 Be2 Qb712 Nxa8 Qxa8 13 Bf3 Bb7 gave Black a degree of compensation for the exchange in L.Bruzon Bautista-V.Malakhov, World Junior Ch., Yerevan 2000, but Bruzon' s subsequent suggestion of 11 Qf3! would have forced a favourable queen exchange, while the aS-rook is not about to run away. 8 e5

This looks like the critical continuation. Instead 8 Be3 Qc6 9 Bd3 0-0 reaches, via a different order of moves, the game M.AI Modiahki-B.Macieja, Turin Olympiad 2006, which continued 10 0-0 Bb7 11 Re1 d612 a4 a613 Nbd4 Qc7 with approximate equality. 8 ... Ng4 9 Qd4 Nxe5! (Diagram 6)

Diagram 5 (B)

Black must tread with great care

Diagram 6 (W) A critical position for 3... Bg7

Offering a very interesting positional exchange sacrifice. This is far more promising than the feeble 9 ...Qxd4?! 10 cxd4 when White dominates the play. 10Qxc5

Accepting the material, but I wonder if this might be the time to look for an improvement. My suggestion would be 10 NxeS!? QxeS+ (10 ... Bxe5? leads to serious problems after 11 Qe4! intending 12 Be3) 11 QxeS Bxe512 f4 a6 13 fxeS axb514 BxbS (Diagram 7), when White's bishop pair ought to suffice for a slight edge. 14 Be3!? also looks interesting. The game continuation is not bad for White, but nor does it bring him any real advantage. 10... Nxf3+ 11 gxf3 bxc5 12 Nc7+ Kd8 13 Nxa8 Bb7 14 Be2 Bxa8 15 Be3 d6 (Diagram 8)

Black is the exchange down but his position may already be arguably the easier to play. As compensation he enjoys one extra pawn, superior minor pieces and a very healthy structure. Note also the lack of open lines for White's rooks and bishops.

15

Starting Out: The Accelerated Dragon 16 0-0 as!

Diagram 7 (B)

Diagram 8 {W)

White should be slightly better

Full compensation for Black

White's only real chance for activity involves opening the queenside, so Malakhov immed ately sets up a clamp on that side of the board. 17 Rab1 Nc618 f4 Kc7 19 a3 RbS 20 Rfc1 NdS!? A logical regrouping. 21b4! This is White's only active plan, so he has to try it. 21... Be4! 22 Rb2 Ne6 23 f3 BfS 24 Bc4 Bh6 25 Re2 Kc6 (Diagram 9)

Diagram 9 {W)

Diagram 10 {W)

White is under pressure

Black wins

26 bS+?! This does not seem to help White's cause. In fact he should probably not be too worried

16

The Hyper-Accelerated Dragon about losing a pawn on b4, as the resulting clearance of the a-, b- and c-files would greatly increase the value of his rooks. 26 Rd1! looks best, removing the rook from the x-ray vision of the h6-bishop, while also hinting at the possibility of the capture bxaS!?, which could prove troublesome when combined with the newly facilitated bishop check on dS. 26... Rxbs!?

Not content with sacrificing just one exchange, Malakhov decides to double his investment. It works very nicely in the game, although 26 ...Kb6 would also have maintained a clear

advantage. 27 BxbS+ KxbS 28 C4+ Kc6 29 Rb2 Nxf4 30 Bxf4 Bxf4 31 Ret?!

31 Rd1 is better; I don't see any value in provoking Black's next. 31 ... e5 32 Rd1 Be6 33 RbS?!

33 Rc2 was essential, but it is understandable that White was not drawn to such passivity, and in any case 33 ... d5!? 34 cxdS+ BxdS 35 Rc3 c4 36 Kf2 gS would leave White facing a miserable and ultimately futile defensive task. 33 ... Bxc4 34 Rxas fs 35 Kf2 e4 36 fxe4 fxe4 (Diagram 10)

We can now see the culmination of Black's strategy, which began with his early exchange sacrifice. The white rooks are helpless against the combined strength of Black's bishops and passed pawns. 37 Ra4 dS 38 Rb1 Bd3 0-1

In the majority of games 6 Na3! has provided White with excellent chances for a strong initiative, although the illustrative example should serve as a reminder that the first player is not guaranteed an easy ride. Should you encounter Malakhov's chosen variation from this game then my suggested improvement on move 10 offers a safe route to a slightly favourable endgame.

Advice for Black We will now address the various ways in which White, after the opening moves 1 e4 cS 2 Nf3 g6, may deviate from the standard Open Sicilian sequence of 3 d4 cxd4 4 Nxd4. As I said before, I am assuming that White players who buy this book will be interested in learning how to play against the Accelerated Dragon, and so the remainder of the chapter should be viewed primarily as a guide for Black. The most popular and challenging of White's options is 3 d4 cxd4 4 Qxd4, while 3 c3 and 3 Bc4 will also be examined later. But before we move on to those variations we must also consider a surprising yet quite dangerous approach.

The Charge of the h-pawn 1 e4 cs 2 Nf3 g6 3 h4!? (Diagram 11)

This will certainly appeal to fans of the unorthodox! 3 h4 may seem like the type of silly idea which might be fun for a blitz game but not much else. In fact it is nowhere near that bad and has been used by a number of very strong players. It has also been the subject of a recent SOS article by Silman and Donaldson.

17


Diagram 11 (B)

Diagram 12 (B)

A choice for creative players

A disaster for Black

3... hs!?

I would recommend this to players looking for a safe, no-nonsense approach, although there are still some pitfalls that must be avoided as we will see. 3 ... Bg7 and 3 ... Nf6 are bot~ playable, but by permitting 4 h5 and 4 e5 Ng4 5 d4 respectively, Black allows his adversar the kind of unbalanced position that he is aiming for with 3 h4. 4 d4 cxd4 5 Bc4!? 5 Qxd4 is not dangerous after 5 ... Nf6, e.g. 6 Bb5!? Bg7!, and compared with the analogous variation after 3 d4 cxd4 4 Qxd4, the availability of the g4-square for the knight makes e4-E a harmless proposition, so Black is doing fine. s... Bg7 6 c3!? e6! I agree with Silman and Donaldson that this is Black's safest choice. WARNING: It is very risky for Black to accept the offer: 6... dxc3?1 7 Nxc3 would give White an improved version of the Smith-Morra gambit due to Black's weakened kingside.

The well-known game A.Rodriguez-B.Larsen, San Martin 1994, continued 7... Nf6 8 Bf4 d6 (8 ... Bg7 was essential) 9 Ng5! e6 10 Nb5! a6 11 Nxd6+ Bxd6 12 Bxd6 (Diagram 12). White had regained the sacrificed pawn with an overwhelming positional advantage and won quickly. After 6... e6! play may continue 7 cxd4 d5 8 exd5 exd5 9 Bb5+ Nc6 followed by ... Nge7 and ...0-0 with approximately equal chances. The line can be compared with the variation 3 Be' covered later in this chapter. It is hard to say whether the mutual advance of the h-pawns favours one side over the other and it would be interesting to see a practical test.

A Second Possible Solution After 3 h4!? Black can also consider the slightly cheeky 3... h6!?, as in V.Golod-M.Berkovicl Rishon Le Ziyyon 2002. The idea is simply to meet h4-h5 with ... g6-g5. Hardly any theory

18

The Hyper-Accelerated Dragon exists on this position and White can try practically anything. The question is: which side will be able to utilize the extra moves h2-h4 and ...h7-h6 to its own advantage? The aforementioned game continued 4 Bc4 Bg7 5 c3 e6 6 Qe2 Ne7 with unclear play. Silman and Donaldson suggest that 5 h5 g5 6 d4 may be critical, mentioning a possible continuation 6... cxd4 7 Nxd4 Nc6 8 Nf5! which looks logical, although I wonder if Black might be able to try 8... Be5!? (Diagram 13),

Diagram 13 (W)

Diagram 14 (W)

An original suggestion

A relatively quiet variation

when 9 g3 d6 and 9 Nd2 Nf6 (or 9 ... e6!?) 10 Nf3 d5! both look highly unclear. For players on the black side of the board who wish to respond to 3 h4 'in kind'- by playing something unexpected and with practically no theory whatsoever - then 3 ... h6!? could be just the ticket.

Theoretical? Very little theory exists on 3 h4!?, although both sides will certainly benefit from some basic knowledge of which lines to avoid.

Statistics 3 h4 scores a total of 52% for White, after which Black's most solid, no-nonsense defence of 3 ... h5 leads to a perfectly balanced 50%. I could only find one example of 3 ... h6!?, which was eventually won by White, but it is clearly too early to talk about statistical trends on the basis of a single game!

A Quiet Bishop Development 1 e4 cs 2 Nf3 g6 3 Bc4

This option tends to appeal to players wishing to avoid a heavy theoretical duel. White is not trying to do anything spectacular in the early stages; he simply develops in an unpretentious manner and, given a few moves, will construct a strong pawn centre with c2-c3

19

Starting Out: The Accelerated Dragon and d2-d4. Black should look to counter the bishop development with a quick ...e7-e6 and ... d7-d5, seizing control of the centre while attacking the exposed bishop. This plan should normally enable him to equalize the play. WARNING: There is a positional trap which is worth remembering, namely that if Black plays ... e7-e6 at the wrong time then White may opt for a quick d2-d4, leading to an Open Sicilian in which ... g7-g6 and ... e7-e6 do not combine together very well.

As long as the second player avoids this positional pitfall, he should not have too many problems. Black should begin with the natural move: 3... Bg7 (Diagram 14)

White's plan from this point will usually involve the construction of a strong pawn centre with c2-c3 and d2-d4. The game may develop slightly differently depending on whether h• does this before or after castling. 4 0-0 will be considered in Game 2, so for the time being we will examine White's immediate attempt to construct a pawn centre: 4C3

Alternatives are toothless: a) 4 d3 Nc6 followed by ... e7-e6 and ... d7-d5 is fine for Black. b) 4 d4 cxd4 5 c3!? is an attempt to lure Black into pawn-grabbing. Unfortunately for White the idea can easily be defused by 5... d3!, leading to a Dragon-like position with material equality in which White's pieces are less than ideally placed. After 4 c3 Black should look to organize an immediate counter with ... 4 ... e6! 5 d4

The cheap trap 5 Qe2 should obviously be met by 5... Ne7 intending ... d7-d5, rather than 5... d5?? 6 exd5 when Black cannot recapture due to the pin along thee-file.

s... cxd4 6 cxd4 6 Nxd4!? is interesting, but compared with the analogous position after 4 0-0 the problem for White it that the pawn advance c2-c3 looks completely out of place here, and this should enable Black to obtain reasonable prospects. 6... Ne7! is perhaps the most accurate move, as the immediate 7 NbS is nothing to worry about in view of 7... d5! 8 exd5 a6! 9 Nd4 (9 d6?? drops a piece after 9 ... axb5 10 Bxb5+ Nec6) 9 ... Nxd5 when Black has no problems whatsoever. 6... d5 (Diagram 15)

Now White will have to waste time moving his bishop again. Usually he will give a check on b5, with or without a preliminary exchange of pawns on d5.

a

Bby

TIP: In my opinion, a good rule of thumb in this variation is that bishop checks on b5 should be blocked with the move ... Nc6! rather than ... Bd7.

7

The alternative is the immediate central exchange with 7 exd5 exd5 8 Bb5+ Nc6!, when 9 ONge710 Bg5 f611 Be3 0-0 12 Nc3 Be6 was fine for Black in M.Van der Werf-D.Heinbuch, Biel1992. 7... Nc6!

20


Diagram 15 (W) White's bishop must waste time

Diagram 16 (B) Preparing to attack the centre

7... Bd7 is playable, but I have my doubts as to whether it can suffice for full equality. The text is more uncompromising. Black does not fear an exchange on c6, as his pair of bishops should fully compensate the slightly devalued pawn structure. Nielsen and Hansen call 7... Nc6 dubious, citing the game P.Juslin-I.Bilek, Boraas 1986, which continued ... 8 es

I agree that this is a bit more challenging than 8 exd5 exd5 which transposes to the above note -7 exdS exdS 8 BbS+ Nc6. 8... Bd7 9 Nc3

And now 9.. .f6 10 exf6 Nxf6 11 Bxc6 Bxc6 12 Bf4 0-0 13 Be5 led to an edge to White. However, I think that 9... Nge7!? (Diagram 16) improves; e.g. 10 0-0 Qb6 when Black is preparing ... Nf5 to attack d4, while incidentally threatening to win a pawn with 11...Nxe5!. Depending on how White plays, Black may also consider the undermining move ... f7-f6 to increase the pressure on the white pawn centre. Overall I think that Black has enough play to provide a fair share of the chances.

Illustrative Game Camel

D R.Hasangatin •

S.Vokarev

Maikop 1998 1 e4 cS 2 Nf3 g6 3 Bc4 Bg7 4 0-0 Nc6

Black should take care to avoid 4... e6?! on account of 5 d4! cxd4 6 Nxd4.

5C3 5 Re1!? can be met by the standard response 5 ... e6. Now White can transpose to familiar paths with 6 c3 Nge7, though he occasionally prefers something different in the shape of 6 Nc3!? Nge7, when 7 d4!? is an interesting attempt to open the centre. Although this may

21

Starting Out: The Accelerated Dragon seem wildly aggressive, the position soon settles down after 7... Nxd4 8 Nxd4 cxd4 9 NbS 0-0 10 Nxd4 (10 e5?! d5! 11 exd6 Nf5 works out well for Black) 10 ... d5 11 exd5 Nxd5 12 c3 (Diagram 17)

Diagram 17 (B)

Diagram 18 (W)

Black should be fine here

Is the d6-square a real problem?

12... a613 a4 Qd6! 14 Bfl (H.Hamdouchi-A.Cabrera, Seville 2002) and now 14 ... Rd8looks like the most straightforward move, with good chances for Black. s... e6 6 d4

If White chooses 6 Re1, then 6... Nge7 is likely to transpose to one of the variations considered elsewhere. 6... cxd4 7 cxd4

7 Nxd4!? is an interesting alternative, when Black should probably reply 7 ... Nge7!? (Diagram 18).

Now 8 Nxc6 bxc6 9 Qd6 Bb7! plans ... Nc8, evicting the queen and thus permitting ... d7-d.': after which Black should be fine. It looks more promising for White to continue with 8 Nl: when 8 ... 0-0 (8 ... d5?! loses a pawn to 9 exd5 exd5 10 Bxd5 Nxd5 11 Qxd5! thanks to the fo on c7) 9 Nd6 was reached in M.Mankinen-V.Partikas, Correspondence 2002, and now instead of 9... Na5?! I think Black should have begun counterattacking with 9 .. .f5!. The game double-edged, but it seems to me that Black should be able to claim a fair share of the chances. 7... ds 8 exds exds 9 Bbs Nge7 (Diagram 19) 10 Bgs!? f6 10... 0-0 11 Bxc6 bxc6 12 Re1 ReS 13 Nbd2 Rb8 14 Nb3 is slightly better for White accordinl

to Bangiev, though 12.. .f6! looks like an improvement to me. 11 Bxc6+ Nxc6?

This is an instructive mistake which leads Black into trouble. Vokarev wishes to avoid compromising his pawn structure, but in doing so loses valuable time. After the superior 11...bxc6 Bangiev gives 12 Bf4 0-0 13 Nbd2 as slightly better for White, but I think Black should be fine as long as he avoids excessive exchanges. One very interesting idea would be to commence immediate kingside operations with 13... g5!?, when 14 Bg3 h5!? and 14 B

22

The Hyper-Accelerated Dragon Nf5 or 14 ... Ng6!? give decent prospects. 12 Re1+ Kf7

12 ... Ne7 13 Bf4 0-0 14 Nc3 intending Qb3 also gives White some initiative.

0

Diagram 19 (W}

Diagram 20 (B)

Chances are balanced

White wins a pawn

NOTE: Although the pawn structure after ... b7xc6 may be viewed as slightly inferior for Black, it is not without its advantages, namely a solid defence of the dS-pawn!

13 Bf4 Bg4 14 Nc3 Qd7

14... Bxf3 15 Qxf3 Nxd4 16 Qd1 f5 is strongly met by 17 ReS!- Bangiev. 15 NbS!

Black is not given a moment's respite! The text targets the vulnerable squares d6 and c7. 1S ... Bf8 16 h3 Bxf3 17 Qxf3 Bb4

No better is 17... Kg718 Nc7. 18 Nd6+! (Diagram 20)

This looks like the most energetic way of continuing the attack. Instead 18 Bd6? Bxe119 Qxd5+ Qe6 (19 ... Kg7?? 20 Bf8+ wins the queen) 20 Qxe6+ Kxe6 21 Rxe1+ Kd5 is beginning to look quite tasty for Black. 18... Kg7 19 Qxds Bxd6

On 19... Bxe120 Rxe1! Rhd8 (20... Rad8 21 Re6! is similar) 21 Re6! Ne7 (22 Rxf6! Kxf6 23 Be5+ was the threat) 22 Qe4 Ng8 23 Nxb7looks extremely good for White. 20Qxd6

20 Re6!? may have been even stronger. 20... Rhd8

20 ... Qxd6 21 Bxd6 Rhd8 22 Re6 Nxd4 23 Re7+ Kg8 24 Bc5 b6 25 Bxd4 Rxd4 26 Rcl would also leave Black in difficulties. 21 Qxd7+ Rxd7 22 Be3 Kf7

23

Starting Out: The Accelerated Dragon 22 ... Nxd4?! 23 Bxd4 Rxd4 24 Re7+ was definitely not an improvement. 23 Rac1 Rad8 24 Red1 Ke7

Once again 24 ... Nxd4 25 Rxd4 Rxd4 26 Bxd4 Rxd4 27 Rc7+ was inadvisable. 25 Rc4 (Diagram 21)

Diagram 21 (B)

Diagram 22 (B)

A long defence for Black

The win is in sight

We have reached an endgame in which I would evaluate White's winning chances asapproximately equal to Black's drawing chances. 25 ... Rd6 26 Kf1 a6 27 a4 bs 28 axbs axbs 29 Res Nxd4 30 Bxd4 Rxd4 31 Rxd4 Rxd4 32 Rxb5

I suspect that if the eleven-piece tablebase existed, it would tell us that Black can hold this position, even if the defensive task is far from trivial in practice. For White the difficulty is that he cannot promote the b-pawn without the help of his king. And if his king goes to th queenside, the black rook will gobble his kingside pawns. The following phase of the gam sees him attempting to 'tidy up' his kingside in such a way as to prevent the black rook from causing too much damage there when the crucial time comes. 32 ... Rd2 33 g4! g5 34 Kg2 Ke6 35 Rb7 h6 36 b3 Rd3 37 b4 Rb3 38 Rb8 Kf7 39 bS Kg6 40 b6 Kg7 41 Kh2 Rf3 42 Kg2 Rb3 43 f3 Rb2+ 44 Kf1 Rb1+ 45 Ke2 Rb3 46 Kd2 Rxf3

White has eventually allowed the loss of a kingside pawn, but has done so under favourable circumstances, as Black has no easy way of creating a passed pawn. 47 Kc2 Rf2+ 48 Kc3 Rf149 Rd8 Rb1 50 Rd6 Kf7 51 Kc4 Ke7 52 Rc6 Rb2 53 Kc5 Rb3 54 Rc7+ (Diagram 22) 54... Ke6 55 b7 Kes 56 Rf7 Ke6 57 Rh7 Kes 58 Kc6 fs 59 gxf5 Rc3+ 60 Kd7 Rd3+ 61 Kc8 Rc3+ 62 Rc7 Rb3 63 Rc6 1-0

White's endgame technique was truly exemplary.

0 24

NOTE: All of Black's troubles can be traced back to the eleventh move when he met 11 Bxc6+ with 11... Nxc6? instead of 11... bxc6. By putting his pawn structure ahead of the welfare of his pieces, he handed White the initiative and was never able to recover.


Theoretical? There is not much theory after 3 Bc4. It is useful for Black to remember a few details, such as avoiding the positional trap of a premature ...e7-e6 (i.e. when the response d2-d4 can be played to good effect), as well as little things like blocking a subsequent Bb5+ with ... Nc6 rather than ... Bd7.

Statistics White scores just 46% after 3 Bc4, which I found slightly surprising as there is nothing particularly wrong with it.

The Pseudo-Aiapin: 3 c3 1 e4 cs 2 Nf3 g6 3 c3 (Diagram 23)

Diagram 23 (B)

Diagram 24 (W)

Planning to build a pawn centre

Fighting for the centre

This is White's most popular alternative to 3 d4. Compared with the well-known Ala pin or c3-Sicilian with 1 e4 c5 2 c3, here Black is already committed to ...g7-g6. This is not necessarily a bad thing, but it does bypass Black's most popular method of meeting that variation, namely 2... Nf6. In Diagram 23 Black can choose between two main options. The most common move has been the natural 3.•. Bg7 followed by 4 d4 cxd4 5 cxd4 dS (5 ... d6 reaches a variation of the Alapin in which White's central control and easy development should provide a slight advantage), at which point White can decide between blocking the centre with 6 es or keeping things more open with 6 exds Nf6. These variations have been tested in many games and in general Black seems to be holding his own. However, as most opponents are likely to be prepared for 3 ... Bg7, I have decided to recommend a more challenging solution for Black in the immediate central strike ... 3...ds!? (Diagram 24)

25

Starting Out: The Accelerated Dragon White's third move left no doubt as to his intention of building a strong pawn centre with d2-d4, so it is highly logical for Black to fight for the centre in this way. This is especially true when you consider that, following the almost inevitable exchange on d5, the b1-knigh can no longer come to c3. In fact, it is no exaggeration to say that White would even be better off if he could somehow 'wish away' his third move. 4exd5

White very occasionally tries 4 e5!? with a view to reaching something similar to one of the 3... Bg7 lines (i.e. 3... Bg7 4 d4 cxd4 5 cxd4 d5 6 e5) after 4 ... Nc6 5 d4 Bg4. Unfortunately for White, the presence of both sides' c-pawns on the board gives him some extra problems, the main one being that his queen's knight can no longer be developed on its optimal square. A secondary point is that Black may be able to develop his king's bishop other thar on g7, where it would be blocked by the central pawns. Therefore, in order to obtain any chance of an opening advantage, White must consent to the central exchange. 4... Qxd5 5 d4 Bg7 (Diagram 25)

Diagram 25 (W) The fianchetto works well here

Diagram 26 (W} The d-pawn offers no advantage

This natural move is clearly best. Black sometimes tries 5 ... Nf6 (more often via 2 c3 d5 3 exd5 Qxd5 4 d4 Nf6 5 Nf6 g6), but this allows White to gain an edge with the powerful6 Qb3!, threatening Bc4, while an exchange on b3 would open the a-file for the white rook. Let us now consider some possible ways in which the game might develop from Diagram 25. a) In contrast to the position after 5... Nf6 (see above), 6 Qb3?! is not such a good idea here, due to the strong response 6 ... Qxb3 7 axb3 cxd4 8 Nxd4 Bxd4! 9 cxd4 Nc6 when the loss of Black's dark-squared bishop is more than compensated by White's crippled queenside structure. 10 Be3 Nb4 11 Na3 Nf6 12 Bb5+ Bd7 13 Ke2 Nfd5 14 Rhcl a6 15 Bxd7+ Kxd7 saw Black take over the initiative and score an eventual victory in V.Stavrakakis-S.Kapnisis, Kavala 2004. b) 6 Be3 threatens dxc5, so Black should respond with 6... cxd4, when 7 cxd4 (7 Bxd4 Nf6 is fine for Black, as the bishop will soon become a target after moves like ...0-0, ... Nc6, ... Rd8,

26

The Hyper-Accelerated Dragon etc) 7... Nf6 should be safe enough for Black. Compared to the dubious line S... cxd4?! 6 cxd4, the white bishop has committed itself to e3, which is not the most active square. 8 Nc3 QaS 9 Be2 0-0 10 0-0 Nc6led to approximate equality in Ni Hua-A.Timofeev, Oropesa del Mar 2000. c) 6 C4!? is slightly more challenging and used to cause some difficulties before the correct methods were found. Black should opt for 6 ... Qd8!. White's idea is to play 7 dS!? (7 dxcS? Qxd1 + 8 Kxd1 Na6 is better for Black) when he hopes that the c4/dS pawn wedge will provide him with the superior prospects, but in the game J.Shaw-A.Dreev, Gothenburg 200S, Black demonstrated the correct antidote with 7... eS! (Diagram 26) 8 Nc3 Ne7, intending a subsequent ... NfS-d4 followed by ... fS with good prospects. Following 9 BgS f6 10 Be3 Na6 11 Be2 NfS 12 0-0 0-0 13 a3 Nd4 14 b4 fS Black already had the better chances and went on to win convincingly. These possibilities illustrate just how easy it is for Black to equalize and perhaps take over the initiative against routine development. White only has one line that can provide any real chances of putting the black position under pressure...

The critical continuation 6 Na3! (Diagram 27)

Now Black must take care, due to the twin possibilities of NbS and Bc4.

Diagram 27 (B)

Diagram 28 (W)

White's only challenging move

The main line

In Diagram 27 Black has two main options. 6... Nf6 is one natural response, but the most popular and most promising move is ... 6... cxd4 (Diagram 28)

It may look risky to spend time on a non-developing move, but tournament practice has

shown the text to be a reliable continuation, if followed up correctly. 7 Bc4

This is critical, although 7 NbS is sometimes employed by players aiming for a safe position with chances for a small edge. Play continues 7... Na6 8 Be3!? (8 Nbxd4 Nf6 is nothing for

27

Starting Out: The Accelerated Dragon White, who has spent three tempi moving the same piece), at which point the rarely seen 8 ... Bd7!? (Diagram 29) deserves attention.

Diagram 29 (W)

Diagram 30 (B)

An effective equalizer?

Where should the queen go?

The game P.Brunelliere-An.Sokolov, French Team Ch. 2003, continued 9 Bxd4 Bxd4 10 Qxd4 Nf6 11 Qxd5 Nxd5 12 Ne5 Nac7 13 Nxc7+ Nxc7 14 Nxd7 Kxd715 0-0-0+ Kc6 16 Be2 Rhd8, when White's advantage was mostly symbolic and Black easily held the draw. This looks like the most straightforward solution, although players wishing for a more complex struggle can also investigate 8... Nf6 9 Qxd4 0-0!?, when 10 Nxa7 Be6 gave Black some compensation for the pawn in M.Kupec-I.Susedik, Slovakia 2003. 7 ...Qe4+ 8 Be3! This excellent move provides the justification for White's play. Now 8 ... dxe3?? loses the queen after 9 Bxf7+!, while 8 ... Nf6? is also unsatisfactory due to 9 Bxf7+! Kf810 cxd4. Therefore Black is practically forced to defend f7 with ... 8... Nh6!

Although this may seem like a concession, the knight is not at all badly placed here and is just one move away from an excellent home on f5. 9cxd4 This is the main line, which will form the subject of Game 3. Instead:

a) 9 Bd3!? Qg4 10 h3 was well met by 10... Qxg2! 11 Rh2 Qxh2 12 Nxh2 dxe3 13 fxe3 0-0 in D.Svetushkin-H.Hamdouchi, Ubeda 1999. Black's rook, bishop and pawn constitute full material equality for the queen, while his superior pawn structure will provide excellent long-term prospects. b) 9 NbS is also possible, when 9 ... 0-0! is Black's best response. Now 10 Nc7?? dxe3 is obviously out of the question, while 10 cxd4 Nc6 reaches a position considered in the note to White's lOth in Game 3. One independent continuation in 10 Bd3!? (Diagram 30). In Diagram 30 Black has tried a few different queen retreats, but I will cut to the chase and say that 10... Qd5! is best, not fearing 11 Nc7 as this is well met by 11...Qa5 12 Bxh6 Qxc713 Bxg7 Kxg714 Nxd4 Nc6 with equality, while if Black wants to unbalance the game he can

28

The Hyper-Accelerated Dragon opt for 12... dxc3!? 13 Bxg7 cxb2+ 14 Kfl Kxg7 15 Nxa8 bxa1Q 16 Qxa1 + f6 with unclear play. White can try 11 Bxh6 Bxh6 12 Nc7, but then 12... Qc6 13 Nxa8 dxc3 14 Qc2 cxb2! 15 Rb1 Bg7 offers Black very good compensation (analysis by Rogozenko). In J.Pinski-T.Markowski, Koszalin 1999, White shied away from these complications in favour of the calm 11 cxd4, after which 11...Nf512 0-0 Nxe313 fxe3 Qd814 Qe2 Nc615 Rad1 e6 was approximately equal. White enjoys greater piece mobility and central control, but Black's bishop pair is a valuable long-term asset. Overall the line with 9 Nb5 certainly deserves to be taken seriously, but it seems that Black can obtain fully satisfactory play.

Illustrative Game Game3

D H.Hamdouchi •

S.Kudrin

FIDE World Championship, Tripoli 2004

1 e4 cs 2 Nf3 g6 3 c3 ds 4 exds Qxds 5 d4 Bg7 6 Na3 cxd4 7 Bc4 Qe4+ 8 Be3 Nh6 9 cxd4 (Diagram 31)

Diagram 31 (B)

Diagram 32 (W)

A critical variation

An untested suggestion

This is possibly the most critical line of the entire 3 ... d5 variation. White simply contents himself with recapturing the temporarily sacrificed pawn before deciding what to do next. NOTE: Although kingside castling may look like an automatic choice for both players (as indeed it should be for Black!), White can sometimes delay castling in favour of the regrouping manoeuvre NbS-c3, improving the knight and harassing the black queen. 9...0-0

In S.Tiviakov-J.Berkvens, Dutch Team Ch. 2006, Black's attempt to seize the initiative with 9... Bg4? backfired badly after 10 Qb3! (breaking the pin while threatening Bd5) 10...e611 Bd3! Qd5 12 Qxd5 exd5 13 Ne5 Be6 14 Nb5 Kd8 15 Rcl Nd7 16 Nc7 when Black resigned.

29

Starting Out: The Accelerated Dragon 100-0

This has been the most popular choice, although the present game provides a persuasive case for Black's chances. White's best try for an advantage may lie with the less common 10 NbS!? Nc6 11 Nc3!, forcing Black to make an immediate decision regarding the future of his queen. 11 ...Qg4 is one option, but then 12 h3! Qd7 (not 12 ... Qxg2?? 13 Rh2 trapping the queen) 13 g4!? Kh8 14 Qd2 Ng8 1S 0-0-0! gave White a very active position in E.MoserZ.Hagarova, Plovdiv 2003. Instead 11 ... QfS seems safer, when R.Juroszek-Z.Hagarova, Trinec 2003, continued 12 Qe2 and now 12... Ng4 looks sensible, followed by ... Nxe3 or ... Nf6, exchanging or improving Black's worst piece. Another possibility is 12 ... QhS!? (Diagram 32), intending ... Bg4 and later ... NfS. The position is very interesting, with chances for both sides. 10... Bg4! (Diagram 33)

Diagram 33 (W) Black develops actively

Diagram 34 (W) Black has full compensation

Also possible is 10... NfS 11 Re1 Nh4!?, but the text is more in keeping with the classical principles of opening play; i.e. bringing all the pieces into the game and avoiding multiple moves with previously developed pieces. 11 Bxh6

The most principled continuation, which used to be considered sufficient for an advantage. Instead 11 NgS?! Bxd112 Nxe4 Bg413 NcS Nc614 h3 Bc8! left Black slightly for choice in F.Maeser-J.Gallagher, Swiss Team Ch. 2003. White's temporary activity is not enough to compensate the weak d-pawn, especially with ... NfS on the way. 11... Bxh6 12 Re1 Qf4!

This sacrifice has revitalized this line for Black. The older continuation 12 ... Bxf3?! 13 Rxe4 Bxd114 Rxd1 BgS 1S ReS! Bf6 16 RbS Nc6 17 dSleft White with a pleasant endgame advantage in E.Sveshnikov-E.Najer, St Petersburg 1998. 13 Rxe7 Nc6 14 Rxb7 Rab8!

The white rook is very active on b7 and so it is highly logical for Black to exchange it. 15 Rxb8 Rxb8 (Diagram 34)

30

The Hyper-Accelerated Dragon Black is temporarily two pawns down, yet he maintains a full share of the chances thanks to his supremely active pieces. 16 Be2? This retreat looks natural, but it was already time for White to start thinking about bailing out to a level endgame: 16 NbS (16 h3 Bxf317 Qxf3 Qxf318 gxf3 Nxd4 is at least equal for Black) 16... a617 Bd5!? should lead to equality after 17... Ne718 Nc3 Nxd5 19 Nxd5 Bxf3 20 Qxf3 Qxf3 21 gxf3 Rxb2 or 17...Rxb518 Bxc6 Rxb219 Qd3 Bxf3!? 20 Bxf3 Bg7. 16... Bxf3 17 Bxf3 Nxd4 18 b3? Missing Black's powerful response, although White is worse even after the superior 18 Rb1 Bg719 Bd5 Rd8 20 g3 Qe5 21 Bf3 Qe6 22 Kg2 Qxa2, as pointed out by Finkel. 18 ... Bg7l (Diagram 35)

Diagram 35 (W)

Diagram 36 (W)

White is in trouble ...

Big trouble!

White evidently overlooked this. The point is that, after the forced reply ... 19Rb1 19 Qcl? Nxf3+ 20 gxf3 Qxcl + 21 Rxcl Bb2 22 Rc7 Bxa3 23 Rxa7 Bd6 would be an easy technical win. 19 ... Be5l ... White is unable to cover h2 as 20 g3? would hang the bishop on f3. 20 Kf1 Qxh2 21 g3 The only way to stave off mate. 21... Nf51 (Diagram 36) Now White must worry about sacrifices on g3. 22 Bg2 Bxg3?1 This looks tempting, but 22 ... Rb6! (Finkel) intending ... Rf6 would have been far stronger, when White would be helpless against the threat of... Nxg3(+). 23Qf3 Now White is able to fight on for a few moves.

31

Starting Out: The Accelerated Dragon 23 ... Bh4?!

Black could still have put the game beyond doubt with 23 ... Rb6! 24 fxg3 Nxg3+ 25 Kel (25 Kf2 Rf6! 26 Qxf6 Ne4+ wins) 25 ... Re6+ 26 Kd2 (or 26 Kdl Re2) 26 ...Qh6+ 27 Kc2 Re2+ etc. 24Rd1

24 Nc4! looks correct, preventing the transfer of the rook to f6. It appears that neither player was particularly tuned in to this possibility. 24... Re8

Once again 24 ... Rb6! looks best. 25 Nc4 Bf6 26 Nd6 Nh4!?

26 ... Nxd6 27 Qxf6 Re6 would maintain Black's advantage, although Kudrin may have been concerned that, in the event of a possible queen exchange, White's bishop could have proven more valuable than Black's knight, especially in conjunction with the queenside pawn majority. 27 Qg3 Qxg3 28 fxg3 Nxg2??

After this blunder the evaluation turns on its head. Black could have maintained a neardecisive advantage with the temporary piece sacrifice 28 ... Rd8! 29 gxh4 Be5 30 b4 Rxd6 31 Rxd6 Bxd6 (Diagram 37).

~

Diagram 37 (W)

Diagram 38 (B)

White faces an uphill struggle

Black must lose his knight

TIP: Positions featuring opposite-coloured bishops are notoriously drawish, especially in endgames in which only kings and pawns remain. This position is an exception due to the favourable (for Black) distribution of pawns.

This will enable him to obtain two connected passed pawns on the kingside, while also keeping a 'spare' queenside pawn- which also happens to be the correct match for his dark-squared bishop. 29 Nxe8

Now it is White who has winning chances! 29 ... Ne3+ 30 Ke2 Nxdl 31 Nxf6+ Kg7 32 Nd5! (Diagram 38}

This is what Black must have overlooked - his knight will be trapped on b2.

32

The Hyper-Accelerated Dragon 32 ... Nb2 33 Kd2 hS 34 Kc2 Na4!

The best chance, making it as difficult as possible for White to create a passed pawn on the queenside. 35 bxa4 gs 36 Kd3 Kg6 37 Ke4 fS+ (Diagram 39)

Diagram 39 (W)

Which way should the king go?

Diagram 40 (B) A popular anti-2 ...g6 system

38 Kes?

Now it is White's turn to miss the win which could have been achieved after 38 Kf3!, which would enable the white king to prevent the enemy pawns from queening while allowing the knight to round up the a-pawn. 38 ... h4 39 gxh4 gxh4 40 Kf4

Now we see the difference - the white king has to waste additional time. 40 ... Kf7

Rushing to the queenside. Black may still appear to be on the brink of defeat, but the remaining moves show that he is just able to survive. 41 Ne3 Ke6 42 as h3 43 Nf1 Kds 44 Kxfs Kcs 45 a6 Kb6 46 Kg4 Kxa6 47 Kxh3 Kbs 48 Kg3 Kb4 49 Kf3 Ka3 50 Ke3 Kxa2 51 Ke4 Yz-Yz

This game and the accompanying notes demonstrate that Black is in reasonably good theoretical shape after 3 ... d5. And if White conducts the opening in lacklustre fashion, Black will often obtain good chances to fight for the initiative at an early stage. TIP: A final bonus is that the present variation can easily be incorporated into a Sicilian player's repertoire as an anti-Aiapin weapon after 2 c3 dS 3 exds Qxds 4 d4 g6!?, when 5 Nf3 Bg7 transposes directly to the lines we have been covering.

Indeed, this is how the majority of games proceeded, although White does possess a tricky alternative in 5 Na3!?, delaying Nf3 for the moment. A detailed examination of this variation is beyond the scope of the present book, but Black should also be ok here. The main

33

Starting Out: The Accelerated Dragon line runs 5 Na3!? cxd4 6 NbS Na6 7 Be3 Bg7 8 Nf3, reaching a position we have considered via the move order 2 Nf3 g6 3 c3 dS 4 exdS QxdS 5 d4 Bg7 6 Na3 cxd4 7 NbS Na6 8 Be3, while 8 Bxd4 and 8 Qa4!? are independent alternatives.

Theoretica I? 3 ... d5 is an ambitious move which can lead to early complications, so it is definitely advisable for Black to spend some time learning how to handle the forcing lines. I believe that your efforts here will be richly rewarded, especially as you can practically kill two birds with one stone by employing the same variation against 2 c3.

Statistics 3 c3 scores a total of 53% for White, but this drops to just 43% after 3 ... d5 4 exd5 Qxd5 5 d4 Bg7. On a precautionary note, it should be remembered that many games in this variation result from the Ala pin move order of 2 c3 d5 3 exd5 Qxd5 4 d4 g6!?, which is frequently used by highly rated players as a means of unbalancing the game against weaker opposition. This might have skewed the statistics in Black's favour slightly, but it's no reason for us not to feel optimistic about Black's chances.

-

The Main Line: 4 Oxd4 1 e4 c5 2 Nf3 g6 3 d4 cxd4 4 Qxd4!? (Diagram 40)

Of the variations examined in the present chapter, 4 Qxd4 is the most popular and probably the most challenging. I consider this to be White's most principled way of fighting against the Hyper-Accelerated Dragon (apart from 4 Nxd4 of course!). By recapturing with the queen White threatens the rook on h8 before Black can respond with ... Bg7. Following the practically forced development of the black knight to f6 White will attempt to harass this piece by advancing his e-pawn, either immediately or after suitable preparation. WARNING: It should hardly need mentioning that the ugly 4...f6?- while not losing by force - is self-evidently a bad move. White's best looks like 5 Bc4 Nc6 6 Qd5! (forcing the following weakening) 6... e6 7 Qd1 when the black kingside is a mess. 4... Nf6

Obviously this is the correct way to deal with the attack on the rook. Now Black plans ... Nc6, highlighting the drawback of White's fourth move. At this point White has three reasonable options: 5 Bb5, 5 e5 and 5 Nc3. I consider the first to be harmless, but the second and third deserve to be taken seriously. Instead, 5 c4?! Nc6 just gives Black a very comfortable version of the Maroczy Bind, while 5 Bg5 Nc6 6 Qe3 Bg7 and 5 Bc4 Nc6 6 Qd3 Bg7 7 0-0 0-0 8 Nc3 d6 both lead to Dragon-type positions in which Black has nothing to worry about.

An Ingenious but Tame Bishop Development 5 Bb5!? (Diagram 41)

34


Diagram 41 (B) Visually shocking, but harmless

Diagram 42 (W) Black's bishops balance his king

This move may appear rather shocking if you are seeing it for the first time! It is quite an ingenious way of preparing e4-e5, but it turns out that Black probably has two ways to obtain a full share of the chances. 5... a6!? leads to unorthodox and unbalanced positions after 6 e5 axb5 7 exf6 Nc6 (7 ...e6!? is also interesting), intending to utilize the doubled b-pawns with ... b5-b4 to restrict the enemy knight. Black's bishop pair should provide full compensation for his fractured pawn structure. This line is perfectly playable for Black, but I have chosen to concentrate on the alternative, which is equally sound yet underestimated.

s... Nc6 This is perhaps Black's most straightforward answer to 5 Bb5, and not only is it objectively sound, I also believe it to be a good psychological choice. In playing 5 Bb5 White has indicated a willingness to enter the murky waters of 5... a6 6 e5, so I can imagine many players feeling less enthusiastic about an early queen exchange. TIP: Many players regard queen less positions as boring and not worth studying, but anyone who holds this attitude is missing out on one of the most deeply fascinating areas of our wonderful game ... not to mention failing to fulfil their potential as a player. 6Bxc6

White has no real choice but to capture the knight. Queen moves have been tried, but in that case what was the point of putting the bishop on b5? 6... dxc6

6... bxc6 is playable, but Black has nothing to fear in the queenless position. 7 Qxd8+ Kxd8 (Diagram 42)

Contrary to the opinions of some commentators, I do not consider Black's chances to be at all inferior in this endgame, as long as he avoids certain pitfalls. Here are what I consider to be the most important features of this position:

35

Starting Out: The Accelerated Dragon 1) White has a lead in development. He has an extra piece in play and will probably castle with check in the near future, thereby gaining a valuable tempo. 2) Black's bishop pair will usually provide him with excellent long-term chances. If Black can complete development without suffering any deterioration in his position (e.g. unfavourable opening of lines, changes in pawn structure, etc) then he will probably gain the advantage. 3) Black will generally aspire to achieve the pawn advance ...e7-e5. This will guarantee him a fair share of space while restricting the white pieces, especially the bishop and the king's knight. 8 Nc3

This is the most natural and almost certainly the best move, even though the knight's own prospects are somewhat restricted by the c6-pawn. In J.Polgar-B.Larsen, Roquebrune 1992, White tried the immediate 8 e5, but her idea was more or less refuted by Black's logical play after 8 ... Nd5 9 Bd2 Bg710 c4 Nb6! (normally the knight would be misplaced here, but Larsen has seen that in the time it takes White to defend c4, he will be able to force a weakening of her structure, after which the knight can reroute itself to a better outpost) 11 b3 Bg4 12 Bc3 Bxf3 13 gxf3 Nd714 f4 Nc5 15 Ke2 Kc7 (Diagram 43), when Black was clearly better, the simple plan being to occupy the d-file and later attack f4 with moves like ... Ne6 and ... Bh6.

Diagram 43 (W) White's strategy has failed

Diagram 44 (B) A potentially critical position

8... Bg7 9 Bf4

This has been by far the most popular choice, but it may not be the best. Indeed, the main line shows that it is White who must fight for equality if Black plays correctly. Alternatives include: a) 9 h3 is too slow to threaten the black position. 9... Nd7! looks best, when 10 Be3 Bxc3+! 11 bxc3 f612 0-0-0 Ke8 (T.Myrvold-I.Rausis, Skei 1995) and 10 Bd2 e511 0-0-0 f612 Be3 Kc7 (R.Sola-A.Rodriguez, Villa Ballester 1999) both seem slightly preferable for Black. b) 9 eS!? may be White's best. Now 9 ... Nd7?! 10 Bf4 makes it difficult for Black to achieve a

36

The Hyper-Accelerated Dragon truly harmonious set-up (10 e6!? is also interesting). Instead I would suggest 9... Nd5, after which 10 Nxd5 cxd5 11 Bf4 (Diagram 44) reaches a critical position for this line. The exchange of a pair of knights is quite helpful to White- he still has one remaining to carry out 'knight duties', whereas Black possesses only 'linear' pieces. White's plan will be to put his pieces on dark squares and aim to stifle the black bishops, while Black should probably aim for queenside counterplay. From the position in Diagram 44, the most accurate move for Black is 1l...Be6!, with a possible continuation 12 Nd4 (12 0-0-0 Kd7 is similar; note that 13 c4? is refuted by the counterpin 13... Rhc8!) 12... Kd7 13 0-0-0 Rhc8 intending a gradual queenside assault. White should probably look for some way to play on the kingside, although 14 h4 can always be met by 14... h6 to keep things closed. 9 ... Nh5! (Diagram 45)

Diagram 45 (W) Equalizing effortlessly

Diagram 46 (B) White's most direct method

9... Nd7 10 0-0-0 Ke8 11 Bc7! was somewhat troublesome in A.Hellmayr-R.Nixon, Groningen 1973, but the text should enable Black to equalize with ease. 100-0-0+

10 Be5 f611 Bd4 was agreed drawn in A.Hait-Y.Meister, Karvina 1992, but I prefer Black after 11...e5, e.g. 12 Bc5 Kc713 0-0-0 b614 Bd6+ Kb7 etc. 10.•. Ke8 11 Be5

11 Bc7 Bg412 Rd3 Bh6+ 13 Kb1 Nf414 Bxf4 Bxf4 gave Black an edge in M.Schlosser-J.Sikora Lerch, Trnava 1989. It is hard for White to turn his lead in development into anything tangible, while the black bishops will be a potent long-term force. 11... f6 12 Bd4 Bh6+ 13 Be3

13 Kb1 e5 14 Bc5 would avoid the doubled pawns, but Black is still more comfortable after 14... Kf7. 13 ... Bxe3+ 14 fxe3 e5

Black had a small but stable advantage in A.Kosten-E.Birmingham, Paris 1988.

37


Theoretical? There is not too much to learn here. I have discussed quite a few possibilities to illustrate how the game may develop, but in general the precise moves do not need to be memorized.

Statistics The statistics in the queenless position after move 7 are perfectly balanced at 50/50. To summarize, I think that Black enjoys a full share of the chances after 5 ... Nc6.

The Immediate Advance of the e-pawn 5 eS (Diagram 46)

This is White's most obvious and direct method and has been the most common choice, accounting for more than 40% of games from this position according to my database. s... Nc6 6Qa4 WARNING: White must keep his wits about him here! A number of players have received a nasty surprise after 6 Qh4? Nxes, winning a pawn due to the queen check on as. White suffers a similar fate after 6 Qe3? Ng4 7 Qe4 Ngxes 8 Nxes Qas+.

On the other hand, 6 Qf4!? Nd5 7 Qe4 is a valid way of transposing to the main line. 6... Nds 7 Qe4 (Diagram 47)

White has occasionally tried 7 Qb3, when play may continue 7... Nc7 8 Bc4 Ne6 9 Bxe6 dxe6 10 0-0 Bg7 11 Bf4 0-0 12 Qe3 Nb4 13 Qe4, as in T.Horvath-K.Meier, Lenk 1994, at which point 13 ... Nd5 would have been logical with approximately equal chances. Black can also opt for the slightly more unbalancing 9 ... fxe6!? 10 Bf4 Bg7 11 Nbd2 0-0 12 Bg3 b5!? with reasonable counterplay, as in J.Dukhin-B.Savchenko, Samara 2004.

Diagram 47 (B)

Diagram 48 (W}

Where should the knight go?

Which bishop to develop first?

Now Black can choose between two knight moves. 7... Ndb4!? is the sharper option, leading

38

The Hyper-Accelerated Dragon to great complications after 8 Bb5 Qa5 9 Nc3 d5 10 exd6 Bf5. Unfortunately I do not completely trust Black's chances here, especially after White's strongest response of 11 Qe5!. I prefer the more reliable alternative: 7... Nc7 This move is solid and less likely to be refuted than 7... Nb4, yet still provides chances of a complex and unbalanced middlegame. 8 Nc3 The usual choice - the knight clearly belongs on this square. 8... Bg7 (Diagram 48) At this point White usually opts for one of two bishop moves: 9 Bc4 and 9 Bf4. The latter is considerably more challenging and will be considered in Game 4. First we will briefly see how Black can achieve a comfortable position against the former. 9 Bc4 This looks quite natural, but it allows Black to begin an immediate counterattack by utilizing some tactical features of the position. 9 ... bs!

This far from new and was in fact first played back in 1898 by Joseph Blackburne. It has stood the test of time and is now considered to bring Black an excellent position. 10Bb3 Accepting the offer brings White nothing but grief: a) 10 Bxb5?! Nxb5 11 Nxb5 Qa5+ 12 Nc3 Rb8 gives Black more than enough compensation for the pawn. b) 10 Nxb5?! is sometimes called a losing blunder in view of 10 ... Nxb511 Bxb5? Qa5+ winning a piece, but it is White's 11th which is the real howler. Instead 11 Qd5 regains the piece due to the attack on f7, though even in this instance Black has a choice of good lines. 11. ..0-0 gives good compensation after 12 Bxb5 Bb7 or 12 Qxb5 Rb8. Also promising is 11...Qa5+!? 12 Bd2 (12 c3 e613 Qxb5 Qxb514 BxbS Nxe5 regains the pawn with an edge) 12...e6 13 Qxb5 (or 13 QcS Qb614 Qxb6 axb6 15 Bxb5 Nxe5) 13 ... Qxb5 14 Bxb5 Nxe5 and in all these lines Black succeeds in recouping the sacrificed pawn, while retaining a slight edge thanks to his central pawn majority. To summarize, 10 Nxb5?! is not a losing blunder in itself, but is still a pretty bad move. 10... Bb7 11 Qe2 11 0-0 Na5 12 Qg4 Nxb3 13 axb3 0-0 14 Re1 a6 was agreed drawn in S.Galdunts-V.Eingorn, Graz 2001. 11... as 12 a4 12 Nxb5 a4 13 Bc4 Nxb5 14 Bxb5 Qc715 Bf4 a3! gives Black good compensation according toADP. 12 ... b4 13 NbS o-o 14 Nxc7 Qxc7 15 Bf4 Nd8! (Diagram 49) and the knight's arrival on e6 ensured Black of an excellent position in S.Brynell-C.Ward, Wrexham 1998. The game continued 16 0-0-0 Ne6 17 Bg3 and now 17... Bxf3 18 Qxf3 Bxe5 19 Bxe5 Qxe5 20 Rxd7 Rad8 21 Bxe6 Qxe6 22 Rxd8 Rxd8 23 Rd1 Rxd1+ 24 Qxd1 Qe4 brought Black slightly

39

Starting Out: The Accelerated Dragon the better of a drawn ending, while 17... Rac8! would have been a straightforward way to maintain a middlegame advantage.

Diagram 49 (W) Black has an excellent position

Diagram 50 (W) Black is cramped, but ok

Theoretical? One of the advantages of 7... Nc7 over 7... Nb4 is the absence of too much heavy theory.

Statistics 5 e5 scores a total of 51% for White. Following the usual5 ... Nc6 6 Qa4 Nd5 7 Qe4, the sharp 7... Nb4 achieves a superficially high 54% for Black, but this is mainly due to a number of games in which White failed to play correctly. If he follows up with the critical8 Bb5 Qa5 9 Nc3 d5 10 exd6 Bf5 11 Qe5! then Black's score plummets to just 23%. 7... Nc7 scores a more modest 39% for Black. After White's strongest continuation of 8 Nc3 Bg7 9 Bf4 my recommendation of 9 ... b6 (see the following game) manages 43%.


0

C.Bauer • V.Colin

French Championship, Besancon 2006 1 e4 c5 2 Nf3 g6 3 d4 cxd4 4 Qxd4 Nf6 5 e5 Nc6 6 Qa4 Nd5 7 Qe4 Nc7 8 Nc3 Bg7 9 Bf4!

This is much more challenging than 9 Bc4 and has been White's most popular choice. 9 ... b6 (Diagram 50)

This looks to me like the most promising continuation. 9 ... Ne6 is playable, but the statistics are not encouraging. Usually I try not to place too much emphasis on them, but when White is scoring 93% you have to wonder!

40


0

NOTE: One of Black's difficult decisions in this line concerns the timing of the move ... Ne6.

In general the knight can exert a good deal of influence on the game from e6. On the negative side, placing the knight here relinquishes Black's control over the important dS-square, and there have been some games in which a white knight has occupied this splendid outpost with powerful effect. After 9 ... Ne6 play may continue 10 Bg3 0-0 (10 ... a6 is too slow and 11 Bc4 b5 12 Bxe6 fxe613 Ng5! 0-0 14 Nxe6 dxe6 15 Qxc6 Bd716 Qe4 ReB 17 Rd1left Black with no real compensation for the missing pawn in B.Macieja-S.Cicak, Warsaw rapid 2006) 11 0-0-0 QaS (Diagram 51), and now 12 h4!? looks dangerous. In G.HernandezJ.Cuenca Jimenez, Albacete 2004, Black went down quickly after 12... Nc5 (12 ...h5!? or 12... b5!? may improve) 13 Qe3 d614 exd6 Be6 15 dxe7 RfcB 16 Bd6 b6 17 a3 Bxc3 1B Qxc3 Qxc319 bxc3 f6 20 Bb5 Kf7 21 Bxc5 bxcS 22 Bxc61-0.

Diagram 51 (W}

Diagram 52 (W}

Ready for mutual king hunting

Creative play by Black

10Qe3 10 h4!? is an untested suggestion of Nielsen and Hansen. 10... h5!? (Diagram 52) Preventing White's intended Bh6. 10... h6 is a playable alternative. The text may appear weakening, but it does have its advantages. For example, Black may be able to follow up with ... Ne6 at some point, when the reply Bg3 can be met by ... Bh6, activating the bishop and harassing the white queen. 11Rd1 11 Bc4 Na5 12 Bd5 Nxd5 13 Nxd5 Bb714 0-0-0 0-0 (14 ... RcB!?) 15 Bh6 Nc4 16 Qf4 ReB 17 Bxg7 Kxg7 was ok for Black in A.Kogan-I.Rausis, French Team Ch. 2004. 11... Bb7 12 Bc4 Na5 13 Be2 Ne6 14 o-o Qc7 14... RcB!? was another way of preparing ... Nc4. 15 Rfe1 Nc4 16 Bxc4 Qxc4 17 Nd4 (Diagram 53) White is essentially playing on the dark squares. He hopes that the b7-bishop will largely

41

Starting Out: The Accelerated Dragon be aiming into thin air, while its brother on g7 will be shut out of the game by the eS-pawn.

Diagram 53 (B) Space vs. two bishops

Diagram 54 (B) A dangerous sacrifice

17 ... Nxd4

17... Nxf4!? 18 Qxf4 0-0 looks more promising to me. Black's bishop pair may not be doing too much at present, but their long-term value should not be underestimated. 18 Rxd4Qc6

The immediate 18...Qe6 was also possible; it is hard to determine which side is more likely to benefit from the insertion of f2-f3. 19 f3 Qe6 20 Qd2

Preventing castling for the moment. 20... Rc8 21 a4

White had a very interesting alternative in 21 Rd6!? QfS 22 Rf6! (Diagram 54) and now: a) 22 ...exf6 23 exf6+ Kf8 24 Qd6+ Kg8 2S fxg7 Rh7 (2S ... Kxg7 26 BeS+ f6? 27 Bxf6+ Qxf6 28 Re7+ wins) 26 ReS Qxf4 27 Re8+ Rxe8 28 Qxf4 intending Qc7 gives White good winning chances. b) 22 ... Bxf6 23 exf6 e6 (23 ...0-0 24 fxe7 Rfe8 2S Bd6 is very good for White) 24 BeS with quite dangerous compensation for the exchange. If these variations really are good for White, then Black should certainly prefer the suggested improvement on move 17. 21 ... Bc6 22 h4

22 NbS is nothing special for White after 22 ... a6 23 Na7 Rc7 24 Nxc6 Qxc6. 22 ... 0-0 23 Re2

23 Bh6? is met by 23 ... BxeS. 23 ... Rc7?!

23 ... Rfd8 looks better, ensuring the survival of the dark-squared bishop. 24 Bh6! (Diagram 55)

42


Diagram 55 (B)

Diagram 56 (B)

Exchanging bishops helps White

White's attack is decisive

24.. .f6?1

Over the past couple of moves Black has lost the thread of the game somewhat, though he manages to recover later. Obviously 24 ... Bxe5?? was no good here in view of 25 Rxe5 Qxe5 26 B£4, but he should have preferred 24 ... Rfc8 25 Bxg7 Kxg7, avoiding further weaknesses and limiting White's advantage to modest boundaries. 25 Bxg7 Kxg7 26 Rd6! Qc4

26 ... exd6? 27 exd6 regains the material with interest. 27 exf6+ exf6 28 Re7+?

Missing a golden opportunity to seal victory with 28 Ne4! Bxe4 29 Rxe4 Qxc2 30 Re7+ (Diagram 56) 30 ...Rf7 (or 30 ...Kg8 31 Qd5+ Kh8 32 Rx£6!) 31 Rx£7+ Kx£7 32 Rxd7+ Kg8 33 Qd5+ Kh8 34 Rd8+ Kh7 35 Qg8+ Kh6 36 Qh8+ Rh7 37 Qx£6 and wins. After the text White is still better, but now he has lost momentum and Colin manages to stabilize his position with some good defensive moves. 28... Rf7 29 Rd4 Qc5 30 Re3 a5! (Diagram 57)

A good move, ensuring that Black will not be overrun on the queenside. 31 Kh2 ReS 32 Ne2 Re7!

Exchanging his own passive rook for White's active one. 33 Rxe7+ Qxe7 34 Nf4 ReS 35 Qd3 f5! (Diagram 58) It may appear anti-positional to fix the pawns on the same colour squares as the bishop, but

there are other more important factors to consider. The text makes far better use of the black queen: instead of being lumbered with defensive duties on £7, it can now remain on the open e-file, while the advance of the £-pawn also forces White to keep a watchful eye on his h-pawn. 36 Nh3 Qe5+ 37 Nf4 Qe1 38 Nh3

White has nothing better than repeating. 38 ...Qe5+ 39 Nf4 Qe140 Nh3 Qe5+ 41 Nf4 Yz-Yz

43


Diagram 57 (W)

Diagram 58 (W)

Black defends well

Black is out of danger

Finally we consider a popular and modern line in which White delays the central push.

Delaying e4-e5 1 e4 c5 2 Nf3 g6 3 d4 cxd4 4 Qxd4 Nf6 5 Nc3!? (Diagram 59)

Diagram 59 (B)

Diagram 60 (W}

Delaying e4-e5 for the moment

A dynamically balanced position

This is a more modern attempt to combat the Hyper-Accelerated Dragon. The evolution of the move is quite easy to understand. In the position after 4... Nf6, White's most obvious fifth move was 5 e5, and it used to be the most popular choice by an overwhelming margin. However, the consensus gradually emerged that Black's position contained sufficient resources, so White began looking for new ways to fight for the advantage. The point of 5 Nc3 is to prepare e4-e5 under more favourable circumstances.

44

The Hyper-Accelerated Dragon s... Nc6

Naturally Black should take the opportunity to harass the queen. 6Qa4!

Other retreats are feeble, e.g. 6 Qd1 Bg7 when Black has an improved version of a Dragon. The point of the text is that after 6... d6, the pin along the a4-e8 diagonal enables him to play 7 es anyway. This is examined in Game 5.

An Offbeat Alternative Before we look at the main line and most reliable continuation of 6... d6, some readers may be interested to learn of a much less well studied yet fully playable alternative in 6... Bg7!?. Then 7 es Ng8 8 Bf4 f6! 9 exf6 Nxf6 (Diagram 60) reaches a little known and dynamically balanced position- a perfect choice for players who like to mix things up while avoiding any heavy theory. White has two main tries: a) 10 Bc4 Qa5 11 Qb3 (11 Qxa5 Nxa5 is nothing for Black to worry about) 11 ... d5! 12 Bb5 0-0 13 0-0 e6 gave Black a nice position in S.Rublevsky-R.Ponomariov, Tomsk (rapid) 2006. His pieces and pawns control a lot of important central squares, and the white queen looks a little odd on b3. b) 10 0-0-0 0-0 11 Bc4+ Kh8 12 Ng5looks critical, when T.Molchanova-A.Elfert, St Petersburg 2004, continued 12 ... d5!? (this isn't forced; Black has a surprisingly safe alternative in 12... Qa5!? when 13 Nf7+ Kg8 leaves White without a particularly convincing discovered check) 13 Bxd5 (13 Nxd5 Qa5 14 Qxa5 Nxa5 15 Nxf6 Nxc4 16 Ngxh7 Rxf6 17 Nxf6 Bxf618 b3 Ne5looks unclear) 13... Nxd514 Nxd5 e5 15 Be3 Nd4 16 Nc3 b5! 17 Qb4 a5 18 Qc5 Bf5 19 Bxd4? exd4 20 h4 ReS 21 Qa7 Rc7 22 Qa6 dxc3 23 b3 Qe7 0-1. Before researching this chapter I had never seriously considered the idea of allowing the knight to be driven back to g8 in this variation. Still, on the limited available evidence it appears quite playable.

Theoretical? There is practically no established theory after 6... Bg7 7 e5 Ng8, and many opponents probably won't even know that it is a 'real' variation.

Statistics 6... Bg7 scores a total of 32%, but the overall statistic is misleading. Following 7 e5 Ng8 8 Bf4, the common yet inferior 8... Nh6?! has yielded a pathetic 14% but the more challenging 8.. .£6! has produced a balanced 50%.

Illustrative Game Games

D J.Rowson •

M.Turner

Edinburgh 2001 1 e4 cs 2 Nf3 g6 3 d4 cxd4 4 Qxd4 Nf6 5 Nc3 Nc6 6 Qa4 d6 7 es (Diagram 61)

45


Diagram 61 (B)

Diagram 62 (W)

A popular position

White's bishops are an asset

White has to play like this, otherwise his preceding moves make no sense. From this point the most combative option is 7... Ng4!?, which can lead to extremely sharp and complicated play. Not surprisingly this move has acquired a considerable body of theory, making it a somewhat impractical choice. TIP: Players wishing to unbalance the game would do better to play the previously investigated 6... Bg7, which leads to equally interesting and promising positions while demanding a fraction of the theoretical preparation required by 7... Ng4!?. 7... dxes

This is much safer and easier to play than 7... Ng4. Black aims to show that the pressure along the a4-e8 diagonal will not be too serious, and simply plans to finish developing and castle. 8 Nxes Bd7! I believe this to be Black's most promising option. White now faces a choice. He can damage Black's pawns with 9 Nxc6, or gain the advantage of the two bishops with 9 Nxd7. 9 Nxd7 This is by far the most popular move; White hopes that his pair of bishops will provide an enduring advantage in this relatively open position. The alternative is 9 Nxc6, after which the obvious 9 ... Bxc6 10 Bb5 gives White a small but pleasant edge. Instead 9 ... bxc6 is a decent move, but my suggestion would be the as yet untested 9 ... Qb6!?, which looks eminently playable. Nisipeanu and Stoica mention the line 10 Be2 Bg7 11 Qa3! leading to a slight edge for White, but 10... Bxc6 looks like an obvious improvement when 11 Bb5 would leave Black a full tempo up on 9... Bxc6 10 Bb5, and after 1l...Bg712 Be3 Qc713 Bxc6 bxc614 0-0 0-0 Black intends ... Nd5 with a reasonable game. 9...Qxd7 (Diagram 62) I must admit that when I first looked at this variation my initial evaluation was that White

46

The Hyper-Accelerated Dragon should automatically be better because of the two bishops. While this assessment carries a grain of truth, closer inspection reveals a number of other factors to be taken into consideration. Black has a slight lead in development and will easily complete the deployment of his forces with ... Bg7 and ...0-0. Just as importantly, the quality of his development is also quite high. The bishop will be perfectly placed on g7, and the knights cover a lot of important squares. Later there will be a simple plan of executing a minority attack on the queenside. As for the loss of the bishop pair, this needn't be so bad after moves like ... e7-e6 and perhaps ... a7-a6/...b7-b5, controlling several important light squares. Having said that, there is no denying that White's position must objectively deserve a slight preference, so let's see how these two Grandmasters attempt to maximize their chances. 10 Be3 Bg7 11 Rdl Qc8

Having lost the light-squared bishop it seems quite natural for Black to deploy his queen on the light squares, although 11...Qc712 Be2 0-0 13 0-0 a6 14 Qh4 Rfd8 also looks quite reasonable. 12 Be2 o-o 13 o-o a6 14 Qh4 Qfs!?

Turner elects to activate his queen immediately. Evidently he regarded this as an improvement over 14 ... Rd8, which he had employed in the earlier game P.Wells-M.Turner, Oxford 1998. 15 Rd2

15 Bd3 could be met by 15... Qa5 or 15 ... Qg4!?. 15 ... Rad8 16 Rfdl Rxd2 17 Rxd2 bS!

Turner continues to play actively. None of Black's moves have been particularly difficult, yet White's bishop pair has not made itself felt in the slightest. 18 a3 Qe6 19 Rdl Qfs 20 Bd3 Qc8

20 ... Qg4!? looks like an even clearer route to equality. 21 Bel Rd8 22 Rel b4 23 axb4 Yz-Yz

A draw was agreed here, probably in anticipation of 23 ... Rd4 24 Qg3 Nxb4 25 Rxe7 Nxd3 26 cxd3 Nd5 27 Rel Nxc3 28 bxc3 Rg4 followed by ... Bxc3 with total equality (but not 28 ...Qxc3?? 29 Qb8+ Bf8 30 Qxf8+! mating).

Theoretical? Once again there is not too much theory to learn here, though it is always helpful to know a few details. The main line of 9 Nxd7 Qxd7 is mostly about positional understanding, and there is little chance of either side suffering a catastrophe in the opening.

Statistics The solid 8... Bd7! has yielded 50%. After the main line of 9 Nxd7 Qxd7 White has managed a total of just 49%, although this goes up to 64% after the strongest plan of 10 Be3 Bg711 Rd1, as seen in Rowson-Turner. The rarer 9 Nxc6!? has scored 57% for White, though the sample size of just seven games is far too small to draw any definite conclusions.

Summary and Conclusions I have no hesitation in recommending 2 ... g6 to Accelerated Dragon players who wish to

47

Starting Out: The Accelerated Dragon avoid the Rossolimo and other anti-Sicilian lines which can occur after 2... Nc6. Over the course of this chapter we have examined a number of ways in which White may avoid the transition to a standard Open Sicilian, but in each case Black seems to be holding his own. The moves 3 h4!7 and 3 Bc4 are not seen too often, but both present their own unique challenges. Fortunately Black can achieve a fully playable position against both of these moves, as long as he combines some basic theoretical knowledge with a few common sense principles. The 'Pseudo-Alapin' variation with 3 c3!7 is an important option, which can also be rather more theoretically intensive than the aforementioned lines. Here I have chosen to focus on the active central strike 3... d5!7, leading to some very interesting positions in which Black has the opportunity to fight for the initiative straight from the opening. As an added boon, this can be employed against 2 c3, a system which many Sicilian players find awkward to meet. The only slight drawback to Black's approach is that a more thorough level of preparation will be required, but in this case I believe the effort to be more than worthwhile. In the final part of the chapter we considered White's most important 'anti-HyperAccelerated' option, namely 3 d4 cxd4 4 Qxd4. Following the standard 4... Nf6 the three main options of 5 Bb5, 5 e5 and 5 Nc3 all present their own unique challenges, but in each case we have seen that Black can obtain a decent share of the chances. Throughout the chapter I have endeavoured to recommend lines that combine theoretical soundness and solidity while still providing some prospects to fight for the initiative. I will conclude by repeating a statement from earlier in the chapter. The move 2 ... g6 is neither 'better' nor 'worse' than 2... Nc6. The choice between them will depend on one's own personal preferences and experiences. Having said that, I would like to make a final, additional point which might influence the decision making process. If, after 2 Nf3, White does not intend to enter the Open Sicilian with 3 d4, then the chances are that his repertoire will be based on the 3 Bb5(+) systems. It follows that if you employ the move order with 2... Nc6, then your adversary will be able to play his favourite line which he is likely to know very well, as 2... Nc6 is a very popular move which can be used to enter many different Sicilians. By contrast, 2...g6 is seen much more rarely, so it follows that the typical 'anti-Sicilian' opponent is unlikely to be so comfortable facing this move. At the very least, they will not have acquired the same level of experience and expertise against a line which they encounter only rarely. Some readers may have particular reasons for preferring 2... Nc6 over 2...g6, or vice versa. But for those who are 'on the fence' or completely new to the Sicilian, I would argue that it makes more sense to base your repertoire on 2...g6, simply because most opponents will be less familiar with the resulting positions. NOTE: From this point onwards the introductory notation in each subsequent chapter will show the standard move order of 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6.

This is in accordance with the classification of the ECO and most other reference books, and does not reflect any sort of personal rejection of the 2... g6 move order. (If anything my personal preference might be slightly the other way, for the reason explained above.)

48

Chapter Two

5 Nc3 - Classical and Other Lines

---

Theoretical Outline: the ••• d7-d5 Thrust A Comparison with the 'Traditional Dragon' Ways for White to Prevent ••• d7-d5 The Nb3 Retreat Delaying the Knight Retreat An Early f2-f4 Push Summary and Conclusions

Starting Out: The Accelerated Dragon 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 (Diagram 1)

Diagram 1 (B)

White forgoes the Mar6czy Bind

Diagram 2 (W) Exploiting the 'Accelerated' move order

In this chapter we will begin to examine the important branch of the Accelerated Dragon in which White declines the invitation to set up the Mar6czy Bind with 5 c4. These 'nonMar6czy' lines almost always involve the move 5 Nc3, followed by one of several methods of development. Occasionally White may employ a different move order such as 5 Be3 Bg7 6 Nc3 (or 5 ... Nf6 6 Nc3), but such deviations are of no more than mere transpositional significance. The queen's knight clearly has no better square than c3, so I will henceforth assume that White utilizes the most flexible move order of 5 Nc3 in order to preserve as many options as possible with respect to the other pieces.

Theoretical Outline: the ... d7-d5 Thrust Beginning from Diagram 1, Black should respond with the standard developing move: 5... Bg7

Aside from being an obviously indispensable developing move, the text also sets up a specific threat against the knight on d4. Now 6 Nxc6? bxc6 would suit Black very well on account of his improved central control and open b-file. 6Be3

6 Nb3!? is sometimes played and will be examined later on. The odd-looking 6 Nde2!? is also seen from time to time. This retreat may look passive but it's not so bad when you understand White's intention. He envisages a positional set-up involving a kingside fianchetto, and may later reroute the king's knight to f4 and perhaps d5. Compare the line 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 g6 6 g3!? Nc6 7 Nde2 against the Dragon. With this being the Accelerated Dragon Black has not yet played ... d7-d6, and I would suggest delaying it for a little longer. After the further moves 6... Nf6 7 g3 it is very interesting to try 7... b5!? (Diagram 2) 8 Bg2 Rb8 9 0-0 0-0 with a nice position for Black. The next move will probably be ... d7-d6, reaching a Dragon in which Black has already assured himself of

so

s Nc3- Classical and Other lines some useful queenside counterplay with ... b7-b5. WARNING: There is an important positional motif which is worth remembering here. If White plays Nds early in the middlegame, Black should be very careful about exchanging knights on d5, as e4xd5 gives White an open e-file and a strong outpost on c6.

Often a better idea is to meet Nd5 with the retreat ... Nd7!, followed later by a timely ... e7e6. (In subsequent chapters we will encounter the identical theme in certain lines of the Mar6czy.) Another good strategy for Black was shown in the game V.Atlas-S.Tiviakov, Chalkidiki 2002. After 10 Nd5 d6 11 c3, the great Dragon specialist began immediate queenside counterplay with 11...b4! 12 cxb4 Nxb413 a3 Nbxd514 exd5 Bb7. Here the above warning does not apply, as White is more concerned about the weakness of his d-pawn. The game continued 15 Nc3, forgoing any hope of getting this piece to its dream home on c6, when Black stood better thanks to his sounder pawn structure. We now return to the most popular move 6 Be3: 6... Nf6!

6... d6 followed by ... Nf6 would reach a traditional Dragon. The text is more accurate, for reasons that will soon become clear. 7 Be2

The most important alternative is 7 Bc4, which is often played with the intention of employing the aggressive Yugoslav attacking formation involving queenside castling. Please refer to Chapters Three and Four for more details. 7 Nxc6!? is another option. The difference between making this capture here compared to the previous move is that the recapture 7... bxc6 can be met by 8 e5, harassing the knight on f6. This and other related variations involving an early Nxc6 exchange form the subject of Chapter Five. 1 ...0-0 8 0-0 (Diagram 3)

Diagram 3 (B)

Diagram 4 (W)

A common position

The Classical Dragon

51

Starting Out: The Accelerated Dragon This natural sequence of moves has been seen in thousands of games, and if you play the Accelerated Dragon on a regular basis you are likely to encounter it time after time.

A Comparison with the 'Traditional Dragon' Before we go any further, the following comparison will help us to understand precisely why Black has every reason to feel optimistic in this particular variation. 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 g6 6 Be2 Bg7 7 0-0 Nc6 8 Be3 0-0 (Diagram 4)

Diagram 4 shows one of the basic positions of the Classical Dragon. At this point White's best and most popular move is 9 Nb3. You may be wondering why this decentralizing move is necessary. One reason is that the natural9 f4 can be strongly met by 9...Qb6!. Ideally White would like to find some way to prepare the f2-f4 advance. 9 Khl and 9 h3 are two such possibilities. The problem with both of these is that Black can obtain a good position with the thematic and excellent move 9 ... d5!.

Back to the 'Accelerated' Version Now let's return to the position shown in Diagram 3. If we keep in mind the analogy with the Classical Dragon, the following should require no explanation. 8... d5! (Diagram 5)

Diagram 5 (W) An important and recurring theme

Diagram 6 (B)

Where to go- e4 or d7?

This move enables Black to solve all of his problems with relative ease. TIP: The ... d7-d5 advance represents Black's most important trump card after 5 Nc3. We will see it time and time again in both this and the following two chapters.

Now 9 exd5 has been the most popular choice, and will form the subject of Game 6. Before looking at this we should also familiarize ourselves with the alternative:

52

5 Nc3- Classical and Other Lines 9 Nxc6 bxc6 10 eS (Diagram 6)

10 exd5?! cxd5 gives Black the advantage thanks to his central pawn majority and potential pressure against the white queenside. In this position Black can choose between two options of approximately equal merit: a) 10... Ne4!? 11 Nxe4 dxe4leads to a position in which Black's active pieces should provide enough compensation for his compromised pawn structure. The game N.Short-J.Yrjola, Manila Olympiad 1992, continued 12 Qxd8 Rxd8 13 Rfd1 Be614 Bd4, and now 14.. .f5!? (Diagram 7) deserved consideration.

Diagram 7 (W)

Diagram 8 (B)

Compare Adams-Khalifman

Black has freed his position

This was played in the famous game M.Adams-A.Khalifman, Las Palmas 1993, with the minor difference of the white king being on h1 instead of g1 (the game was reached via the aforementioned 9 Kh1 d5! variation of the Classical Dragon). Compared with AdamsKhalifman, the position of the king on gl instead of h1 must help White slightly, but the course of the game is highly instructive in any case. Play continued 15 a4?! K£716 a5?! Rxd4! 17 Rxd4 Rb8! with excellent compensation for Black, who went on to win a fine game (perhaps Adams was counting on 17...Bxe5?! 18 Rb4! with advantage to White). b) Although the above seems satisfactory for Black, it is not to everyone's taste to enter an endgame while simultaneously saddling oneself with a broken pawn structure. For that reason, the alternative 10... Nd7 may represent a more straightforward path to equality. Play may continue 11 f4 e6 12 Na4 f6!. TIP: We will encounter this blocked central pawn structure (e5/f4 vs. d5/e6/f7) numerous times over the course of this chapter. In almost all cases, Black's immediate priority should be the removal of the eS-pawn through the undermining move ...f7-f6.

Anyone familiar with the French Defence will know how crucial this resource can be in that opening. In the present variation, with the bishop on g7, it is impossible to overstate its importance. 13 exf6 Qxf6 14 c3 (Diagram 8)

53

Starting Out: The Accelerated Dragon It is easy to see that the quality of the black pieces has improved immeasurably now that

the e5-pawn has disappeared. From here, the game A.Henriquez-D.Barria, Santiago 1994, was agreed drawn after 14 ... Qe7 15 Qd2. While this seems reasonable enough, the positioning of the knight on a4 might have tempted me to open things up with 14 ... e5!? 15 fxe5 Qxe5 with a possible continuation 16 Rxf8+ (or 16 Bd4 Qg5) 16... Bxf8!? 17 Bd4 Qg5 and at least equal chances. Both kings are slightly exposed, but it looks to me as though Black has good prospects with ...Bd6 and perhaps ... Ne5-g4 on the agenda. To summarize, 9 Nxc6 bxc610 eS gives Black nothing special to worry about, with 10... Ne4!? and 10 ... Nd7 both offering a fair share of the chances.


D A.Jain

• A.Greet

Crowthorne 2006

1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Be3 Bg7 6 Nc3 Nf6 7 Be2 o-o 8 o-o ds! 9 exds Nb4!? (Diagram 9)

Diagram 9 (W)

Diagram 10 (W)

Delaying the recapture

A promising sacrifice

Delaying the recapture for one move in order to try to enable a knight to settle on the superb dS-square. The text may come as a mild shock to the average opponent who will likely be anticipating the more mundane 9... Nxd5, which, it must be said, is also fully sufficient for equality (and often more in practice). To be honest the choice is very much a subjective one, so let us embark on a quick diversion to investigate the immediate recapture. In that case, the natural and almost universal continuation has been 10 NxdS QxdS 11 Bf3, when 11...Qc4 has proved to be perfectly reliable, while 11...Qa5!? (Diagram 10) can also be a lot of fun!

54

5 Nc3- Classical and Other lines Now 12 Nxc6 bxc6 13 Bxc6?! Rb8 14 QdS Qc7! gave Black tremendous compensation in G.Basanta-J.Donaldson, Seattle 1987, and many others. White has no chance of retaining his extra pawn and will more than likely end up losing material thanks to Black's powerful queenside pressure. Instead of the greedy 13 Bxc6?!, White should play more conservatively with 13 c3. Then 13 ... Rd8 14 Qc2 Bf515 Be4 Rab8 gave Black a comfortable position in J.Nelson-M.Turner, British Ch., Norwich 1994, although there was certainly no need for White to opt for the overly compliant 16 Bxf5?! Qxf5 17 Qxf5 gxf5 18 Rfd1 Rd5 19 Rxd5?! cxd5 with a big advantage for Black. To summarize, 9... Nxd5 and 9... Nb4!? are both fully playable for Black- the choice is entirely up to you. 10 Qd2 (Diagram 11)

Diagram 11 (B)

Diagram 12 (W)

Timid play from White

Black is already better

There is nothing terribly wrong with this move, but it can hardly be viewed as anything more than an equalizing attempt. If Black is allowed to carry out his plan of recapturing with a knight on d5, he will have good chances to obtain the bishop pair; hence the text move, by which White intends to solve that problem with a subsequent Bh6. Instead, the only remotely challenging option for White here is 10 d6! Qxd6 when there are two main options: a) 11 Ndb5?! (the wrong knight!) enables Black to gain excellent chances after 11 ... Qb8! (Diagram 12). Following the consistent 12 Bc5 Nc613 Bf3 a6 14 Na3 (14 Nd4 Nxd4 15 Bxd4 Rd8 16 Nd5 Nxd5 17 Bxg7 Kxg718 c4 Be6 19 cxd5 Bxd5! won a clear pawn in R.BonhamN.Freeman, Correspondence 1964; while 16 Re1? Ng4! led to a swift demise for White in J.Isepy-J.Gonczi, Sarospatak 1997) 14...Be6 was comfortable for the second player in R.Cardoso-F.Visier Segovia, Las Palmas 1975. He might also have done well to consider 14... Qc7!?, e.g. 15 Nc4 Be616 Bb6 Qf4! 17 Ne3 Qb4! when White has serious difficulties. b) 11 NebS! is better, the point being that after 1l...Qb8 White is able to gain some space with 12 c4. Now 12... a6 is possible, but the most straightforward continuation looks to be 12... Nc6! (Diagram 13) re-centralizing the knight.

55


Diagram 13 (W)

Diagram 14 (B)


Black is already at least equal

The game D.Yanofsky-P.Benko, Stockholm Interzonal1962, continued with the natural13 Nxc6 bxc614 Nd4 Bd715 Bf3 Qc716 Nb3 Rfd8 17 Nc5, at which point Nielsen and Hansen suggest 17... Bg4! 18 Qe2 Bxf3 19 Qxf3 Nd7! with comfortable equality. To conclude, it is fair to say that 10 d6 is the best way to keep some complexity in the game, though even here there is no reason to evaluate Black's chances as in any way inferior. Returning to the game after 10 Qd2: 10... Nbxd5 11 Nxd5 Nxd5 12 Bh6 (Diagram 14) 12... Qb6

This is good enough for equality, although if I had this position again I would prefer 12...e5!. F.Van Hoorne-L.Kavalek, Leningrad 1960, continued 13 Bxg7 Kxg714 Nf3 Nf4! 15 Qe3 f6 when Black already had a slight edge due to his superior central control. 13 Bxg7 Kxg7 14 Nb3 Be6 15 Qd4+ Qxd416 Nxd4 Nf417 Nxe6+ Nxe6 18 Bc4 Nc5 19 Rfe1 e6 {Diagram 15)

56

Diagram 15 (W)

Diagram 16 (B)

A comfortable endgame for Black

White's kingside is slightly weak

5 Nc3- Classical and Other Lines Objectively the position should be fairly level, but personally I would slightly prefer Black. True, White may possess a queenside pawn majority and an alleged advantage of 'bishop over knight in an open position', but it is not so easy for him to utilize those factors. In fact it is hard to suggest a really useful role for the white bishop (thanks in no small part to Black's important extra central pawn), while his queenside pawns may have difficulty advancing without creating weaknesses. The rest of the game illustrates quite nicely how Black can play to win such a position. 20b3

Preventing ... Na4. 20... Rac8 21 Bf1 Kf6 22 Rad1 Rfd8 TIP: When playing for a win in such positions, it is often a good idea to allow the exchange of one, but not both pairs of rooks. 23 g3

24 f4? (Diagram 16)

It may appear natural to gain some space while making way for the king to approach the

centre, but the text is actually a serious positional error. 24... h61 25 Bg2 gS

Highlighting the problem of White's 24th. 26 Rf1 Ke7 27 fxgs?l

There was no need to make this exchange yet, though White may have been worried about being saddled with an isolated pawn on f4. 21 ... hxgs 28 g4?!

Preventing Black from gaining more space with ... £7-fS, but in this case the cure is worse than the disease. 28 ... Nd7

Heading for f4, while gaining time by attacking c2. 29 c4 NeS 30 h3 Nd3 31 Be4 Nf4 (Diagram 17)

Diagram 17 (W)

Diagram 18 (W)

Are bishops better than knights?

A depressing ending for White

57

Starting Out: The Accelerated Dragon What a difference twelve moves can make! The white position has become highly unpleasant. 32 Kh2 Rh8 33 Bg2 RedS 34 Rfe1 f6!

Gradually increasing control over the position, while also emphasizing the lack of constructive moves available to White. 35 Rf1 aS 36 Rfe1 Nd3! 37 Re3 Ne5!

The point of my last two moves is that the threat of ... Nxg4+ now forces White to relinquish the d-file. 38 Rxd8 Rxd8 (Diagram 18}

Now the rook will invade. 39 Re2 Rd4 40 Kg3 Ng6 41 Rc2 Kd6 42 Bf1 Kc5 43 Kf3 Rf4+ 44 Ke2 Re4+ 45 Kf3 Re1 {Diagram 19)

Diagram 19 {W}

Diagram 20 (B)

White is almost paralysed.

A logical retreat

46 Kf2 Ra1 47 Rd2 Nf4 48 Bg2 Kb4 49 Kg3 Rg1 0-1

White resigned, as he is unable to avoid a hopeless king and pawn ending.

Theoretical? Not terribly so. The most important rule of thumb for Black is that if ... d7-d5 can be played, it will almost always be a good move. Once you reach this stage, a bit of knowledge can of course be helpful, but it is more likely to be your opponents who are in need of theoretical advice.

Statistics They speak for themselves: Black scores a very healthy 59% after 9 ... d5!.

Ways for White to prevent ... d7-d5 We have seen that White's most natural scheme of development involving Bcl-e3, Bfl-e2

58

5 Nc3- Classical and Other Lines and 0-0 gives Black no problems whatsoever, thanks to the strong possibility of ... d7-d5. In this section we will turn our attention to some of the ways in which White can fight against this strategy. The first point to mention is that Black can transpose to the traditional Dragon variation by playing ... d7-d6 at any time. However, I will assume that most Accelerated Dragon players will, as a matter of principle, prefer to keep the game in independent territory whenever possible.

The Nb3 Retreat One of the most obvious ways for White to inhibit ... d7-d5 is to perform the knight retreat Nd4-b3, as in the traditional Classical Dragon shown in Diagram 4. This can be done as early as the sixth move: 6 Nb3 (Diagram 20)

The advantage of playing this immediately (as opposed to 6 Be3 followed by a subsequent Nb3) is that White retains the possibility of developing the dark-squared bishop on the active square g5. Despite this extra flexibility there is nothing wrong with the standard development 6... Nf6 followed by ...0-0 and ... d7-d6, leading to a traditional Dragon in which Black has a fair share of the chances. But as I said before, it makes more sense for us to keep the game in Accelerated Dragon territory. That is why I would like to call to your attention the surprising continuation ... 6... Bxc3+!? 1 bxc3 Nf6 (Diagram 21)

Diagram 21 (W)

Diagram 22 (W)

An uncompromising approach

A highly atypical position

White's queenside structure is in ruins, and Black is slightly ahead in development. On the negative side, Black will have to take a certain amount of care to avoid suffering on the dark squares. The next few moves show the correct way to proceed. 8 Bd3

Other moves such as 8 f3 have been tried, but the text is the most natural.

s... ds! 59

Starting Out: The Accelerated Dragon Fighting for the initiative! 8 ... d6?! is nowhere near as challenging. 9 exds Qxds 10 o-o Bfs (Diagram 22) I think it is fair to evaluate the position as unclear with approximately balanced chances. Psychological factors may also come into consideration: after all, I would imagine that most players who opt for 6 Nb3 are probably looking for a quiet game in the early stages, and may find it difficult to adjust to such an irregular position. As matters are far from straightforward for either colour, let's take a quick look at how the game may develop. 11 C4 is sensible, and after 11 ...Qd7 possibilities include: a) 12 Bb2 Rd8led to an eventual draw in L.Arnold-M.Steinbacher, German League 1996, but it looks more logical to evacuate the king from the centre with 12 ... 0-0-0!, when Black is in excellent shape. b) 12 NcS Qc8 13 Bxf5 Qxf514 Nxb7 0-0 is good for Black according to Nielsen and Hansen, but 13 Bh6 requires some precise handling. Best is 13 ... Ng4! 14 Bg7 Qc7! (Diagram 23), preparing to castle, while inducing the following weakening of the white kingside: 15 g3 Rg8 16 Bc3 0-0-0 (16 ... Nce5 led to an eventual victory for Black in D.Josenhans-R.La Flair, Long Island 1992, but I think I would prefer the text) 17 Rb1 b6 18 Na6 Qb7 19 c5?! (perhaps White already needs to think about repeating moves with 19 Nc5 when Black can acquiesce with 19... Qc7 or keep the game going with 19... Qa8!?) 19... Nce5! is very promising for Black, as pointed out by Suba (cited by both Martin, and Nielsen and Hansen).

Diagram 23 (W)

Diagram 24 (B)

A fascinating variation

Preventing ... ds for the moment

Theoretical? This line is pretty sharp, so it wouldn't do any harm to memorize at least some of the above. On the positive side, I can't imagine that many of your opponents are likely to know much about it.

Statistics I was surprised to see 6 ... Bxc3+!? scoring just 37%. Things do improve slightly: if we get as

60

5 Nc3- Classical and Other Lines far as 11...Qd7 then the figure rises to 43%, which is just about acceptable, though admittedly still not great. What can I say? I believe it to be objectively fine for Black, but such an uncompromising approach does not come without risks.

Delaying the Knight Retreat Interesting as the ...Bxc3+!? idea is, it can only be used against the immediate 6 Nb3. Let's turn our attention to the following move order: 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Be2 o-o 8 Nb3!? or 7 Nb3!? o-o 8 Be2 (Diagram 24)

Now 8 ... d5 is prevented, while 8... d6 would reach a traditional Dragon. Although this is theoretically fine for Black, it might still be viewed as something of a moral victory for White. Fortunately there is a good alternative which allows us to remain on Accelerated Dragon territory. 8...as!?

Now White almost always responds with ... 9a4 It can hardly be a good idea for White to allow the pawn to march to a4 and a3, as this

would lead to a serious weakening of the h8-a1 diagonal. The point of Black's play, however, is that after the response ... 9... Nb4! ... he will be able to play ... d7-d5 after all! White's usual response is to castle, and this will be the subject of Game 7. For now we will consider a couple of alternatives: a) 10 eS seems premature, and 10... Ne8 11 f4 d6 begins to undermine the white centre. After the natural 12 0-0, 12... Qc7! looks best, creating a threat against e5 while simultaneously setting up some pressure along the c-file which prevents the c3-knight from moving. b) 10 f4!? is more interesting. White reinforces the e5-square so that after the standard moves 10... d5 11 e5, thee-pawn is already securely defended. Now after 11...Ne4, White may be able to retain an edge with 12 NbS! f6 13 exf6 Bxf6 14 c3 Nc6 15 0-0 as seen in M.Crepan-M.Coklin, Slovenian Ch. 1991. Perhaps the most principled try for Black is 11...Bf5!? (Diagram 25), as seen in the game F.Fiol-A.Meszaros, Eger 1990, which continued 12 Nd4 (12 exf6 Nxc2+ 13 Kf2 Nxe314 Kxe3 Bxf6leads to a position in which the two pawns and exposed position of the white king should offer Black decent practical compensation for the piece) 12... Ne4 13 Nxe4 Bxe4 14 c3 Nc6 15 0-0 g5!? 16 g3 gxf4 17 gxf4, and here 17... Kh8 would have been logical, when the white king looks slightly uncomfortable.

Illustrative Game Game 7

0 V.Gurevich • B.Chatalbashev Cappelle Ia Grande 1998 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Nb3 o-o 8 Be2 aS 9 a4 Nb4 10 0-0

61

Starting Out: The Accelerated Dragon This is the main line. As usual, the point of Black's opening strategy is revealed after the thematic move: 10... ds! {Diagram 26) 11 es

Diagram 25 (W}

Diagram 26 {W}

Complicated, but balanced

That move again!

11 exd5 is rarely seen and should not be feared. J.Faure-A.Adorjan, Geneva 1987, continued 11...Nfxd512 Nxd5 Nxd5 13 Bd4 Nf4 14 Bxg7 Kxg715 Bf3 e5 16 Qxd8 Rxd8 17 Rfd1 and now 17... Rxd1 +! 18 Rxd1 Bf5 19 Rd2 ReS 20 c3 b6 would have left Black with slightly better chances. His pieces are active, while his pawns are doing a fine job of restricting the enemy knight. 11... Ne4 12 f4! {Diagram 27)

Diagram 27 (B)

Diagram 28 (B)

A critical position

What would you play as Black?

This is the most challenging move. Instead 12 Nxe4 dxe4 13 Qxd8 Rxd8 gives White noth-

62

5 Nc3- Classical and Other lines ing; e.g. 14 Bb6 (14 Rad1 Rxd11S Rxd1 BxeS16 Rd8+ Kg717 c3 Nc618 Re8 Rb8 saw Black untangle his position while remaining a pawn up in J.Maskova-M.Botan, Moscow 1994) 14... Rf8 (this looks simplest, although 14... Rd7!? and 14 ... RdS!? could also be considered) 1S NxaS (1S c3 may be better, though Black looks to be fine anyway) 1S ... Nxc2 16 Racl Nb4 17 :'>lc4?! Rxa4 and Black went on to win in I.Jurcisin-A.Cvicela, Nove Zamky 1999. This is a very important position for the 8 ... aS variation, in which Black must choose between three contrasting options. 12 ... sts

This logical move can give rise to some highly complex and fascinating positions, though I must say that it would probably not be my personal choice. a) 12 ... gS!? has only been tried once, but is certainly worth considering if you enjoy murky positions. D.Barua-L.Yurtaev, Commonwealth Ch. 2000, continued 13 Nd4!? (13 fxgS BxeS looks unclear, as does 13 Nxe4 dxe4 14 Qxd8 Rxd8 1S Bb6 Rf8; the aS-pawn may be dropping, but Black has pressure against c2 andeS) 13 ... gxf4 14 Bxf4 Nc6 (another idea is 14 ... Qb6!? 1S Kh1 Nc6 to keep the queens on) 1S Nxc6 bxc616 Nxe4 dxe4 17 Qxd8 Rxd8 18 Bc4 BfS with a complex endgame. Martin evaluates it as equal, whereas I would say that Black's shaky pawn structure may give White the upper hand. At the same time matters are far from straightforward, and in the game Black went on to score a hard fought victory. b) I think my personal preference would be for 12... Nxc3, which leads to a very interesting struggle after 13 bxc3 Nc6 14 Bf3 e6. White has some pawn weaknesses, but his pieces are quite active and he controls more space. I think that Black should be fine, as long as he makes the undermining move ... f7-f6! his top priority. For example: bl) 1S BcS (Diagram 28) looks tempting, and was first played in S.Hamann-J.Collado Galarza, Las Palmas 1973. In the four games in my database which reached this position Black replied with the predictable 1S...Re8. There is nothing terribly wrong with this, but when I first looked at the position I instinctively and immediately felt that 1S.. .f6! was crying out to be played. Very often my snap judgements tum out to be spectacularly wrong, but on this occasion further investigation revealed this exchange sacrifice to be fully justified. Following 16 Bxf8 Qxf817 exf6 Qxf6 Black will pick up the pawn on c3, after which his splendid minor pieces will provide more than sufficient compensation for the tiny material investment. I rather enjoy making positional exchange sacrifices and would certainly relish playing the black side here. b2) Another sensible move is 1S Nd4, after which Black should again retaliate with the undermining move 1S .. .f6!. G.Scholz Solis-P.Olivier, French Team Ch. 1990, continued 16 exf6 Bxf617 NbS, and now the regrouping 17... Ne7! (Diagram 29) 18 Bd4 NfS, followed by completing development with ... Bd7 and ... Rc8, looks complex but approximately equal. We now return to the game continuation of 12 ... BfS!?. 13 Nd4!

This centralizing move must surely be the best, defending c2 while placing the knight on its optimum square. 13 ... Rc8 (Diagram 30)

Logically bringing another piece into the game. 14Ncbs

14 NxfS?! gxfS would only strengthen Black's centre.

63

Starting Out: The Accelerated Dragon 14...f6!?

Diagram 29 (W)

Diagram 30 (W)

A double-edged middlegame

Another tense position

The consequences of this move are by no means clear, but it is definitely the most principled and consistent continuation. In other words, if analysis demonstrates this move to be bad, then the whole line with 12... Bxf5 should probably be avoided in favour of one of the alternatives that we explored at move twelve. 15 exf6 15 e6!? looks like a serious and possibly critical alternative. White now threatens to trap the bishop with 16 g4, so Black must play 15... Nd6 to make an escape square on e4. The position looks double-edged, but I personally don't trust it for Black. One fairly forcing possibility is 16 c3 Nc617 Nxd6 Qxd618 Nxf5 Qxe6!? 19 Bf3! Rfd8 20 Bb6 (Diagram 31) and on this occasion I have my doubts as to whether Black can obtain enough compensation for the exchange.

64

Diagram 31 (B)

Diagram 32 (W)


A tough battle

5 NC3- Classical and Other lines 1S ... Bxf6 16 c3 Nc6 (Diagram 32)

Typically for this line, the position is complex and hard to evaluate. Computers tend to prefer White slightly, but in a practical encounter it is probably just unclear. 17 Bf3 Kh8 18 Kh1 es!?

An interesting attempt to increase the scope of the black pieces. 19 Nxfs

For what it's worth, a few minutes with Fritz indicates that White could have gained an advantage with 19 Bxe4! dxe4 20 NxfS Qxd1 (or 20 ...gxf5 21 Nd6) 21 Raxd1 gxfS 22 BcS!. 19... gxfs

Now it looks like Black is holding his own, and could easily gain the upper hand in the event of lacklustre play by White. 20 Bxe4 fxe4 21 fxes Nxes 22 Qhs Bg7 23 Nd4 Qd7 24 NfS Nc4 25 Bh6 (Diagram 33)

Diagram 33 (B) Heading for a perpetual

Diagram 34 (B) An important variation

Both sides have played fairly consistently over the last few moves. For a game involving a complex opening and some highly uncompromising middlegame play, it is somewhat ironic that the contest will shortly peter out to a draw by perpetual check. 2S ... Bxh6 26 Nxh6 Qg7 27 Nf7+ Kg8 28 Nh6+ Kh8 29 Nf7+ Kg8 30 Nh6+ Yz-Yz

The end of this game may have been an anticlimax, but we have obtained plenty of food for thought along the way!

Theoretical? 8... a5 can lead to some highly complicated positions, so a degree of theoretical knowledge will certainly be beneficial.

Statistics 8... a5 scores a reasonable 47% for Black.

65


An Early f2-f4 Push Here is White's final weapon in his fight against the ... d7-d5 break. With the f-pawn already on f4, he will be ideally placed to meet ... d7-d5 with e4-e5. And compared with the previous section (with an early Nb3), the knight will be better off on d4 in the ensuing Frenchlike structure. The key position arises after the following moves: 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Be2 0-0 8 f4!? (Diagram 34)

Compared with the harmless line 8 0-0 d5, White is now able to advance e4-e5, gaining space and keeping his centre intact.

How should Black Respond? 8... d6 is one option, reaching a traditional Dragon. Then 9 0-0 Qb6! is known to be at least equal for Black. A more challenging alternative is 9 Nb3! followed by 0-0, or even an early kingside expansion with g2-g4!?. Black should have enough resources in either case, but the entire variation with 8 ... d6 belongs to the traditional Dragon, and I would not be making the best use of my already limited space if I discussed it in full detail here. What other moves should we consider? Having mentioned that 8 ... d6 9 0-0?! Qb6! gives Black excellent chances, it is natural to investigate the same idea a move earlier: 8...Qb6?!. Unfortunately it does not have the desired effect. White easily obtains the advantage after 9 e5! Ne8 (in case of 9 ...Qxb2?, R.McCart-H.Doel, Canberra 1993, the most clinical response would be 10 Ndb5! when Black is unable to parry the twin threats of 11 exf6 and 11 Rb1, winning the queen) 10 Nxc6 Qxc6 (10 ... Qxe3? 11 Nxe7+ Kh812 Ncd5 is dreadful for Black) 11 Nd5 Kh812 Bf3 (Diagram 35), when White's active pieces gave him a clear advantage in K.Grosar-M.Ankerst, Kranj 1999.

Diagram 35 (B)

Diagram 36 (W)

Not recommended for Black

The consistent and principled move

What other moves can be considered? A few strong players have tried 8... e5!?, but this can hardly be sufficient for equality (though it is probably not as bad as it appears). Instead Black should stick with the more promising and highly thematic central strike:

66

5 Nc3- Classical and Other Lines B.••ds!? (Diagram 36)

The ... d7-d5 thrust has been the central theme of this chapter, so the text would certainly make for a coherent repertoire choice. It may also, objectively, be the most promising move. 9e5

Now that White has played f2-f4, he had better avoid an exchange of the central pawns. Note that 9 Nxc6 bxc6 10 e5 NeB would just give White an inferior version of the main line, in which he has needlessly strengthened the black centre by exchanging on c6. 9... Ne8!

I do not fully trust 9 ... Ne4 10 Nxe4 dxe4 11 Nxc6 bxc6 as Black must surely be a bit worse due to his numerous pawn weaknesses. The text may look passive, but after investigating it further, I consider it to be fully playable. 10Bf3

White almost always takes this opportunity to pressurize the d5-pawn. After 10 0-0 Black could opt for the murky 10... Nc7!? 11 Qd2 g5!?, but I would prefer the standard undermining move 10.. .f6! 11 exf6 Bxf6!? (Diagram 37), when possibilities include ... Ng7-f5, or ... e6 and ... Nd6. The position holds approximately equal chances for both sides.

Diagram 37 (W)

Diagram 38 (W}

A complex and balanced position


After 10 Bf3, Black must decide how to defend the e5-pawn. 10... Nc7 is one option, but the knight is rather passive here and White should be better after 11 Qd2. The following game will demonstrate a more promising path for the second player.

Illustrative Game GameB

0

E.Hintikka • A.Koponen

Kuopio 1989

1 e4 cs 2 Nf3 Nc6 3 Nc3

67

Starting Out: The Accelerated Dragon This move order is often used by players who wish to avoid the Sveshnikov variation, but it need not be of concern to fans of the Accelerated Dragon. 3... g6 4 d4 cxd4 5 Nxd4 Bg7 6 Be3 Nf6 7 Be2 o-o 8 f4!? ds!? 9 es NeB! 10 Bf3 e6! (Diagram 38) Compared with 10... Nc7, the benefits of the text move are obvious. First and foremost, the d5-pawn is securely defended. Secondly, following the inevitable ... f7-f6, the e8-knight will be able to come to a much more active home on d6. This is all great news, but the one drawback of 10...e6 is that White can- and must, if he is hoping for any sort of advantagewin an exchange with ... 11 Nxc6 The less ambitious 11 0-0 f6! 12 exf6 Bxf6led to double-edged play in D.Schneider-J.Mont Reynaud, Nashville 1998, but 12...Qxf6! would have been more logical, after which Black can complete development with ... Nd6 and ... Bd7, obtaining a fine position. 11 ... bxc6 12 Bcs f6! (Diagram 39)

Diagram 39 (W)

Diagram 40 (W)

Deja vu!

Is Black's compensation sufficient?

The similarity between this and Diagram 28 is quite striking. Personally I love to play like this, sacrificing an exchange in return for the initiative and superior quality of position.

TIP: From a psychological point of view I think this type of strategy can be especially useful when playing with the black pieces. As White we tend not to anticipate having to defend in the early stages, which can make it more difficult to adapt to that situation. 13 Bxf8 Kxf8 13 ... Bxf8 was also possible, but I think Hintikka was correct to keep the bishop on its best diagonal. After the text Black's compensation for the exchange consists of: 1) A monstrous dark-squared bishop, which- in the short term, at least- is likely to exert a far greater influence over the game than either white rook. 2) A solid wall of central pawns which control a lot of important squares and restrict the white minor pieces.

68

5 NC3- Classical and Other Lines 3) Better quality of pieces. Black's remaining rook will enjoy an open b-file, and his knight may play an important role from d6. The white king is also slightly less safe than its opposite number. At the same time, I must be careful not to overstate Black's case. Though the above factors should not be underestimated, we should equally not forget that an exchange is an exchange, and if Black follows up in lacklustre fashion then he could easily end up in a lost endgame. 14Qd4 It is hard to decide whether White should give preference to this or the alternative 14 exf6 Bxf6 (Diagram 40). [14 ... Qxf6 is also playable, the only disadvantage being that it allows

White to castle without hindrance, compared to the main line where ... Qb6+ is a constant menace. Please also note that 14... Nxf6?! would be a mistake, as the knight will have far brighter prospects on d6.] In general, the aim of this book is not to provide reams of complicated analysis. However, this variation is a particularly critical one, which also happens to have undergone almost no practical testing. For those reasons, the following analysis will attempt to probe a little more deeply than usual. As far as I can see, the main possibilities are as follows: a) 15 0-0? should be avoided on account of 15 ... Qb6+ 16 Kh1 Qxb2. b) 15 Qd2 should probably be met by 15 ... Ba6!? (15 ... Nd6 16 Be2looks slightly better for White), and now 16 0-0-0 Nd617 b3 Qe7! threatens ... Nc4 with good attacking chances, while 16 Be2 Bxe2 17 Nxe2 Qb6 also provides decent compensation. c) 15 Rb1!? may be White's best; the long diagonal is a constant source of worry for him, so the text will probably be necessary sooner or later. Now Black's possibilities include: cl) 15 ... Qb6 was the first move I looked at, but it seems that 16 Qd2 Nd617 b3 Rb8 (17 ... Nc4!? 18 bxc4 Bxc3 regains the material, but 19 Qxc3! Qxb1+ 20 Bd1 Kg8 21 0-0 leaves Black vulnerable on the dark squares) 18 Na4 Qb419 Qxb4 Rxb4 20 g3 results in a position in which Black's compensation is not quite sufficient.

Diagram 41 (W)

Diagram 42 (B)

Black has decent compensation

Trying to stabilize the centre

69

Starting Out: The Accelerated Dragon c2) 15... Ba6! is better. Following the logical 16 Be2 (White has to try and get castled at some point) 16... Bxe2 17 Nxe2 (17 Qxe2 Qb618 Nd1 Nd6looks like decent compensation: White is still unable to castle, while ... Re8 and ... e6-e5 is threatened) 17... Qb618 c3 Nd619 Nd4 Nf5! (Diagram 41) 20 Nxf5 (20 Nxe6+? Kg8 is too dangerous) 20 ...exf5 21 Qd2 Re8+ 22 Kfl Re4 23 g3 Qa6+ 24 Kf2 Qxa2 25 Ra1 Qb3, Black's extra pawn and active pieces should provide enough for the exchange. I hope the above variations will give the reader a reasonable impression of how the game may develop after 14 exf6 Bxf6. Although it is too early to give a definitive theoretical verdict, I think it is fair to say that most players would probably not enjoy the white side of these positions. Returning to the game after 14 Qd4 (Diagram 42): 14...Qc7?! This move is tricky, but objectively it seems to be an error. Instead 14 .. .fxe5! looks best, reaching a similar position to the game, except that the black queen is better off on d8 than c7, as will become clear. After the natural sequence 15 fxe5 Kg8 16 0-0, Black has a strong resource available in 16... Nd6! 17 Rae1 Nf7!, when the e5-pawn will come under heavy fire. The black position is rock-solid, and it is hard to suggest a useful plan for White. Black, on the other hand, can aim to win the e5-pawn and eventually set the central pawns in motion I think I would take the black position if given the choice. 15 0-0 There now follows a sharp tactical phase in which White should come out on top. 1S .. .fxes 16 fxes Bxes 17 BxdS+ Kg8 (Diagram 43)

Diagram 43 (W)

Diagram 44 (W)

White must play precisely

Black threatens ... Qh4+

18 Qcs?! Correct was 18 Qf2! Qg7 (18 ... Bxh2+? can hardly help Black after 19 Kh1 Bd6 20 NbS!, whil€ 18 ... Nf6? 19 Ne4! is also hopeless) 19 Bxc6! Bd4 20 Bxa8 Bb7 21 Bxb7 Bxf2+ 22 Rxf2 Qxb7 23 b3, resulting in an endgame in which White enjoys a material advantage. 18 ... Bxh2+ 19 Kh1 Bd6 20 Qxc6 Qe7 (Diagram 44) 21 Ne4?? One of those inexplicable blunders. Instead, there were two decent moves which would have maintained the delicate equilibrium of the position:

70

5 Nc3- Classical and Other lines a) 21 Rf3 exd5 22 Nxd5 (22 Qxd5+ Kg7) 22 ...Qh4+ (22 ... Qe5? 23 Qxe8+! Qxe8 24 Nf6+ K£7 25 Nxe8+ Kxe8 26 Rei+ Kd7 27 Rdl Rb8 28 c4! leaves Black unable to prevent the loss of one of his bishops) 23 Kgl Qh2+ 24 Kf2 Qh4+ looks like ending up in a draw by perpetual check. b) 21 g3!? exd5 (2l...Rb8!? is possible, but 21...Bxg3? 22 Qxa8! Qh4+ 23 Kg2 Qh2+ 24 K£3 escapes the checks) 22 Rael Bb7! (forced; anything else loses instantly) 23 Qb5 Be5 (23 ... d4+? 24 Kglleaves Black without a good follow-up) 24 Rxe5 Qxe5 25 Qxb7 Nc7 26 Kh2 with an unclear, but approximately equal semi-endgame. 21 ... Qh4+ 22 Kgl Bh2+ 0-1

White resigned, as it is mate in two. Once again, an exciting game was marred by a somewhat anticlimactic finish, but the first twenty moves contained several fascinating and critical moments.

Theoretical? To be honest, there is not too much to memorize here: 8 ... d5, 9... Ne8, 10... e6, sac the exchange and have fun playing the position! It may be helpful to learn a few of the finer points which can occur later on, but the most important thing is to understand where Black's compensation comes from, and how best to make use of it.

Statistics 8... d5 scores an impressive 58% for Black, which rises to an even better 64% after 9 e5 NeB.

Summary and Conclusions The Classical system involving an early Be2 has always been a fairly popular choice against the traditional Dragon. White intends simply to develop his bishops in the centre, castle on the kingside, and aim for a safe middlegame with a slight initiative. It is, therefore, completely natural that a lot of players should attempt to carry out the same strategy against the Accelerated Dragon. In the first part of this chapter we learned about Black's chief trump card in this variation, namely the ...d7-d5 break. If White develops naively with Be3, Be2 and 0-0 then we know that Black can equalize almost effortlessly with 8 ... d5! as in Game 6. White's more sophisticated strategies all involve some kind of prophylactic measure to prevent, or take the sting out of ... d7-d5. In the event of an early Nd4-b3 retreat, it is very interesting for Black to adopt the uncompromising exchange ... Bxc3+, followed by a quick ... Nf6 and ... d7-d5. If he delays with Be3, Be2 and only then Nb3, then Black should react with 8 ... a5!, threatening to harass the knight with ... a5-a4. After the almost universal reply 9 a4, Black can continue with 9... Nb4 and 10... d5 anyway. Finally White can aim to strengthen his influence in the centre by means of f2-f4. In that case Black should once again proceed with ... d7-d5, retreating the knight to e8 after the inevitable e4-e5 advance. The golden rule of these French-like structures is that Black must make it his mission to destroy the e5-pawn by means of ... f7-f6. We have seen two occasions in this chapter in which Black was willing to sacrifice an exchange just to achieve that goal. I hope to have convinced the reader of the merits of a well-timed positional sacrifice. The ability to recognize such opportunities can play a pivotal role in a player's success, and I hope that the present chapter has shed a few rays of light on this subject.

71

Chapter Three

Yugoslav Accelerated Dragon: Introduction

----

Background Information the Traditional Dragon How does the 'Accelerated' Version Differ? An Unplayable Line for White Another Sub-standard, but Common Move Order The Correct Move Order for White 7••• Qas!?- Forcing Kings ide Castling The Knight Retreat Keeping the Bishop on the a2-g8 Diagonal Summary and Conclusions

Yugoslav Accelerated Dragon: Introduction 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 (Diagram 1)

Diagram 1 (W) A major branch ofthe Accelerated

Diagram 2 (W) A critical line of the Dragon

In the previous chapter we looked at a sensible but fairly timid scheme of development for White involving Be2 and 0-0. The system discussed in this chapter is much more ambitious. The Yugoslav Attack is an aggressive anti-Dragon system involving the moves Be3, f2-f3, Qd2, 0-0-0 (Bc4 is optional depending on circumstances), followed by h4-h5, opening the hfile with a potentially deadly attack.

Background Information -the Traditional Dragon To understand precisely what White is aiming for, we should first familiarize ourselves with the Yugoslav Attack against the traditional Dragon. Everything which follows later in the chapter can be understood much more clearly when viewed in the light of this comparison. Here is the relevant variation from the traditional Dragon: 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 g6 6 Be3 Bg7 7 f3

This is much more aggressive than the Be2 systems which we encountered in the previous chapter. White is planning Qd2, long castling, followed by opening the h-file in an effort to checkmate the black king. NOTE: The move f2-f3 is almost always necessary in order to prevent ... Ng4. If White allowed the e3-bishop to be exchanged for a mere knight, then the g7-bishop would become tremendously powerful. 7...0-0

Black can delay castling, but this, the old main line continuation, will enable us to draw a parallel with the Accelerated Dragon version encountered in this chapter. 8 Qd2 Nc6 (Diagram 2)

At this point White can choose between two major options: 9 0-0-0 and 9 Bc4. Theory has favoured both moves at different times, but in general it is fair to say that both can lead to

73

Starting Out: The Accelerated Dragon immensely complicated play, with the choice largely depending on personal preference.

Option 1 9 0-0-0 (Diagram 3) is perhaps the most obvious move, when 9... Nxd4 10 Bxd4 Be6 is one

major possibility, but current theory suggests that White can stay on top here. Therefore most experts agree that Black should take the opportunity to fight for the centre with 9...dS!?. I hope that the mere mention of this move will cause the proverbial light bulb to flash above the head of the reader, and we will indeed find that in the positions resulting from our 'Accelerated' move order, careless play by White can allow Black to save a crucial tempo in certain variations by playing ... d7-d5 in one move.

Diagram 3 (B)

Diagram 4 (B)

White castles long immediately

Preventing ... d6-d5

Option 2 The positions occurring in the 9 0-0-0 d5 variation are somewhat irregular and often hard to evaluate, and a lot of players prefer to prevent the possibility with 9 Bc4 (Diagram 4), developing another piece and postponing castling for another move or two. Here most games continue 9 ... Bd7 10 0-0-0 or sometimes 10 h41?, leading to an ultra-sharp middlegame in which both sides will attempt to checkmate the other.

How does the 'Accelerated' Version Differ? The first point for both sides to remember is that, just as in the previous chapter, the ... d7d5 break can sometimes save Black a full tempo compared with the traditional Dragon. The following section will demonstrate the classic example of this theme.

An Unplayable Line for White 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 f3?!

74

Yugoslav Accelerated Dragon: Introduction This is a perfect example of how easy it is for White to go wrong. Yes, this move is an essential part of the 'Yugoslav' attacking system, but in this variation timing is everything. Instead 7 Bc4! is the only correct move order, as we shall see later. 1 ...0-0 (Diagram 5)

Diagram 5 {W)

Diagram 6 (W)

White already has problems!

Black has gained a full tempo

All of White's moves have been completely natural. In fact, they are precisely the same as those which constitute his most highly regarded system against the Dragon. Nevertheless, the surprising truth in Diagram 5 is that White may already be struggling to equalize! It all becomes clear after the continuation ... 8Qd2 8 Bc4 is also possible, and will be considered later. The text is frequently employed by players who like to use the 9 0-0-0 line against the tradition Dragon (as shown in Diagram 3). But anyone using it against the Accelerated sequence is in for a nasty shock ... 8... d5! (Diagram 6)

At this point, I invite the reader to refer back to Diagram 3- 'Option 1'- showing the position after 9 0-0-0 in the Yugoslav. As I mentioned there, the main line for Black is the principled central strike 9 ... d5!?. Diagram 6 shows the same position- except that White has not yet castled! In other words, by playing ... d7 -dS in one move instead of two Black has saved a whole tempo.

0

NOTE: To gain an extra tempo after only eight moves is quite an achievement under any circumstances. When we are dealing with a sharp, tactical variation involving opposite-sided castling, the importance of every move increases enormously.

It would be too much to claim a large advantage for Black in Diagram 6, but at the same time it is certainly White who will have to work hard to equalize the game. 9 0-0-0 It seems unlikely that White has anything better. The alternative is 9 Nxc6 bxc6 and now:

75

Starting Out: The Accelerated Dragon a) 10 0-0-0 Qa5 (10 ... Rb8 is also promising) gives Black a strong attacking position. In N.Seki-F.Visier Segovia, Nice 1974, White attempted to win a pawn by tactical means with 11 exd5 cxd5 12 Nxd5, but after 12... Qxa2 13 Nxe7+ Kh8 14 Qb4 Be6 15 Qa3 Qxa3 16 bxa3 Rfe8 17 Nc6 Bd518 Nb4 Rxe319 Nxd5 Nxd5 20 Rxd5 Rxa3 the white position was in ruins. b) 10 e5 Ne8! 11 f4 (S.Beshukov-S.Dvoirys, Elista 2001, continued 11 Bh6 Qc712 Bxg7 Nxg7 13 f4 f614 exf6 Rxf615 g3 Bg4 16 Be2 Bxe2 17 Nxe2 e5 18 0-0-0 e4 when the central pawn wedge gave Black the advantage) 11 .. .£612 exf6 Nxf6 (Diagram 7) 13 Bd4 (13 0-0-0?! is bad after 13 ... Ng4!, e.g. 14 Bd4 e5! 15 fxe5?? Bh6 winning the queen) 13 ... Rb8 14 Bd3 (14 0-0-0 Rb4 gives Black a strong initiative; an exchange sacrifice could well be on the cards!) 14... Rxb2 15 Bxa7? (White had to try and struggle on a pawn down; the text is almost suicidal) 15 ... Nh5!? (15 ... Ng4! intending ... Qa5 also looks tremendously powerful) 16 Ne2 Nxf4! 17 0-0 Nxd318 Qxd3 was R.Zelcic-D.Feletar, Pula 2000, and now 18... Bf5! would have been the most efficient route to victory, although 18 ... Rxfl+ also proved more than sufficient to bring Black the full point.

Diagram 7 (W)

Diagram 8 (W)

Black stands clearly better

White must struggle to equalize

After 9 0-0-0 we reach a topical theoretical Dragon tabiya ... except that it is Black to move instead of White! 9... Nxd4 (Diagram 8)

9... dxe4 is also perfectly fine and gives Black some chances of an advantage if followed up correctly. However, the text is even more straightforward. 10 Bxd4

This is the usual reaction, although others have also been seen: a) 10 Qxd4 Nxe4 11 Qxd5 Nd6 12 Bd4 Be613 Qg5 was played in N.Upadhyay-R.Saptarshi, Calcutta 2000, and now 13 ... Bxd4 14 Rxd4 Qb6 would have left Black with a slight initiative. b) 10 e5!? (Diagram 9) has hardly ever been played in this position, but it is probably no worse than the alternatives. Obviously this move could come as quite a shock, so we should take a quick look at some lines. Incidentally, if we were to place the white king on b1 instead of c1 then we would

76

Yugoslav Accelerated Dragon: Introduction arrive at a very well-known variation of the 9 0-0-0 Dragon. The following brief analysis will show how the placement of the king can make a big difference ... The simplest solution is probably 10... Nf5 11 exf6, when Black has two tempting options: bl) 1l...exf6 12 Nxd5 (12 Bc5? is a good move in the analogous variation with the king already on b1, but here it is refuted by 12... Bh6! 13 f4 Qc7!) 12... Nxe3 13 Qxe3 Be6 is fine for Black.

Diagram 9 (B)

Diagram 10 (B)

Visually shocking, but harmless

Black can sacrifice the queen!

b2) Also strong is 11...Bxf612 Nxd5 (Diagram 10), when 12... Qxd5! 13 Qxd5 Nxe314 Qd2 Nxd115 Qxd1 Be6 is a well-known queen sacrifice in the 9 0-0-0 Dragon, in which the move Kb1 would already have been played. In this position, the attack on the a2-pawn will probably compel White to play 16 Kb1 anyway, giving Black a full extra tempo in a line which is already considered to give him reasonable chances. Now the strongest continuation looks like 16... Rfd817 Qe1 Rd6 followed by ... Rad8 when Black's beautifully placed pieces more than make up for the small material deficit. Returning to the position after 10 Bxd4: 10...dxe4 11 fxe4 11 Nxe4 Nxe4 12 fxe4 Bxd4 13 Qxd4 Qc7 gave Black a slight edge in K.Buerkle-A.Dietrich, Wittlich 1980, due to his superior pawn structure. 11... Be6 12 Kb1 Qas (Diagram 11) The primary threat is ... Nxe4 exploiting the vulnerability of a2, while ... Rfd8 is a secondary idea to establish an awkward pin. Therefore White has no real choice but to try and simplify the position with Nd5. He can do this with or without a preliminary exchange on f6, but in both cases he must struggle for equality. 13 Bxf6 13 Nd5 also leaves White struggling after 13... Qxd2 14 Nx£6+ (14 Nxe7+? Kh8 15 Rxd2 Nxe4 is terrible for White) 14 ...Bxf615 Rxd2 Rfd8 (alternatively 15... Bxd4 16 Rxd4 Rfd817 Rxd8+ Rxd8 18 Bd3 Kg7 19 Rf1 h5 left Black slightly for choice in M.Andersen-S.Jorgensen, Esbjerg 2006) 16 Be3 (or 16 c3 Rd6 intending to double up) 16... Rxd2 17 Bxd2 Rd8 18 Bd3 Bc4!? 19

77

Starting Out: The Accelerated Dragon Bxc4 Rxd2 gives Black a nice ending; opposite bishops or not, White will have to suffer to make a draw here. 13... Bxf6 14 Nds Qa4! (Diagram 12) Now the game T.Abergel-J.Goupe, French Junior Team Ch. 1996, continued 15 Bd3 (Black would win a pawn after 15 Nxf6+?! exf6), and here 15 ... Bxd5! 16 exd5 Rfd8looks very nice for Black, who can look to organize an attack on the white king.

Diagram 11 (W)

Diagram 12 (W)

Black threatens ... Nxe4!

White's problems persist

TIP: It is well known that opposite-coloured bishops can exert a drawish influence in certain types of endgames. In the middlegame, however, they can be highly conducive to a successful attack on the enemy king.

That concludes our coverage of the variation 7 f3?! 0-0 8 Qd2 d5!. This variation is often (and not unjustifiably) written off as 'simply bad for White'. Indeed, from a theoretical viewpoint it certainly should be! At the same time, I am conscious of the fact that many readers may have little or no knowledge of the analogous variation of the Dragon. And when you are sitting at the board with the clock ticking, it is not enough simply to know that you have a good position -you still need to be able to play good moves! The other reason why I have paid so much attention to such a poor line is that a lot of players do actually play like this as White! I found well over a thousand games in this variation (many of them involving highly rated White players who should have known better!), so there is a very good chance that you will encounter it at some time or another. After familiarizing yourself with this section, you will have an excellent chance of punishing your opponent's sloppy play. WARNING: If you are reading this chapter with a view to employing the Yugoslav Attack as White, then you have no choice but to aim for the Bc4 version shown in Diagram 4. If you try to reach Diagram 3 (with 9 0-0-0 instead of 9 Bc4), your opponent will be able to play ... d7-d5 in one move.

78


Theoretical? Even though this is a substandard line from White's point of view, I would urge Black players to learn at least a few of the main lines presented here. That way you will maximize your chances of punishing anyone who unsuspectingly stumbles into this opening pitfall.

Statistics From Diagram 6 (after 8 ... d5!) Black has scored an impressive 69%.

Another Sub-standard, but Common Move Order The following has also been seen time and time again: 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 f3?! 0-0 8 Bc4 (Diagram 13)

Diagram 13 (B)

Diagram 14 (W)

Another inaccurate sequence

White must be very careful

At first glance this looks much better than the line with 8 Qd2, since ... d7-d5 is now prevented, but Black has another way to exploit the temporary 'looseness' of the white position. Note that this could also arise via 7 Bc4! (the correct move!) 7...0-0 8 f3?! -here White should play 8 Bb3, for reasons that will soon become obvious. 8... Qb6! (Diagram 14)

8 ... d6 would reach a normal Dragon, but the text is far more troublesome. White can probably maintain equality if he plays accurately, but he will have to kiss the Yugoslav checkmating attack goodbye. I still have painful memories of an early experience on the white side of the position shown in Diagram 14. I was eleven or possibly twelve years old, and was competing in an important closed tournament against some other top English juniors. Facing a boy named David Tompson (who has a few games on the database from the late 1990's, but who seems to have given the game up since then) I declined the invitation to play the Mar6czy Bind, thinking that I would be able to achieve an easy transposition to a normal Dragon. The text

79

Starting Out: The Accelerated Dragon move came as a total shock, and I suffered a swift, painful and humiliating defeat. I don't have the game score and can't remember precisely what happened, but as far as I can recall the continuation was 9 Qd2?! Nxe4! (Diagram 15). WARNING: If you play these positions as White, you should always be wary of the discovered attacks ... N(x)e4 and ... N(x)g4. As Black, you should constantly be on the lookout for opportunities to implement them!

Diagram 15 (W)

Diagram 16 (B)

Black is winning material

This b-pawn is truly poisoned!

I can't remember any more details of the game (they do say that the brain sometimes represses painful memories!), but I went on to suffer a quick and humiliating defeat. In any event, in Diagram 1S White probably has nothing better than 10 fxe4 (10 Nxc6? Nxd2 11 Nxe7+ Kh8 12 Bxb6 Nxc4 is practically winning for Black, while 10 Nxe4 Bxd4 also leaves White a clear pawn down) 10... Bxd4 11 Bxd4 Qxd4 12 Qxd4 Nxd4 13 0-0-0 Nc6 14 Nd5 Kg7 when White had no real compensation for the pawn in R.lbrahimov-A.Konstantinov, Novosibirsk 1962. Instead of 9 Qd2, White's best move is probably ... 9 Bb3!

Here is summary of the alternatives: a) 9 Nxc6?? is obviously no good due to 9 ... Qxe3+. b) 9 NfS?? Qxb2leaves White unable to avoid the loss of a piece. c) 9 Na4?? Qb4+ wins the bishop on c4. d) 9 NebS?? also loses a piece after 9 ... a6! 10 NfS QaS+ 11 Bd2 Qd8. e) 9 Qd3?? loses after 9... NeS10 Qe2 Qxb2. f) 9 BbS? has been tried, but 9 ... Nxe4! 10 NdS QaS+ 11 c3 Nf6 leaves Black a safe pawn up. g) 9 0-0?! loses a pawn for insufficient compensation after 9 ... Qxb2, e.g. 10 NebS Qb4 11 Qe2 QaS. h) 9 a3!? (Diagram 16) is probably White's only truly playable alternative to 9 Bb3. The idea is that 9 ... Qxb2?? 10 Na4 will win the queen.

80

Yugoslav Accelerated Dragon: Introduction TIP: This method of defending a b-pawn by indirect means (i.e. meeting the move ... Qb6 with a2-a3, or for Black, meeting Qb3 with ... a7-a6) crops up in many different openings and is well worth remembering.

Black should also avoid 9 ... Nxe4? as 10 NdS! gives him some problems. The calm 9 ... d6, as played in P.Keres-B.Larsen, Beverwijk 1964, is perfectly alright, as White is unable to make good use of the opposition of bishop and queen. However, I think that the strongest move for Black is 9 ... Qc5! (sidestepping a potential NdS attack), when 10 Bb3 (10 Ba2 Ng4! is similar) 10... Ng4! 11 fxg4 Bxd4 12 Bxd4 Nxd4 left Black with a slight edge due to his better structure in R.Freitag-F.Drill, German League 1993. Returning to 9 Bb3!: 9... Ng4!? (Diagram 17) Also possible is 9... Nxe4 10 NdS QaS+ 11 c3 NcS 12 Nxc6 dxc6 13 Nxe7+ Kh8 14 Nxc8 Raxc8 with equality. The text is more ambitious.

Diagram 17 (W}

Diagram 18 (W}

Striving for an advantage

A good ending for Black

10fxg4

10 NdS fails to impress after 10... Nxe3 11 Nxb6 Nxd1 12 Nxa8 Nxb2 13 c3, when Black can choose between 13 ... Be5!? (P.Simacek-A.Saric, Split 2002) keeping the knight trapped on a8 (though capturing it is another matter), or the more straightforward 13 ... Nxd414 cxd4 Bxd4 (Diagram 18) when Black stands better. He has two pawns for the exchange plus a powerful pair of bishops, whilst the white knight on a8 is badly misplaced. 10... Bxd4 (Diagram 19) Black is quite happy to allow the exchange of these bishops. Compared with a normal Yugoslav attack, he is in no danger of being checkmated on the kingside. 11 Bxd4 11 NdS QaS+ 12 Bd2 Qd8 13 c3 Bg7 was seen in Liu Dede-I.Rogers, Singapore 1998. Black can follow up with ... d7-d6 and ... NeS, with a slight advantage. 11... Qxd4 12 Qxd4 White has no particularly enticing way to avoid the endgame, so he should settle for mis-

81

Starting Out: The Accelerated Dragon placing the black knight. 12... Nxd4 13 NdS Nc6 (Diagram 20)

If Black can find a way to complete development without making any concessions, then he will simply stand better. White suffers from a permanently damaged pawn structure, and if the black knight can settle on e5 it will become the best piece on the board.

Diagram 19 (W)

Diagram 20 (W)

Exchanging bishops suits Black

Structure vs. development

14g5!?

14 Ba4 Kg715 0-0-0 Rb816 h3 b5 17 Bb3 d6 saw Black consolidate his position in Y.Sakharov-L.Stein, Kiev 1964. From here, a logical plan would be ... Be6 and, if permitted, ... Bxd5 followed by ... Ne5 when the discrepancy between the quality of the minor pieces would be quite striking. 14... d6 15 Ba4 Kg7 16 0-0-0 (Diagram 21)

82

Diagram 21 (B)

Diagram 22 (B)

Black is fine if he takes care

White's only correct move order

Yugoslav Accelerated Dragon: Introduction We have been following the game S.Sulskis-N.Andrianov, Los Angeles 2003, which continued 16 ...Bg4?! 17 Rd3 Rac8 18 Rb3 Na519 Ra3 Nc6 20 Bxc6 bxc6 21 Ne3 Be6 22 Rxa7 and White won a pawn, which he later converted to victory. The move that Black would like to play would be 16... Bd7, but here that would allow 17 Nxe7. 16... Rb8? (intending ... b7-b5) would also be a mistake due to 17 Bxc6 bxc6 18 Nxe7 when Black is unable to trap the knight. Instead, my suggestion is 16...Rd8!. The idea is simply to prepare ...Bd7 without allowing the reply Nxe7. Following 17 Rhfl (17 Bxc6? bxc618 Nxe7? Bb7 really does leave the knight in trouble) 17... Bd7, intending ... Ne5, Black seems absolutely fine, and may have some chances to press for a slight advantage thanks to his superior pawn structure. Once again, we have spent a considerable amount of time looking at what can only be described as a non-critical, and probably just plain inferior option. The reason is the same as before: despite being an inaccurate move order, 7 f3?! 0-0 8 Bc4 remains a frequent visitor to tournament practice, a database search revealing over a thousand games, including plenty of titled players.

Theoretical? I echo my comments from the previous section: it is not enough to know that your opponent has played an inaccurate move order- you also need to know how to punish him! To be fair, after 7 f3?! 0-0, 8 Bc4 is a slightly better move than 8 Qd2, and while 8... Qb6! should give Black at least equality, certain lines require a bit of precision. So once again, a bit of additional theoretical knowledge can really help to maximize your chances.

Statistics 8...Qb6! has scored a whopping 73%.

The Correct Move Order for White Finally, after familiarizing ourselves with the frequently played yet inferior alternatives, we can investigate White's most accurate way of aiming for his desired Yugoslav formation. 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! (Diagram 22)

If White is aiming for a Yugoslav Attack, then he cannot do without the development of this bishop on c4. The text is necessary to inhibit ... d7-d5, while there is no need to rush with f2-f3, as Black is not yet threatening ... Ng4.

From Diagram 22, the main continuation is 7...0-0, after which White should play 8 Bb3!, taking the sting out of a possible ... Qb6 and preparing to reach the desired Yugoslav formation with f2-f3 and Qd2. This complex and critical variation will form the subject of Chapter Four. For the time being, we will turn our attention to a very important alternative:

7...Qas!?- Forcing Kingside Castling 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf67 Bc4 Qa5!? (Diagram 23)

83


Diagram 23 (W) A solid choice

Diagram 24 (W) White has problems

The text may appear peculiar at first glance, but it carries one extremely attractive feature, namely that White is practically forced to castle on the kingside, and will therefore have to forgo his plan of attacking down the h-file. This is a significant achievement, although the drawback is that the queen has committed herself to the aS-square at a very early stage. 8 0-0

There is no sense in trying to avoid this, as shown by the following brief analysis: a) 8 Qd2? loses material after 8... Nxe4! 9 Nxc6 Qxc3!. b) 8 f3?! Qb4! 9 Bb3?! (9 Nxc6 should be preferred, though 9... bxc6 is pleasant for Black: an exchange on c6 is almost always helpful for the second player due to his improved central control and newly opened b-file) 9 ... Nxe4! 10 Nxc6 (10 fxe4 Bxd4leaves Black a safe pawn up) 10... Bxc3+ 11 bxc3 Qxc3+ 12 Ke2 dxc6! leaves Black two pawns up, as capturing on e4 would lose the queen. c) 8 Nb3?! is a mistake when played before castling, as after 8 ... Qb4! (Diagram 24) White has problems. From here the least of the evils is probably 9 Nd2!? (White is not helped by 9 BdS Nxe4! or 9 Bd3 Nxe4!) 9...Qxb210 NbS (threatening to trap the queen with c2-c3), though even here 10 ...QeS 11 f4 Qb8 12 eS NxeS!? 13 fxeS QxeS gave Black three pawns and a strong initiative for the piece in V.Kupreichik-V.Veremeichik, Minsk 1973. Returning to 8 0-0 (Diagram 25): Black has achieved a small 'moral victory' in forcing his adversary to settle for short castling after all. White probably stands slightly better thanks to his superior development and active pieces, but Black is fairly solid and this variation may be a good choice for players who wish to achieve a safe position from which they can gradually try to outplay the opponent, without having to memorize too much theory. 8 ... 0-0 (Diagram 26)

It should come as no surprise for you to hear that the gluttonous 8 ...Qb4? brings Black

nothing but misery after 9 Bb3 Nxe4? (9 ...0-0 should be preferred, but then what was the point of his eighth?) 10 Nxc6 bxc6 (10 ... dxc6 changes nothing) 11 a3! Nxc3 12 Qf3! Ne2+ 13 Kh1 and Black resigned in L.Romanescu-A.Popescu, Calimanesti 2000.

84


Diagram 25 (B)

Diagram 26 (W)

No Yugoslav attacks today!

Knight or bishop to b3?

At this point White must choose between two principal options. He could expel the black queen from its active position with 9 Nb3, angling for a slightly improved version of a Classical Dragon. On the other hand he could opt for the standard prophylactic retreat 9 Bb3, again leading to a slightly improved version of a well-known Dragon line involving Bb3 and 0-0, due to the premature deployment of the black queen to a5. Both options give White chances of a slight edge, though this should not necessarily deter the reader from playing the black side of the position. The early queen development may not be ideal against short castling, but it is hardly disastrous either. We will look at both of White's options in tum.

The Knight Retreat 9 Nb3 Qc7 (Diagram 27)

Diagram 27 (W)

Diagram 28 (W)

Another choice for White

White struggles to equalize

85

Starting Out: The Accelerated Dragon Now White can choose between 10 f4 and 10 Bg5!?. Both moves present their own unique challenges, and will be covered in Games 9 and 10 respectively.

Keeping the Bishop on the a2-g8 Diagonal 9 Bb3

This is a very logical move. Compared with 9 Nb3 White maintains his bishop on its most active diagonal, and avoids retreating the knight from the centre. WARNING: The latter point is something of a double-edged sword; the knight may be active on d4, but it is also tactically vulnerable. Both sides should be on the lookout for ... Nxe4 ideas at every turn.

9... d6 There is little chance of ... d7-d5 becoming a possibility, so the text is a necessary part of Black's development. Now ... Ng4 is threatened, hence White's next... 10h3

10 f3 has also been played, but does not combine so well with short castling. Black should continue with the logical10 ... Bd7 11 Qd2 Rfc8, intending the typical queenside counterplay with good chances. 10 h3 is better because White may wish to play commence active operations in the centre with f2-f4. 10... Bd7

This is a typical starting position for the 9 Bb3 variation. Now the main line is the active and aggressive 11 f4 (Game 12), while the more positional 11 Re1 can be found in Game 11. Please note that 11 Qd2? is a mistake due to 11...Nxd4! (Diagram 28). [The right move order; instead 11...Nxe4?! 12 Nxc6 Nxd2 13 Nxa5 Nxf114 Kxfllooks slightly better for White.] Then White has no choice but to settle for 12 Qxd4 (12 Bxd4? Nxe4! wins material) 12... Ng4, after which the long and relatively forcing sequence 13 Qd5 Qxd5 14 Nxd5 Nxe3 15 Nxe7+ Kh816 fxe3 Rae817 Nd5 Bxb218 Rab1 Be519 Nf6 Rd8 20 Nxd7 Rxd7 21 Bd5 b6 (S.Khmelnitsky-D.Ionescu, Bucharest 1997) results in a near-equal endgame in which Black is slightly for choice thanks to his better pawn structure. WARNING: When the black queen is on as, the move Qd2 can often leave White tactically vulnerable against ... Nxe4 ideas, so both sides should constantly be on the lookout for this possibility.

Illustrative Games Game9

0

J.M.Degraeve • P.Lerch

French Team Championship 2005

1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! Qas!? 8 o-o o-o 9 Nb3 Qc7 10 f4 d6 11 Be2

White usually sees fit to remove this piece from the firing line of the black queen, even if

86

Yugoslav Accelerated Dragon: Introduction there is no immediate threat. 11... b6! (Diagram 29) This move has two points in its favour. The most obvious is that the light-squared bishop will be very well placed on b7. The second point is that later in the middlegame Black can often make use of the resource ... Nc6-a5!?, allowing the compromising of his pawn structure in order to open the b-file. We will see some examples of this plan in the notes to the present game.

Diagram 29 (W)

Diagram 30 (W)

The double fianchetto works well

A double-edged middlegame

12 Bf3 This is the most natural and common continuation, although the more aggressive 12 g4!? has also been seen. Play normally continues 12... Bb713 g5 Nd714 Nd5 Qd8 (Diagram 30) 15 Rb1 with a complex and unclear position. White has a space advantage and kingside attacking chances, but it will take some time to organize a truly dangerous assault. Meanwhile Black is very solid and can drive the knight away with ... e7-e6, followed by ... Nc5. In Diagram 30 White has sometimes tried sacrificing a pawn with 15 f5!? Bxb2 16 Rb1, although it is hard to say whether this will provide full compensation. 12 ... Bb7 13 Qd2 Again this has been the most common, though 13 Qe1!? and 13 Qd3!? have also been tried by top players. Generally speaking, all three queen moves lead to broadly the same types of positions, and Black should proceed in similar fashion with moves like 13... Rac8 or 13 ... Na5!? according to personal preference. 13 Nd5 is a more direct attempt to seize the initiative. White's idea is that the newly changed pawn structure after 13 ... Nxd5 14 exd5 will enable him to obtain pressure along thee-file. Indeed, Black would be in trouble were it not for the strong and thematic resource 14 ... Na5! 15 Nxa5 bxa5 (Diagram 31), leading to a position in which Black's doubled a-pawns are compensated by his impending pressure along the open b- and c-files. The game E.Vasiukov-T.Giorgadze, Tbilisi 1973, continued 16 c3 Qc4 17 Qb3 Ba6 with balanced chances and an eventual draw.

87


Diagram 31 {W)

Diagram 32 {W)

The b-and c-files give counterplay

A recurring theme

TIP: The idea of playing ... Nas, by which Black purposefully saddles himself with doubled pawns in order to open the b-file, is an important theme which should be remembered, as it can arise in many different Sicilians. 13 ... Na5!? (Diagram 32) Once again we see the same idea. However, while very interesting, the text is not the only playable move. Those looking for a less committal approach may prefer the straightforward 13... Rac8, leading to a tense and rich middlegame with chances for both sides. 14 Nxas bxas 15 Rad1 Rfc8 This is arguably the most principled and best move in the position. 15 ... Rfd8 also deserves consideration, in order to discourage a possible e4-e5 advance; the drawback is its relative passivity, and White gained an edge with 16 Bd4 Rac8 17 Qe3 in B.Bachmann-A.Gundrum, Correspondence 1996. 16 Bd4 Bc6 17 es!? {Diagram 33) White is in no mood to hang about! Players preferring a more measured build-up could opt for something like 17 Qe3. In that case White could claim a slight advantage, although in practical terms both sides will have their chances. 17 ... Ne8? This passive move is the first real mistake of the game. 17... Bxf3 should have been preferred, when 18 Rxf3 dxe5 19 fxe5 (19 Bxe5 achieves little after 19... Qb6+) 19... Ng4looks pretty unclear. White is rather active, but I don't see a knockout blow and there is always a risk that his e-pawn could become a target in the endgame. 1s Nds! This energetic move is clearly the best. 18 ... Bxds 18 ... Qb7 would give White a pleasant choice between the safe pawn grab 19 Qxa5, or the calm 19 c4!? with a clear positional advantage. 19 Bxds Rab8 20 Qe3 Qxc2 21 fS! {Diagram 34)

88


Diagram 33 (B)

Diagram 34 (B)

Energetic play

A bad situation for Black

Excellent play, blasting the position open. 21 e6?! f5 would enable Black to keep things blocked up for the time being. 21... gxf5

Probably the best defence, as Bxf7+ was a serious threat. For example, after 2l...Nc7 there is 22 Bxf7+! Kxf7 23 fxg6+ Kg8 24 exd6 exd6 25 Bxg7, winning back the piece with a neardecisive advantage, as 25 ... Kxg7?? 26 Qe7+ forces mate. 22 exd6?!

This is not altogether bad, but it is probably not the best way for White to capitalize on his superiority. Perhaps the most accurate was 22 Rd2!? (forcing the queen to a less active position) 22 ... Qa4 (22 ...Qc7 23 e6 changes little) 23 e6!, when Black probably has to play the ugly 23 .. .f6. In that case he may have avoided an immediate kingside catastrophe, but his once prized 'Dragon' bishop is a sorry sight. 22 ... Bxd4 23 Rxd4 Nxd6 (Diagram 35) 24 Qg5+

24 Qxe7 is less good due to 24 ... Qc5!. 24... Kf8 25 Qh6+ Ke8?!

Better was 25 ... Kg8 26 Rh4 Qxb2! when the queen covers the key h8-square from afar. The black position looks perilous, but there doesn't seem to be a forced win. A possible continuation is 27 Qxh7+ Kf8 28 Bxf7 Nxf7 29 Qxf5 Qb6+ 30 Khl Qf6 31 Rhf4 Qxf5 32 Rxf5 Kg8 33 Rxf7 Rbl 34 Rxbl Kxf7 when the rook ending should be a draw. 26 Qg7 Kd7 27 Bxf7 Rxb2 (Diagram 36) So far the game has been quite lively and entertaining, and now 28 Bb3 would have kept up the pressure. What happens instead is rather puzzling to say the least! The game score shows that White, a strong Grandmaster, blunders by allowing a mate in two! However, it turns out that Black, himself rated over 2300, is in no mood for such an easy finish and instead opts to continue the fight! I am not sure what to make of this - perhaps a data error could be to blame, as it seems almost inconceivable that two strong players could both miss what follows. Either that, or they were both terribly short of time.

89


Diagram 35 (W)

Diagram 36 (W)

Dangerous, but no forced win

Strange events are in store ...

28 Re1?? Rb1?? 29 Be6+ Kc6 30 Qe5? Rxe1+ 31 Qxe1 Rb8 32 Rd1 Qc5+ 33 Kh1 Rb2! 34 Bb3 Ne4 35 Bd5+ Kb5 36 Bxe4 fxe4 37 Qxe4 Qc6?? 38 Qe5+ 1-o

Despite the comedy of errors at the end, this game was very useful in demonstrating some of the attacking and positional motifs that can crop up in this line. Game10

D

5.Karjakin • V.lvanchuk

European Championship, Warsaw 2005

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! Qa5!? 8 o-o o-o 9 Nb3 Qc7 10 Bg5!? (Diagram 37)

90

Diagram 37 (B)

Diagram 38 (W)

An active and tricky variation

Unusual, but playable for Black

Yugoslav Accelerated Dragon: Introduction This is less popular than 10 f4, but has been the choice of some very strong players including Anand. The idea is to 'exploit' the early development of the black queen on c7 by menacing Bxf6 followed by Nd5. 10... a6!? Another idea is 10... e6!?, stopping any Nd5 ideas once and for all. It looks a little risky to weaken the d6-square, but so far no-one has shown a really convincing way to take advantage of this. D.Pruess-J.Donaldson, Santa Monica 2004, continued 11 Qd2 a612 Bf4 Ne5 13 Be2 b514 f3 Bb7 (Diagram 38} with an interesting position, somewhat reminiscent of a Kan Sicilian (2 Nf3 e6 3 d4 cxd4 4 Nxd4 a6) in which both sides have their chances. 11 Be2 It makes sense to retreat the bishop out of harm's way. The immediate 11 Bxf6 Bxf612 Nd5 achieves little after 12... Qe5 (this is the reason why Black has delayed ... d7-d6). In case of 11 a4, Black can follow a similar recipe as in Chapter Two by playing 11 ... Nb4!, e.g. 12 Be2 d5! 13 exd5 Rd814 Bf3 Be6 15 Qe2 Nfxd5 16 Nxd5 Bxd5 17 Qxe7 Qxe7 18 Bxe7 Nxc2 (Diagram 39).

Diagram 39 (W)

Diagram 40 (W)

leading to an equal endgame

Bishops or pawn structure?

This position was reached in two separate games of an internet match in 2005. The first encounter G.Guseinov-K.Aleksander, continued 19 Bxd8 Nxa1 20 Bxd5 Rxd8 21 Bxf7+ Kxf7 22 Rxa1 Bxb2 when Black's strong bishop and queenside majority gave him the advantage, although the game was eventually drawn. On the next occasion Guseinov tried the superior 19 Rad1, and gained a clear advantage after 19... Rd7?! 20 Bh4 Nb4 21 Nc5, which he later converted to victory. But if Black had opted for the superior 19... Re8! he would have had no problems whatsoever. 11...d6!? A very interesting idea from the ever-creative Ukrainian. The previous top level encounter V.Anand-L.Van Wely, Wijk aan Zee 1999, had continued ll...e612 Qd2 b513 a3 Rb814 Rfd1 with a clear advantage to the pride of India, who went on to win a nice game. 12 Bxf6 exf6 (Diagram 40)

91

Starting Out: The Accelerated Dragon The position is tough to evaluate. Black's pawns are obviously not great, but his position contains a certain amount of dynamic potential and the unopposed g7 -bishop is an obvious trump card. 13Qd2

13 Nd5 Qd8 14 c3 was seen in M.Mueller-V.Schroeder, Baden 2002, and now 14... f5! looks like the most logical move. Karjakin prefers to delay NdS so that the subsequent arrival of a rook on d1 will threaten the d6-pawn. 13 ... Ne7

13 .. .£5 also looks quite satisfactory. 14 Rad1 f5 15 Bf3 (Diagram 41)

Diagram 41 (B)

Diagram 42 (W)

An unusual, unclear position


15 ... Be5!?

Black could also consider 15... Bxc3!? 16 bxc3 fxe4 17 Bxe4 BfS, relinquishing his strong bishop in order to mess up White's pawns, with approximate equality. Ivanchuk's move keeps things more complex. 16 exf5 Bxf5 17 Nd4 Bd718 Nd5 Nxd5 19 Bxd5 Bc6 20 Bb3 Qb6 (Diagram 42)

This position is about equal. Black has an isolated d-pawn, but this is compensated by his strong pair of bishops. 21 Kh1?!

The start of an ambitious but slightly questionable plan involving the advance of the £pawn. 21 c3 would have been more solid. 21... Rad8 22 f4 Bf6 23 Nf3 a5! (Diagram 43)

It is important to create some trouble on the queenside. 24f5?!

This show of aggression is unjustified, though Black's position already looks slightly preferable. 24 Nd4 Be4 would see the bishop settle on a good square, but perhaps White could have considered 24 a4!?, when 24 ... Bxb2?! 25 Rb1 Qa6 (the only move) 26 f5! provides dan-

92

Yugoslav Accelerated Dragon: Introduction gerous attacking chances. 24... a4 25 Bc4

The computer prefers 25 Bd5, but Karjakin was clearly pinning his hopes on the attacking power of this bishop, and 25 ... Qxb2 26 fxg6 hxg6 27 Bxc6 bxc6 28 Qf4 d5 29 Qxa4 Qc3 would leave Black with a clear positional advantage. 25 ... Qxb2 26 fxg6 hxg6 27 Ng5 Rd7! (Diagram 44)

Diagram 43 (W)

Diagram 44 (W}

Beginning counterplay

Black's defences are secure

White has some practical chances, but lvanchuk's defensive skills are more than up to the challenge. 28 Bd3 Kg7 29 Qf4?

29 Rbl! Qxa2 30 Rb4! would have been a better attempt, intending to swing the rook to f4 orh4. 29 ... Rh8 30 Nf3 Qxa2!

\lot only paying another pawn into the bank, but also turning his own a4-pawn into a potential match winner. 31 Rde1

31 Nd4?? is obviously impossible as 31...Be5 would lead to disaster on h2. 31... Qb2! (Diagram 45)

Ivanchuk has provided a model demonstration in how to refute an unsound attack. Unafraid of ghosts, he simply calculates variations as normal, puts his pieces on good squares and, in the end, succeeds in making one of the rising stars of world chess appear quite ordinary. 32 c3?!

This hardly helps, though the position was becoming pretty hopeless in any case. White has no real attack, and the a-pawn will soon decide the game. 32 ... Qxc3 33 Re3 Qc5 34 Be4 a3 35 Bxc6 bxc6 36 Re2 Ra7 37 g4 a2 38 g5 a1Q 0-1

93


Diagram 45 (W) No problem- everything is covered!

Diagram 46 (B) A positional approach

We will now examine a couple of games in which White chose to maintain his bishop's active positioning on the a2-g8 diagonal. Game 11

D Z.Pioch • Rowy 2002

K.Toma

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! Qa5!? 8 o-o o-o 9 Bb3 d6 10 h3 Bd7 11 Re1 (Diagram 46)

This useful move improves the rook in preparation for a subsequent opening of the e-file. In general White's plan will be to centralize his forces and play Nd5 at a suitable time. Then in the event of an exchange on d5, he will recapture with thee-pawn, when the rook will bE ideally placed to exert pressure against e7. 11... Rfe8!?

A useful prophylactic measure, shadowing the position of the white rook. An example of what Black must avoid is shown by the plausible sequence 1l...Nxd4 12 Bxd4 Bc6 13 Nd5 Bxd5?! 14 exd5 when Black is passive and permanently weak along thee-file. 12Qe2

12 Qd2?! is an instructive mistake which should be met by 12 ... Nxe4! 13 Nxc6 (Diagram 47), when 13... Qxc3! 14 bxc3 Nxd2 15 Nxe7+ Rxe7 16 Bxd2 leads to an ending in which only Black can hope to be better. 12 ... Qh5!? (Diagram 48)

If anyone has a chance to develop kingside attacking chances then it is likely to be White. Therefore it is quite desirable for Black to offer an exchange of queens. Instead, 12 ... Nxd4 13 Bxd4 Bc6 14 Rad1 intending Nd5looks quite pleasant for White. WARNING: 12... Nxe4?? must definitely be avoided here, due to 13 Nxc6 Bxc6 14 Nxe4 Bxe4 15 Bd2 when White's extra piece prompted Black's immediate resignation in L.Espig-I.Radulov, Raach 1969.

94

Yugoslav Accelerated Dragon: Introduction 13 Nf3 Nes 14 Nxe5 Qxe5 15 Qd3 Bc6 16 Bd4

Diagram 47 (B)

Time for a pseudo queen sacrifice

Diagram 48 (W) A queen exchange suits Black

In M.Golubev-V.Tutunnik, Internet 1999, the continuation was 16 BdS!? NxdS 17 exdS Bd7 18 Qd2 QhS with approximate equality, but perhaps the pawn grab 18 Bxa7!? could have been tried. 16...Qa5 17 NdS (Diagram 49)

Diagram 49 (B) White wants to open thee-file

Diagram 50 (W) An ideal position for White

17 ... Nd7!

The game M.Gonschior-R.Lerch, Correspondence 1997, provides a perfect example of what White is hoping for from the 11 Re1 variation: 17... Nxd5?! 18 exdS Bd719 Bxg7 Kxg7 (Diagram 50) 20 Re3! QcS 21 Rae1 BfS 22 Qd2 aS 23 g4! Bd7 24 Rxe7 a4 25 Rxd7 axb3 26 Rxe8 Rxe8 27 axb3 bS 28 c4 bxc4 29 Qc3+ Kg8 30 bxc4 and Black resigned.

95

Starting Out: The Accelerated Dragon 18 Bxg7 Kxg7 19 Qd4+ Kg8 19.. .f6 would be inadvisable here. There are times when it can be useful for Black to augment his control over the dark squares, especially following an exchange of bishops, but in this particular case the white minor pieces are a bit too close to e6. 20 c3 Rac8 White has more space and an outpost on d5, but Black is very solid. 21 Re3 QcS 22 Qd2 aS! {Diagram 51)

Diagram 51 {W)

Diagram 52 (B)

Commencing counterplay

Black can equalize now

A well-timed thrust which ensures that White will not just be able to build up on the kingside with impunity. 23 Rd1 a4 24 Bc2 NeS 25 b3 bS 26 f4 Nd7 26 ...Bxd5 27 fxe5 (27 exd5?! Nd7leaves Black with the superior minor piece) 27 ... Be6 28 exd6 exd6 looks about equal. 27 Kh2 {Diagram 52) 27 Nb4!? may have been White's last chance to claim a slight advantage. The knight would have been very well placed here. 27 ... axb3 28 axb3 Bxds 29 exds Nf6 Both sides have slight pressure on their respective flanks, and the net result is equality. If anything it may be White who has to be careful as his bishop is slightly inferior to the black knight. 30 Rf3 Kg7 31 fs Ra8 32 Qgs?! {Diagram 53) 32 b4!? looks better, fixing the b5-pawn as a potential weakness. 32 ... Ra2 32 ... h6! followed by ... g6-g5 would have been slight better for Black, due to his superior minor piece. The text looks active, but the course of the game shows that White is fine. 33 fxg6 hxg6 34 Rdf1 Qxds 34 ... Rxc2 35 Rxf6 exf6 36 Qxf6+ Kh6 should also lead to a draw by perpetual.

96

Yugoslav Accelerated Dragon: Introduction 35 Qxd5 Nxd5 36 Rxf7+ Kh6 (Diagram 54)

Now the game peters out to a draw 37 Be4 Nxc3 38 R1f4 Nxe4 39 Rxe4 Rb2 40 Rh4+ Kg5 41 Rg4+ Kh6 1/z-1/z

Obviously not 41...Kh5?? 42 Rh7 mate! Neither side can realistically avoid the perpetual.

Diagram 53 (B) Black has a chance to be better

Diagram 54 (W) leading to equality

Game12

0

F.Amonatov •

E.lnarkiev

Russian Team Championship 2007 1 e4 c5 2 Nf3 Nc6 3 Nc3 g6 4 d4 cxd4 5 Nxd4 Bg7 6 Be3 Nf6 7 Bc4 Qa5 8 o-o o-o 9 Bb3 d6 10 h3 Bd7 11 f4 (Diagram 55)

Diagram 55 (B) A more aggressive try

Diagram 56 (W) An embarrassment for White

11••• Nxd4!

97

Starting Out: The Accelerated Dragon This is definitely the soundest defence. 1l...Rac8 and 1l...Qh5 could both be met by 12 Nf3!, when Black may regret neglecting to exchange the knight, and meanwhile White can build up in the centre. 12 Bxd4 Bc6 13 Qd3! I believe this to be White's most accurate move. Alternatives are less challenging, and in some cases White can fall into early difficulties. a) 13 Qe2? looks quite natural, but can be refuted by the excellent move 13...Qb4! (Diagram 56) 14 Rad1 (no better is 14 Rfd1 Nxe4! 15 Bxg7 Kxg716 Nd5 Bxd517 Bxd5 Nf6) 14... Nxe4! 15 Bxg7 Kxg716 Nxe4 (or 16 Nd5 Bxd5 17 Bxd5 Ng3) 16 ... Qxe4 17 Qf2 Qf5, which left White with no compensation for the missing pawn in O.Andreieva-A.Muslimova, USSR Team Ch. 1967. b) 13 Qe1 should also be met by 13 ... Qb4!, when 14 Rd1 Nxe4! 15 Bxg7 Kxg716 Nd5 Qc5+ (16 ...Qxe1 ?! 17 Rfxe1 allows White to regain the pawn maintaining a slight edge) 17 Kh2 Nf6 18 Nc7 Bxg2 19 Qxe7 Bxh3 20 Kxh3 Qh5+ leads to a perpetual and has been seen in several games. In D.Janosevic-B.Parma, Zenica 1963, Black played more ambitiously with 17... e6!? 18 Qxe4 exd519 Qd4+ Qxd4 20 Rxd4 Rae8 21 f5. White eventually won this game, but it seems to me that at this point 2l...Re5! would have left Black with slightly better chances, though a draw would still be the most likely result. c) 13 Nd5 has been played in quite a few games, though it seems a little odd to combine it with an early f2-f4. Black can best highlight that fact with the excellent, unstereotypical move 13 ... Rae8! (Diagram 57), when capturing on f6 would leave the e4-pawn mortally weak. Now 14 Qd3 Nxd5 15 exd5 Bb5 forces 16 c4, blocking the b3-bishop, and there follows 16 ... Bxd4+ 17 Qxd4 Bd7 intending ... Qc5, followed by ...e7-e6 with no problems. White can also try 14 Qe1, after which 14...Qxe115 Raxe1 Nd716 Bxg7 Kxg7 17 Re3 Nc518 Rd1 Nxb3 was agreed drawn in Skovgaard-Svensson, Correspondence 1984.

Diagram 57 (W)

Diagram 58 (W)

Excellent use of the rook

X-raying the white queen

13 ... Rad8! (Diagram 58) Just as in line 'c' above, Black deploys the queen's rook in an atypical but very appropriate

98

Yugoslav Accelerated Dragon: Introduction fashion, given the demands of the position. Apart from following a generally useful principle of placing one's rooks on the same file as the enemy queen, the text has a concrete purpose in preparing a central push involving ...e7-e5 and ... d6-d5. 14 Rad1! This accurate move prevents Black's ambitions for the time being. After the less precise 14 Rae1 Black can carry out his plan with 14 ... e5 15 Be3 d5!? (also good is 15... exf4 16 Bx£4 d5) 16 fxe5 dxe417 Qc4 Qxe518 Bxa7 Rd719 Be3 Nd5 20 Nxd5 Bxd5 21 Qb5 Bc6 22 Qxe5 Bxe5 when Black was better in J.Macura-I.Privara, Decin 1972. 14... Nd7 The value of White's last can be seen in the variation 14...e5?! 15 Be3 d5? 16 exd5 when White wins material. Black would have to prefer 15 ... exf4, but then 16 Rxf4leaves White dearly better: all his pieces are active, while the open £-file combined with the b3-bishop spells trouble for Black. 15 Bxg7 Kxg7 (Diagram 59)

Diagram 59 (W) Black is very solid

Diagram 60 (B) Black is in no real danger

Both sides have played the opening very accurately. White maintains a small edge, but Black is very solid. 16 Kh2!? 16 Kh1 has been the most popular move here, but a few strong players evidently believe the king to be better off a square further forward. One common sequence has been 16 ... Nc5 17 Qd4+ (this forcing line does not work out so well, so perhaps White should prefer 17 Qe3) 17... e518 fxe5 Nxb3 19 axb3 dxe5 20 Qf2 £5! and Black has already solved all of his problems. 16 ... Qc5 16... Nc5 also looks fine. 17Qe2 17 Bd5!? Qb4 18 b3 ReS 19 R£3 Qc5 20 Bxc6 Rxc6 21 Rd2 Nf6 22 Nd5 Nxd5 23 exd5 Rc7 24 f5 left Black under some pressure in V.lvanchuk-L.Van Wely, Monte Carlo (blindfold rapid)

99

Starting Out: The Accelerated Dragon 1999. Perhaps 18... Nf6!? could be an improvement, intending to follow with ... e7-e6. 17... bs 18 Bds b419 Bxc6 bxc3

19 ... Qxc6? 20 Nd5 forks e7 and b4. 20 Bxd7 Rxd7 21 b3 (Diagram 60)

White can perhaps claim a tiny edge here due to the possibility of activating his rooks along the d-file plus the slight weakness of the c3-pawn, but Black is not in any real danger, and after some accurate play by both sides the game ends with the natural sharing of the point. 21 ... e6 22 Rf3 ReS 23 Qe3 ds 24 exds exds 25 Qes+ Kg8 26 Re1 d4 27 Qxcs Rxcs 28 Rd3 fs 29 Kg3 Kf7 30 Kf2 RedS 31 Re2 RSd6 32 Re1 Re6 33 Red1 Rde7 34 Rxd4 Re2+ 35 Kg3 Rxc2 36 R1d3 hS 37 Rc4 h4+ 38 Kxh4 Rxg2 39 Rdxc3 Rxa2 40 Rc7 Rxc7 41 Rxc7+ Kf6 Yz-Yz

After competing for as long as possible, the players now agreed to a draw as there is no chance for anyone to win.

Summary and Conclusions The Yugoslav Attack is White's most popular and highly regarded method of fighting against the traditional Dragon involving an early ... d7-d6. Therefore it is hardly surprising that a great number of players choose the 'economical' approach in attempting to employ the same system against the Accelerated version. These individuals tend to fall into one of two categories. First, there are a great many players, especially at club level, who remain largely ignorant of the subtle differences between the two types of Dragon formation. On meeting the early ... g7-g6, they naively assume that they can churn out their usual Yugoslav Attack. Very often they will make the mistake of playing an early f2-f3, which is customary in the Dragon but inaccurate against the Accelerated. The early part of the chapter shows very clearly how Black can exploit this sloppy strategy. Readers intending to play the Yugoslav Attack with either colour should take great care to familiarize themselves with this section. For White it will show you what to avoid, while for Black you can see exactly how to make the most of your opponents' inaccurate opening play. The second group are those who do understand the differences, such as the need to prevent the tempo-gaining ... d7-d5 break. The main rule of thumb is that the bishop development Bc4 and the prophylactic retreat Bb3 must come before f2-f3 or Qd2. Obviously these players will present the greater theoretical challenge, but Black still benefits from a greater variety of options compared with a normal Dragon. One significant 'bonus option' is the move 7...QaS!?, which practically forces White to castle short. This can be a useful way to throw the opponent off their game plan, although the early queen development represents a slight concession in the positions with short castling. The analysis and illustrative games contain plenty of suggestions for both sides. Overall I consider 7... Qa5!? to be a fully playable variation, which would make an excellent choice for players looking for avoid any heavy theory while guaranteeing a sound position with mutual middlegame chances.

100

Chapter Four

Yugoslav Accelerated Dragon: Main Line

---

Inaccurate Move Orders, and how to Punish them Two Interesting Sidelines after 8 Bb3! A Near-transposition to the Dragon with 8 ••• d6 A Complex and Independent V ana • t•1on: 8••• as •. I? The Critical Line: 9 f3 ds! The Main Line: 10 Bxds! Summary and Conclusions

Starting Out: The Accelerated Dragon 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! If White wants to employ the Yugoslav formation then this is the only way to do it. In the previous chapter we saw that 7 f3?! 0-0 leads to trouble after 8 Qd2 d5! or 8 Bc4 Qb6!.

7...0-0 (Diagram 1) This natural move is, unsurprisingly, the main line. As you will recall from the previous chapter, 7... Qa5 would have more or less forced White to settle for short castling with 8 0-0 0-0. Still, one can't help but feel that the immediate 7...0-0 must represent the most principled continuation for Black.

Diagram 1 (W) This chapter's starting position

Diagram 2 (W) Hunting the valuable bishop

Now White had better continue with the prophylactic retreat 8 Bb3!, shifting the bishop to a more secure location while shielding the sensitive b-pawn. Before moving on to this, we will briefly look at a few of the inferior alternatives.

Inaccurate Move Orders, and how to Punish them a) 8 f3?1 Qb6! is just another way for White to reach an inferior line which was dealt with in the last chapter via 7 f3?! 0-0 8 Bc4 Qb6!. b) 8 Qd2?! is the type of mistake often made by inexperienced players. White fails to safeguard his valuable dark-squared bishop from exchange, and this can be exploited by 8... Ng4! (Diagram 2).

If Black is allowed to exchange on e3 then he will stand much better, as his now unopposed bishop on g7 will become a monster. White cannot move the bishop away due to the attack on d4, so he had better settle for 9 Nxc6. Now 9 ... bxc610 Bd4 is not completely clear, but 9...dxc6! leaves Black slightly better.

These variations demonstrate very clearly the reasons why White must be precise in his move order. The moves he would like to play are Qdl-d2 and f2-f3. 8 Qd2 is no good because of 8 ... Ng4. If he prevents ... Ng4 with 8 f3 then he runs into 8 ... Qb6!. Thus he must proceed in strict accordance with the following sequence:

102

Yugoslav Accelerated Dragon: Main Line 1) Position the king's bishop on b3, in order to take the sting out of ... Qb6. 2) Now that ...Qb6 is not a problem, play f2-f3 to prevent ... Ng4. 3) Once ... Ng4 has been prevented, complete development with Qd2. 4) Start attacking! Before we look at the possibilities after 8 Bb3!, we should also consider the consequences of kingside castling. After 8 0-0 Black can continue with 8 ... d6, reaching a relatively harmless variation of the traditional Dragon. In fact this would be a direct improvement for Black over the previous chapter, because his queen has not had to commit herself to aS. Satisfactory as that may be, there exists a forcing possibility which strikes me as being much more in the spirit of the Accelerated Dragon: the temporary piece sacrifice 8... Nxe4! (Diagram 3).

Diagram 3 (W} An instant equalizer

Diagram 4 (W} Equality after just twelve moves

9Nxe4

After 9 Bx£7+ Rx£710 Nxe4 Black can try 10...Bxd4!?, 'sacrificing' the strong bishop in order to mobilize the central pawns: 11 Bxd4 d5! (ll...e5? 12 Be3 d513 Bg5 turned the tables in W.Watson-M.Chiburdanidze, Brussels 1987), and Black will achieve ...e7-e5 with a good game, e.g. 12 Ng5 R£5 13 Be3 e5 with the initiative. 10...Qa5 also gives Black chances for an edge. 9 ...d5 10 Nxc6 bxc6 11 Bd3 dxe4 12 Bxe4 Ba6!? (Diagram 4)

The more common 12... Qc7 should also be fine. Now after the possible continuation 13 Qxd8 Rfxd8 14 Rfb1, it is interesting for Black to try 14 ...Bd4!?; e.g. 15 Bxc6 Rac8 16 Ba4 Bxe3 17 fxe3 Rd2 18 Rd1 Rcd8 19 Rxd2 Rxd2 when Black's activity easily compensated for the pawn in I.Makka-T.Papadopoulou, Athens 2000. I won't include any illustrative games as these are only minor variations, and I think it is important to reserve as much space as possible for the more critical lines that we will encounter later in the chapter.

Theoretical? There is not too much to know here. The main thing for White is to know which move or-

103

Starting Out: The Accelerated Dragon ders to avoid, while for Black it is very useful to be familiar with some simple variations. As I have said before, you are absolutely certain, sooner or later, to encounter opponents who muddle up their move orders, so it pays to know the best ways to exploit their inaccuracies.

Statistics It should come as no surprise that the inaccurate 8 Qd2? scores just 35% for White after 8 ... Ng4. 8 0-0 fares slightly better, but 8... Nxe4! still scores a healthy 56% for Black, rising to 58% after 12... Ba6!?.

Two Interesting Sidelines after 8 Bb3! 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! o-o 8 Bb31 (Diagram 5)

Following this important prophylactic measure, White is finally ready to implement his plan: f2-f3, Qd2, followed by 0-0-0 and h2-h4 with a kingside attack.

Diagram 5 (B)

Diagram 6 (W)

Necessary prophylaxis with 8 Bb3

An interesting sideline

After 8 Bb3 Black must choose between a number of contrasting possibilities. The two main moves are 8... d6, heading for a Dragon-like position with a subtle difference, and 8... a5!? which can lead to some fascinating complications. We will come to these two lines in due course, but for the moment we will turn our attention to a couple of less popular but still very interesting sidelines.

Sideline 1: S... QaSI? 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! o-o 8 Bb3! Qa5!? (Diagram 6)

At first glance this may seem like an inferior version of the 7...Qa5 variation. White normally continues ... 9f3

104

Yugoslav Accelerated Dragon: Main Line 9 0-0 would reach the previous chapter, but the text is the natural choice. Compared with the analogous line after 7 ...Qa5 8 f3?, in which 8 ...Qb4! was strong, here the second queen move would be pointless with the white bishop already on b3. So has Black just pointlessly muddled up two different variations? 9 ... ds! ~o! The text provides the tactical justification. White can still obtain some advantage, but the whole line is likely to carry considerable surprise value against most opponents. 10 exds Nb4 11 Qd2 (Diagram 7) The usual choice and probably best, although 11 d6!? could also be considered. Then the simple recapture 11 ... exd6 is playable, but White may be able to claim a slight edge due to the isolated pawn. Another idea is 11...Nfd5!?, when S.Dueck-N.Nuesken, German Junior Ch. 2005, continued 12 Nxd5 Nxd5+ 13 Qd2 Qxd2+ 14 Bxd2 Bxd4 15 Bxd5 e6, and now 16 c3! Be5 17 Be4 Bxd6 18 0-0-0 would have left White with a slight but stable advantage thanks to his greater mobility and queenside pressure. Finally, 11...Qe5!? 12 Qe2 Qxd6 is an untested possibility that deserves attention.

Diagram 7 (B)

Diagram 8 (W)

Time for a temporary sacrifice

Regaining the piece

After 11 Qd2 Black must resort to a temporary piece sacrifice if he is to regain his pawn. ll... NbxdS!

Only this move will do. 11...Rd8? 12 0-0-0 Nbxd5? loses material after 13 Nxd5 Qxd2+ 14 Rxd2 Nxd5 15 NbS!. 12 Nxds Qxd2+ 13 Kxd2 Nxds 14 Bxds Rd8 (Diagram 8) 15 c4 15 Bb3?! makes no attempt to challenge the black position, and after 15 ... Bxd4 16 Bxd4 Rxd4+ 17 Ke3 Rd8 18 Rhd1 Bd7 Black had fully equalized and later went on to win in R.Sluka-D.Navara, Olomouc 2000. 1S ... e6 At this point I think White's most straightforward and probably best option is to return the piece with ... 16 Be4!

105

Starting Out: The Accelerated Dragon 16 NbS? exd517 Nc7 dxc4+ 18 Ke2 Bf519 Nxa8 Rxa8 was clearly better for Black in P.Simacek-J.Bauma, Czech League 2000. However, 16 Rhe1!? is a legitimate alternative, and 16...exd5 17 c5! was a little better for White in S.Movsesian-I.Smirin, Sarajevo 2002, thanks to his secure blockade of the d4-square and the possibility of a queenside pawn advance. At the same time he must always be mindful of Black's pair of bishops. 16... Bxd4

16.. .£5!? has been suggested as an improvement, after which White should probably avoid 17 Bxb7?! Bxb7 18 Kc3 K£7 with compensation. The simplest looks like 17 Bd3! when 17... Bxd418 Bxd4 Rxd4 19 Kc3 (P.Dely-B.Soos, Skopje 1967) is similar to the main game. 17 Bxd4 Rxd4+ 18 Kc3 (Diagram 9)

Diagram 9 (B)

Diagram 10 (W)

Black faces a long defence

Another interesting sideline

18... Rd7

18 ... e5 19 Rad1 does not help Black. 19 Rhd1 Rb8 20 Rxd7 Bxd7 21 Rd1

was clearly better for White in P.Svidler-C.Lutz, Frankfurt 1999. All three of White's pieces are more active than their black counterparts, and he can quite comfortably 'play for two results', pressing for the win without taking undue risks. The variation with 8 ... Qa5 has both plus and minus points. It could have considerable surprise value and it can be nice to uncork moves like 9 ... d5! and ll ... Nbxd5! against an unsuspecting opponent. On the other hand, if White is able to find the correct path then he should be able to secure a fairly comfortable advantage. If you play the Accelerated Dragon with Black and are looking for a non-theoretical antidote, with a 'worst case scenario' of a somewhat inferior endgame, then 8 ... Qa5 could be worth considering, especially at fast time controls when surprise value becomes all the more relevant.

Sideline 2: 8... e6!? 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! o-o 8 Bb3! e6!? (Diagram 10)

106

Yugoslav Accelerated Dragon: Main Line In a way the text is quite thematic for the Accelerated Dragon, in that Black is once again hoping to achieve the ... d7-d5 break. This move may not carry quite the same force here as it does in certain other variations, but it will still enable Black to mount an effective challenge for the central squares. White's most popular reply has been the logical and safe ... 90-0 The aggressive plan involving long castling is less attractive here, as the opening of the centre will make it much harder for White to execute a successful kingside attack. It makes far more sense to change tack, castle short and prepare to play against the isolated pawn which will soon be appearing on d5. The more active 9 f4!? may also appeal to some players, with the possible continuation 9... d6 (9 ... Ng4?! has been considered dubious since A.Grischuk-V.Zvjaginsev, Poikovsky 2004: 10 Qxg4 Nxd4 11 0-0-0 Nxb3+ 12 axb3 when White had the initiative and went on to win a fine game) and now: a) If you plan to play this variation with either colour it is worth knowing that 10 Ndb5 allows Black to equalize in unexpected fashion with 10 ... Nxe4! (Diagram 11). I.NatafV.Zvjaginsev, Internet 2004, continued 11 Nxe4 d5 12 Ned6! (12 Nd2?! a6 looks too dangerous for White, while 12 Nec3? d4 is obviously no good) 12... a6 13 Nxc8 (13 c3?! axb5 14 ~xb5 Qh4+! 15 g3 Qh3 16 Qe2 Na5! 17 Bc2 Nc4 is better for Black- Mikhalevski) 13 ... axb5 14 Nb6 Ra6! 15 Nxd5 exd5 16 Qxd5 Bxb2 with balanced chances.

Diagram 11 (W)

Diagram 12 {W)

Amazing, but it seems to work!

Extra pawn vs. strong bishops

b) 10 Qf3 is a natural move. Now it looks very interesting to try 10... e5!? (in B.MaciejaM.Carlsen, Drammen 2005, Black played the more solid but slightly passive 10... Nxd4 11 Bxd4 Bd7) 11 Nxc6 bxc6 12 fxe5 Ng4 13 exd6 Nxe3 14 Qxe3 Qxd6 15 Rd1 Qf6 (Diagram 12) when the strong pair of bishops gave Black just about enough compensation in M.CarlsenV.Malakhov, Sarajevo 2006. The fact that a player of Magnus Carlsen's ability can be found playing both colours in this variation speaks volumes about the richness of possibilities for both sides. 9... ds 10 exds exds

107

Starting Out: The Accelerated Dragon 10... Nxd5 led to a slightly dry position after 11 Nxd5 exd5 12 c3 Na5 13 Qd3 Nxb3 14 axb3 in P.Svidler-M.Carlsen, Morelia/Linares 2007, where White had a slight edge and was able to play for a win without taking any risks, although the young prodigy eventually succeeded in holding the draw. After the text White is also slightly better, but the presence of an extra pair of knights provides a lot more room for creativity. A possible continuation is 11 h3 Re812 Qd2 Be613 Rad1 (Diagram 13) with the usual slight advantage in P.Smirnov-B.Macieja, Saint Vincent 2005, although Black is solid and either side has scope to outplay the opponent.

Diagram 13 (B)

Diagram 14 (W)

An interesting, atypical position

Almost a main line Dragon

8 ...e6 is a very interesting move which could come as a surprise to a lot of opponents. White's most promising reaction may be the positional treatment with 9 0-0. This should be enough for a slight advantage, but some aggressively minded players may find it psychologically and emotionally difficult to play this way, having originally intended a Yugoslav Attack.

Theoretical? Neither of these sidelines is particularly theoretical. 8...Qa5 generally heads for a relatively forced transition into the endgame, from which neither side can realistically deviate. 8... e6 can lead to very sharp play after 9 f4!?, but 9 0-0 d5 leads to a fairly non-forcing middlegame with an isolated pawn, in which the result will be determined by positional understanding rather than theoretical knowledge.

Statistics 8...Qa5 9 f3 d5 scores 46% for Black, although this drops to 37% if White gets as far as 13 Be4. The second sideline, 8... e6, scores 42%. To summarize, White should enjoy a slight edge in both of these variations. However, both are perfectly playable for Black, and could be worth considering as a surprise weapon, or

108

Yugoslav Accelerated Dragon: Main Line simply as a means of forcing an opponent to think for themselves rather than churning out 25 moves of theory.

A Near-transposition to the Dragon with 8... d6 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! 0-0 8 Bb3! d6 (Diagram 14) This move may seem to represent a small moral victory for White, who has obtained his desired transposition to a normal Dragon...hasn't he? Well, yes ... sort of! In terms of ECO classification, we are most definitely in Dragon territory, and many games do proceed along the normal lines of that opening. The only small difference is that White has retreated his king's bishop to b3 slightly prematurely, whereas he would normally only do so after being provoked by, say, the arrival of a black rook on c8. This detail gives Black one extra possibility involving an early queenside pawn expansion, as we will shortly see. 9f3

White proceeds with the Yugoslav formation. Obviously 9 Qd2? would be a mistake in view of 9 ... Ng4. White occasionally settles for short castling with 9 h3 Bd7 10 0-0, but Black has nothing to worry about here and in fact enjoys a slightly improved version of the previous chapter, not having committed his queen to aS so early. 9... Bd710Qd2

The standard continuation, though it may be a slight inaccuracy. Now we will see Black's extra possibility which is unique to the positions with an early Bb3. 10... Nxd41? 11 Bxd4 b5! (Diagram 15)

Diagram 15 (W)

Diagram 16 (B)

Beginning a queenside march

A precise move order

This is the big idea; the retreat of White's bishop to b3 has lessened his control over bS, and the text is the logical way to exploit that detail. Black intends to follow with ... a7-a5 with the makings of a promising queenside assault, not to mention the threat of trapping the b3bishop. This is a fascinating line, but I have reluctantly decided not to cover it in detail here, for two

109

Starting Out: The Accelerated Dragon reasons. Although 8 ... d6 enables Black to avoid the 'absolute' main lines of the Dragon, the resulting positions can often become wildly tactical and intensely theoretical. In other words, they are probably not what a practitioner of the Accelerated Dragon is likely to want to aim for. Secondly, and just as importantly, White has a more accurate tenth move which may well negate the whole issue! It is to this possibility that we shall now turn. 10 h4! (Diagram 16)

This seems to be more accurate than 10 Qd2. White makes it his priority to open the h-file as quickly as possible. Black's soundest response is generally considered to be 10... h5, which can be found in Game 13. But for now, let's see what happens if Black employs the plan mentioned above. 10... Nxd411 Bxd4 bs 12 hs (Diagram 17)

Diagram 17 (B)

Diagram 18 (W}

White's attack is very dangerous

Complex, but it looks risky

12 ...e6

This is a common theme in these positions, preventing a knight from landing on d5. Readers who are familiar with the Yugoslav Attack will not be surprised to learn that 12... Nxh5? is practically suicidal. D.Kontic-G.Todorovic, Cetinje 1992, continued 13 Bxg7 Kxg714 Qd2 Rh8 15 g4 Nf6, and now White could have wrapped up the game in thematic style with 16 Qh6+ Kg8 17 g5 Nh518 Rxh5! gxh519 g6!, e.g. 19... e6 20 g7 Qh4+ 21 Kd2 and wins. It is much more interesting for Black to play 12... a5!? 13 hxg6 hxg6 (Diagram 18). This may represent Black's best attempt to make this line work, although I still get the impression that White ought to be better. Play may continue 14 Nd5 (not the only move; White can also consider the calm 14 a3!?, as in M.Damjanovic-J.Begovac, Sombor 1974; instead 14 Qd2 a415 Bd5 b416 Ne2 e517 Bxa8 exd418 Bd5 Nxd519 exd5 was M.GolubevS.Alonso, Dos Hermanas 2004, when 19... a3! would have given Black reasonable compensation) 14... Nxd5 15 Bxd5 Rc8 16 Qd2 Rxc2 (16 ... Bxd4?? 17 Qh6! wins) 17 Qxc2 Bxd418 0-0-0 Bg719 Kb1 Qb6 (M.Perez Candelario-G.Guseinov, European Ch., Kusadasi 2006), and now 20 Rcl looks best with a slight edge to White. 13 hxg6 hxg6 14 Qd2 as 15 a4 bxa4

110

Yugoslav Accelerated Dragon: Main Line 15... b4?! 16 Ne2leaves White with a free hand on the kingside. 16 Nxa4 Bc617 g4 Rb8 (Diagram 19)

Diagram 19 (W)

Diagram 20 (W)

The h-file is a worry

A position to avoid as Black

This position was reached in Z.Lanka-P.H.Nielsen, Moscow Olympiad 1994, and now Nielsen's suggestion of 18 Qh2! ReB 19 Be3 intending Bh6 would have given White a very dangerous attack. The above analysis should give a good idea of why I will not be advocating this line for Black, and I hope it will prove useful for readers who wish to take the white side. 10 h4! is well worth remembering, as compared with the more common 10 Qd2, the ability to open the h-file a move earlier can make a big difference to the evaluation of certain variations.

Illustrative Game Although the contestants were both strong Grandmasters, I must mention that this was ·only' a blitz game and should be regarded as such. Nevertheless, it does provide a neat demonstration of what can happen if Black stubbornly refuses to transpose to a standard Dragon position. Game13

:::J A.Cirischuk •

D.Ciurevich

World Blitz Ch., Rishon Le Zion 2006

1 e4 cs 2 Ne2 Nc6 3 Nbc3 g6 4 d4 cxd4 5 Nxd4 Bg7 6 Be3 Nf6 7 Bc4 d6 8 Bb3 0-0 9 f3 Bd7 10

h4! hs!? Theoretically speaking, this is probably Black's 'safest'- which is, of course, a relative term! 11Qd2 Nxd4?

Aiming for a sort of hybrid between the traditional Dragon and' Accelerated ...b7-b5' lines, but the combination is an unhealthy one. Objectively Black's strongest continuation is 11...Rc8, when 12 0-0-0 Ne5 reaches one of the main lines of the Dragon.

111

Starting Out: The Accelerated Dragon 12 Bxd4 bs (Diagram 20)

Compared with Diagram 15, the mutual advances of the h-pawns will have a profound impact on the evaluation of the position. Far from stifling the white attack, the h5-pawn provides a lever by which White can wrench open the black kingside. 13 g4! hxg4?!

This doesn't help, though it is hard to suggest anything satisfactory. a) 13 ... a5 doesn't solve Black's problems after 14 Bxf6 Bxf615 gxh5 Kg716 hxg6 e6 17 a4!? (Golubev). b) 13 ... e5 has been tried in a few games, but 14 Be3 still leaves Black in a mess; e.g. 14... a5 (14 ... hxg4 15 Bg5 aS 16 h5! a4 17 h6! led to another quick victory for White in A.FernandesA.Antunes, Royan 1989) 15 Bg5 (15 gxh5 Nxh5 16 Qxd6 also looks good) 15... a416 Bd5 Rc8 17 gxh5 was awful for Black in K.Klundt-B.De Greif, Siegen Olympiad 1970. 14 hS! (Diagram 21)

Diagram 21 (B)

Diagram 22 (B)

A crushing attack

Game over

Now there is no defence. 14... es

Golubev gives the variations 14 ... Nxh515 Bxg7 Kxg716 Rxh5! gxh517 Qg5+ and 14... gxh5 15 Bxf6 Bxf6 16 Qh6 h4 17 Rxh4! Bxh4+ 18 Qxh4 e5 19 Qh6, both ending in victory for White. 15 h6! Nhs

Desperation, as 15... Bh8 loses to 16 h7+! Nxh717 Qh6. 16 hxg7 Kxg7 17 Be3

Black is a piece down with a rotten position. 17 ... Rh8 18 Qxd6 (Diagram 22)

Black could have resigned here or even earlier, but plays on for a few more moves, perhaps hoping for a miracle connected with the fast time limit. 18 ... Qf6 19 o-o-o Qxf3 20 QxeS+ Qf6 21 Qxf6+ Nxf6 22 Bd4 Rxh1 23 Rxh1 g3 24 NdS 1-0

112

Yugoslav Accelerated Dragon: Main Line

Theoretical? Yes, these lines can be intensely theoretical and can often cross over into pure Dragon territory, one of the most heavily analysed openings in all of chess. I would guess that the typical Accelerated Dragon player would generally prefer to avoid this sort of thing. On the other hand, anyone interested in the white side will presumably be happy to play the Yugoslav Attack against the Dragon, and will, by definition, be the type of player who relishes a theoretical challenge. The main thing for these players is to remember 10 h4!.

Statistics 8... d6 scores 44% for Black. After the further moves 9 f3 Bd7, I think it is quite telling that 10 Qd2 scores a fairly average 53% while 10 h4! yields a significantly better 67%.

A Complex and Independent Variation: 8... as!? The following is the last major variation we will examine in this chapter, and also happens to be one of the most fascinating in the entire Accelerated Dragon. The only drawback is that many of the critical lines have been analysed through to the endgame. 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc41 0-0 8 Bb3! aS!? (Diagram 23)

Diagram 23 (W}

Diagram 24 (W}

A dynamic system

White has no hope of an advantage

Black's idea, in a nutshell, is to give himself the possibility of flicking in a timely ... a5-a4 to divert White's pieces away from controlling the centre. This point will become clearer when we reach the main lines after 9 f3 d5!.

Can't White just play 9 a4 and exploit the hole on bs? 9 a4 is probably the first move that we should look at, as it is likely to be the instinctive reaction of a lot of players encountering this variation for the first time. The first clever

113

Starting Out: The Accelerated Dragon point of Black's play is seen after the response: 9... Ng4! 10 Qxg4

Instead 10 Nxc6?! is not at all challenging and is usually employed either as a drawing attempt against a higher rated opponent, or occasionally by a strong player who doesn't know any proper theory! Play continues 10... Nxe311 Nxd8 (11 Nxe7+?! Qxe712 fxe3 Bxc3+ 13 bxc3 Qh4+ 14 g3 Qxe415 Kf2 Ra6! was slightly better for Black in S.Easterbrook SmithI.Rogers, Sydney 1996) 11...Nxd112 Nd5 (12 Nxf7? Nxc3 was just winning for Black in C.Ripolles Garcia-D.Doncevic, Palma de Mallorca 1992, while 12 Rxd1?! Bxc3+! 13 bxc3 Rxd8 also left White worse in L.Mista-M.Damjanovic, Reggio Emilia 1970) 12 ... Rxd8 (Diagram 24) 13 Rxd1 (no better is 13 Nxe7+ Kf8 14 Nxc8 Nxb2) 13 ... Kf8 14 Nb6 Ra6 15 Nxc8 Rxc8 16 Rxd7 Rb6 regains the pawn. The ending should then be about equal, though Black has gone on to score 2!f2/3 in my database. 10... Nxd4 (Diagram 25)

Diagram 25 (W)

Diagram 26 (W}

White's structure will be damaged

Driving the white pieces back

Now we see another clever point behind Black's 8th move: a subsequent ... Nxb3 will wreck White's queenside structure (compare the line 8... Ng4?! 9 Qxg4 Nxd4 10 Qd1 Nxb3 11 axb3! where White has a nice position). 11Qh4

Despite the comment about the pawn structure, White should not consider 11 Bxd4? Bxd4. True, he would avoid the doubled b-pawns, but the loss of his powerful bishop would be too high a price. The text is the usual choice and looks the most active, although 11 Qd1 has also been employed at a high level. Then 11...Nxb312 cxb3 d613 Qd2 Be614 Nd5 Bxd515 exd5 Qd716 0-0 b5 was fine for Black in D.Janosevic-M.Damjanovic, Sarajevo 1969. Another interesting possibility is 12... Ra6!?; e.g. 13 0-0 Re6 14 Qc2 b6 15 Rfe1 Bb716 f3 f5! 17 exf5 gxf5 18 Nb5 Rg619 Qf2 e5 and Black went on to win with a crushing kingside attack in O.Fossum-K.Stokke, Bergen 2002. TIP: I advise the reader to make a mental note of the rook lift ... Ra6(-e6). It is an important resource for Black, and a recurring theme in the 9 a41ine.

114

Yugoslav Accelerated Dragon: Main Line 11...d6

The immediate exchange with ll ... Nxb312 cxb3 Ra6! 13 0-0 Bf6 also looks perfectly fine. 12 NdS ReS 13 Rd1 Nxb3 14 Bb61?

14 cxb3 Ra615 Bg5 f616 Bh6 Rc6 is fairly harmless. 14...Qd7 15 cxb3 Ra6! (Diagram 26)

The invasion on c7 looked like a problem, but the text demonstrates that Black has everything under control. Z.Almasi-V.Kramnik, Monte Carlo (blindfold rapid) 2003, continued 16 Bd4 Qd8 17 0-0 Be6 18 Bxg7 Kxg719 Nf4 Qc8 20 Rd3 Rc6 21 Qg3 f6 22 Ne2 Bf7 23 Nd4 Rcl (Diagram 27) when Black's superior structure gave him a slight edge which he eventually converted to victory.

Diagram 27 (W)

Diagram 28 (B)

Black looks slightly better

A positional option

To summarize, 9 a4 Ng4! seems to give Black fully satisfactory play. Before we move on to the critical 9 f3, I should briefly call your attention to the quieter 9 o-ol? (Diagram 28). Although the 8 ... a5!? variation has acquired a reputation for highly combative and dynamic play, the present variation does enable White to spoil the party to a certain extent. The safest response is probably ... 9... Nxd4

9... d6 is best met by 10 Ndb5!, avoiding the knight exchange. A.Fedorov-V.Malakhov, Moscow 2004, continued 10... b6 11 £3 Ba6 12 Qd2 Ne5 13 a4 with a slight edge to White thanks to the b5-outpost. For those determined to reach an unbalanced middlegame, 9 ... a4!? deserves consideration. Most experts agree that White is for choice after 10 Nxa4 Nxe4 11 Nb5!, with one high level example, P.Svidler-V.Topalov, Linares 1999, continuing 1l...Ra6 12 Qe2 d613 c4 N£614 Rfd1 Bd715 Nac3 Qb8 16 h3 when White was still slightly better due to his queenside control. At the same time the position is quite irregular and difficult to play for both sides, and Topalov eventually went on to win after a tough battle. 10Bxd4d6

The position resembles the 7 ...Qa5 8 0-0 0-0 9 Bb3 variation examined in Chapter Three,

115

Starting Out: The Accelerated Dragon except that Black has placed a pawn on aS instead of his queen. In general White will strive for activity in the centre and kingside, while Black will aim for ...Bd7-c6, ... Nd7 and play on the dark squares.

11 Qe2 Bd7 (Diagram 29)

Diagram 29 (W) Is ... as-a4 a real threat?

Diagram 30 (W} That move again!

Now after the obvious 12 a4 Bc6 Black has no real worries and can follow with ... Nd7-c5. Therefore, in the top level encounter T.Radjabov-S.Tiviakov, Wijk aan Zee 2007, White preferred 12 Rad1!? when 12... a4 13 Bc4 Bc6led to an interesting middlegame. Strictly speaking I suppose White must be just a shade better, but the black position is sound and in practice both sides will have their chances. To conclude, I would say that after 9 0-0 Black's soundest move is 9... Nxd4, while players who prefer an unbalanced position may wish to consider 9 ... a4.

The Critical Line: 9 f3 dS! 9 f3

This is the most principled and challenging continuation. After all, if we assume that White is aiming for a Yugoslav Attack, then this is, by default, likely to be his next move. As an added bonus he fortifies the e4-pawn, thus depriving Black of the ... a5-a4 resource. So what is Black's idea? If you haven't encountered this variation before then you may be in for a surprise ...

9... ds! (Diagram 30) Over the course of the past three chapters we have seen that this move is one of the central themes of the entire Accelerated Dragon after 5 Nc3, and here it is yet again. Here 9 ... d5 is a principled and highly ambitious choice: Black seeks to blast the game open while the white king is still in the centre and is willing to sacrifice material to achieve his goal. Naturally, such an aggressive policy does not come without risk, but if you enjoy complications then look no further!

116

Yugoslav Accelerated Dragon: Main Line I vividly recall my feelings of bewilderment when I first became aware of this variation. To put it bluntly, 9 ... d5 appeared- to my eyes at least- as though it must surely be unsound! Ok, I thought, Black may be able to stir up some complications and set some traps, but it can't possibly be objectively correct to play like this ... can it? Well, although a final theoretical verdict has not yet been reached, the fact is that the present variation is a lot more than a bag of tricks. We have seen that the ... d7-d5 advance is a recurring theme in the non-Mar6czy lines of the Accelerated Dragon, and in the present case Black is merely applying the same idea in slightly more radical circumstances. Part of the justification comes from the fact that White's last move (9 f3) is made to look worse than redundant in the resulting positions, whereas Black's a-pawn, by contrast, can often play a key role by sacrificing itself on a4 to draw one of the white pieces away from the centre, a point which will soon become clear. So how should White respond? Obviously he should take on d5, but which of the three options should be preferred? [I will briefly mention that 10 Nxc6? turns out badly after 10... bxc6, e.g. 11 exd5 a4!? (11...Nxd5 is also strong) 12 Nxa4 Nxd513 Bxd5 cxd514 c3 d415 cxd4 e5 16 d5? Qh4+ and Black won a piece in A.Solozhentseva-R.Weemaes, Le Touquet 2005.] If you plan to play either side of this variation then you should definitely be aware that 10 NxdS? is a weak move. Black quickly seizes the initiative after 10... Nxds 11 exds (11 Bxd5?? loses a piece: 1l...Nxd4 12 Bxd4 Bxd4 13 Qxd4 e6) 11... Nb412 C4 a41 (Diagram 31) 13 Bc2 (13 Bxa4? Qa5 is no good, while 13 a3 axb314 axb4 Rxa115 Qxa1 e5! was also much better for Black in J.Horvath-Z.Eberth, Correspondence 1991) 13 ... es! 14 NbS (14 Ne2 Qh4+ is mentioned by Martin, and 14 Bd2 didn't help either after 14... Nxc2+ 15 Nxc2 Qh4+ 16 g3 Qxc4 in J.Miralles Brugues-F.Diaz Rubi, Barcelona 1995) 14... Qh4+ and Black regained the pawn while retaining a strong initiative in D.Fatyga-T.Hagberg, Correspondence 1990.

Diagram 31 (W)

Diagram 32 (W)

The a-pawn proves its worth!

Black is doing well

10 exds is a bit more challenging, but Black still looks to be in good shape after 10... Nb4 (Diagram 32). Black now threatens to recapture on d5 with an excellent game, so White is more or less

117

Starting Out: The Accelerated Dragon forced to move the d4-knight, either to e2 or b5. Regardless of which square he chooses, Black's answer will be the same: he will advance the a-pawn to a4, drawing one of the white pieces away from d5 so as to be able to recapture anyway. True, he will still be a pawn down, but his active pieces easily make up for this minor deficit. 11 Nde2

Probably best. 11 Ndb5 is also met by 11...a4! 12 Nxa4 (12 Bc4 a3!? was irritating for White in A.Sewambar-M.Berkovich, Andorra 2002, and 12...Bf5 also good, too) 12... Nfxd5 13 Bf2 (13 Bd2 Bd7 gives Black excellent play) 13... Bd714 Nbc3 Bxa4 15 Nxa4 b5 16 Nb6 Nxb6 17 Qxd8 Rfxd8 18 Bxb6 Rdc8 was better for Black in B.Ivanovic-T.Wedberg, Eksjo 1980. 11... a4 12 Nxa4

12 Bxa4?! Nfxd5leaves Black superbly placed. 12 ... Nfxds (Diagram 33)

Diagram 33 (W)

Diagram 34 (W)

Sufficient activity for a pawn

Complicated, but probably a draw!

Tournament practice has shown that Black's active pieces provide at least enough compensation for his slight material deficit. One or two imprecise moves can easily land White in trouble, but even when he plays accurately Black should always be able to regain the pawn and reach an equal endgame. 13 Bf2

A couple of other moves have been tried: a) 13 Bd4 Bxd4 14 Qxd4 (14 Nxd4?? loses to 14 ... Ne3 15 Qd2 Qxd4!) 14... Bf5 15 Nac3 (15 Qd2 is met by 15 ... Bxc2!, while 15 0-0-0?! e5! 16 Qd2 Rxa4 17 Bxa4 Qa5 gave Black a terrific attack in H.Glauser-L.Karlsson, Lugano 1982) 15 ... Nxc2+ 16 Bxc2 Nxc3 17 Qxd8 Rfxd818 Bxf5 Nxe2 19 Kxe2 gxf5 was equal in J.Kostro-G.Szilagyi, Polanica Zdroj 1969. b) On 13 Bd2 it is interesting for Black to try 13... b5!? (13 ... Be6 is also reasonable), e.g. 14 Nac3 Be6 15 a3 (15 Nxb5?! Qb6) 15 ... Nxc3 16 Bxc3 (V.Mezentsev-M.Ho, San Francisco 2003) and now 16... Bxc3+ 17 bxc3 Bxb318 cxb3 Nd3+ 19 Kfl e5 would have given Black more than enough for the pawn. 13 ... Bf5 (Diagram 34)

118

Yugoslav Accelerated Dragon: Main Line Here is one of the many slightly frustrating instances in modern chess in which a seemingly rich and complex middlegame has been more or less analysed out to a forced draw. White has tried two logical moves: a) 14 0-0, when 14... b5! 15 Nac3 Nxc3 16 Nxc3 Qxd117 Rfxd1 Bxc2 18 Bxc2 Nxc2 19 Racl Bxc3 20 Rxc2 Bf6 (Diagram 35) led to a level endgame in A.Shirov-J.Lautier, Tilburg 1997.

Diagram 35 (W)

Diagram 36 (B)

One equal endgame ...

...and another!

b) 14 a3 also seems to fizzle out to equality after 14... Nxc2+! 15 Bxc2 Qa5+! 16 b4 Nxb4! 17 axb4 Qxb4+ 18 Qd2 Qxd2+ 19 Kxd2 (Diagram 36), when 19 ...Bxa1 (19 ...Rfd8+ 20 Nd4 Bxd4 21 Bxd4 Rxd4+ 22 Kc3 Bxc2 also looks like a draw) 20 Rxa1 Rfd8+ 21 Kcl Rdc8 22 Nd4 Bxc2 23 Nxc2 Rc4 24 Rb1 Raxa4 25 Rxb7 Ra2 26 Rb2 Rxb2 27 Kxb2 e5 again led to a sharing of the point in P.Lea-J.Jordan, Internet 2004. Finally, we come to White's strongest capture and the only real way to test Black's idea:

The Main Line: 10 Bxds! 10 Bxd5!

Having looked at the above variations, the reader will have noticed the disruptive power of a well-timed ...a5-a4 thrust. Taking that into consideration, the rationale for the text becomes self-evident. 10... Nxd5 11 exd5

11 Nxd5 is an important alternative, when it is important for Black to retaliate vigorously with 11...£5!, details of which can be found in Game 14. 11... Nb4 12 Nde2 (Diagram 37) 12 ... Bf5

12... e6!? is another possibility. It is generally considered too risky for White to capture on e6, so this line often leads, after Black captures on d5, to isolated pawn positions somewhat resembling those found after 8 ... e6!?. Although this is a valid approach for Black, I have chosen to concentrate on 12 ... Bf5, as this somehow feels more in the spirit of the 8 ... a5 variation.

119

Starting Out: The Accelerated Dragon 13 Rc1 bS! (Diagram 38)

Diagram 37 (B)

Diagram 38 (W)

A critical line for 8... as!?

Sharp, but still well-known

13 ... Rc8 and 13 ... Ra6 have also been tried, but the text is the most energetic. Black's idea is revealed after 14 a3 Nxc2+! (or 14 ...Bxc2!), details of which can be found in Game 15. 140-0 ReS Now Black threatens ... Nxc2 when a subsequent ... b5-b4 will regain any lost material with interest. Therefore White's next is practically forced. 15 Nd4 (Diagram 39)

Diagram 39 (B)

Diagram 40 (W)

An important crossroads for Black

A messy middlegame

15 ... Bxd4!?

One of two major options. See Game 16 for the interesting exchange sacrifice 15 ... Rxc3!?. Other moves have been tried, but it seems to me that these two are the most promising.

120

Yugoslav Accelerated Dragon: Main Line 16 Qxd4 Nxc2 17 Rxc2 Bxc2 (Diagram 40) 18 Bh6 The usual choice, but White might prefer the rarely played 18 Nxb5!?. J.Hector-S.Dvoirys, Skelleftea 2001, continued 18... Re8 (Black could also consider 18...e5!? 19 Qxe5 Re8 20 Qd4 Rb8- compare 18 Bh6 e5) 19 Na7 Ra8 20 Nc6 Qd6 21 Bf4 Qf6 22 Qe3 (22 Be5! looks better, e.g. 22 ...Qf5 23 Re1 with strong pressure) 22 ... Ba4 23 Bg5 Qxb2 24 Nxe7+ Kg7 25 Bh6+ Kh8 26 d6 when White had very good compensation for the slight material deficit, though Black held the draw in the end. 18...es! The only good move. 19 Qxes f6 20 Qd4 20 Qe6+?! is not dangerous after 20 ... Rf7 21 Ne4 Bxe4 22 fxe4 Qd7! (Diagram 41) when it is probably White who must be careful to draw. N.De Firmian-E.Pigusov, Moscow 1989, continued 23 Rxf6 (23 Qxd7?! Rxd7 24 Rxf6 Re7 25 Rf4 Rce8 26 d6 was R.Volf-J.Fendrych, Correspondence 1994, and now 26 ... Rf7! would have given Black some winning chances) 23 ... Re8 24 Qxd7 Rxd7 25 Rf2 when a draw was agreed.

Diagram 41 (W}

Diagram 42 (W}

White has no advantage here

Black can equalize ... probably!

20... Rf7 20... Re8? led to a swift disaster in J.Hector-D.Pirrot, Metz 1988: 21 d6 Bf5 22 Ne4 Bxe4? (losing instantly, but Black was already busted) 23 Qa7! and Black was forced to resign. 21 Nxbs Rb7 (Diagram 42) White has two pawns for the exchange and a strong passed d-pawn, but analysis seems to indicate that Black can reach a drawn ending with careful play. 22 Nc3 22 a4 Qb6 23 Qxb6 Rxb6 24 Be3 Rbb8 25 Bf4 Ra8! was fine for Black in J.Hector-J.Dorfman, Clermont Ferrand 1989. 22 ...Qb6 23 Qxb6 Rxb6 24 Rf2 Rxb2! An improvement over 24 ... Bf5?! 25 Be3 which was good for White in E.Prandstetter-

121

Starting Out: The Accelerated Dragon J.Doery, Dortmund 1987. 25 Bc1 Rb3 26 axb3 Rxc3 (Diagram 43)

Diagram 43 {W}

Diagram 44 (W}

Black should hold the draw

Softening up the light squares

This should lead to a drawn opposite bishop ending, as demonstrated by Nielsen and Hansen. First they mention White's most obvious try: 27 Bb2 Rc8 28 Bxf6 Bxb3 29 d6 Rcl!+ 30 Rfl Rxfl + 31 Kxfl Be6 with an elementary draw. TIP: The drawish tendencies of opposite-coloured bishops are at their highest when other pieces are absent from the board. In this endgame it is crucial for Black to exchange rooks, otherwise White would obtain good winning chances.

With the above in mind, White's only realistic winning plan involves keeping the rooks on the board with 27 d6!?. But even here Nielsen and Hansen point out the continuation 27 ... Bf5 28 Bd2 Rxb3 29 g4 Be6 30 Bxa5 Rd3 31 Bc7 Kf7 when Black's active pieces should enable him to hold the draw with little difficulty. Summing up, 15 ... Bxd4!? seems to be holding up for Black in theoretical terms, though White may do well to investigate the possible improvement noted at move 18. The main problem for Black is that, even if everything goes to plan, it is not a lot of fun to base your opening preparation on securing a draw from a slightly inferior endgame. For that reason I would encourage the reader to consider the exchange sacrifice 15 ... Rxc3!? in Game 16. This line carries its own risks, but it also gives White more chances to go wrong.


D N.Mitkov •

E.Perelshteyn

Chicago 2006

1 e4 cs 2 Nc3 g6 3 Nf3 Bg7 4 d4 cxd4 5 Nxd4 Nc6 6 Be3 Nf6 7 Bc4 0-0 8 Bb3 as 9 f3 dS 10

122

Yugoslav Accelerated Dragon: Main Line Bxd5 Nxd5 11 Nxd5 This move is theoretically less dangerous than 11 exd5, but could nevertheless be an attractive choice for some players. There is not too much theory to learn, and most games soon end up in a near-equal endgame in which White still has some nagging chances to play for a win. 11.. .f5! (Diagram 44) Although this is unlikely to be the first move to pops into one's head, it is very logical and fully consistent with the demands of the position. White has just exchanged a valuable bishop, so Black wastes no time in softening up some newly weakened light squares. 12 Nxc6 This is the normal move, though 12 c3!? is also possible and may give White a faint edge: 12... fxe4 13 fxe4 e6 14 Nf4 (14 Nxc6? Qh4+! 15 Kd2 bxc6 16 Nb6 Rd8+ 17 Kc2 Qxe4+ 18 Kb3 a4+ 19 Ka3 Bf8+ 20 b4 Rxd1 21 Raxd1 Qxe3 22 Nxa8 Qxc3+ 23 Kxa4 Qxb4 mate was '\.Fullbrook-J.Donaldson, Seattle 1989) 14 ... Nxd4 15 cxd4 Bxd4 (Diagram 45).

Diagram 45 (W)

Diagram 46 (B)

Regaining the pawn

White should maintain a tiny edge

Black has regained the pawn and simplified the position, but he will still need to take some care over the weak dark squares (imagine a situation with a white bishop on h6 and rook on the f-file creating a permanent mating threat on f8). The game I.Berzinsh-V.Maidla, Tallinn 1996, continued 16 Bxd4 Rxf417 Be3 Rf8 (17... Qxd1+ 18 Rxd1 Rxe4looks too dangerous after 19 Rd8+ Kg7 20 Kd2) 18 Qxd8 Rxd8 when Black was very close to equality, although White eventually managed to grind out a victory. 12... bxc6 13 Nb6 Rb8 14 Qxd8 Rxd8 15 Rd1 Rxd1+ 16 Kxd1 (Diagram 46) 16...fxe4 16... Bxb2 17 Nxc8 Rxc8 18 exf5 gxf519 Ke2 is close to equal, but again White has a nagging edge and has scored surprisingly highly from here. 17 Nxc8 Rxc818 b3! li White is to achieve anything he must keep his queenside pawns intact. 18 fxe4 Bxb2 19 Ke2 Be5 is fine for Black. 18 ... exf3 19 gxf3

123

Starting Out: The Accelerated Dragon White has a tiny edge, for two reasons. First, he has three pawn islands vs. Black's four. Second, his three vs. two majority on the queenside has more chance of yielding a passed pawn than Black's majority on the kingside. Nevertheless, a draw is still the most likely outcome due to the extensive simplifications that have taken place. 19... Kf7!? {Diagram 47)

Diagram 47 (W}

Diagram 48 (W}

Black should be ok

The draw is almost secure.

19... a4 has been the most popular move here. Perelshteyn obviously does not believe it to be necessary, though it is probably just a matter of taste. In any case, his defensive technique turns out to be perfectly adequate and provides a convincing demonstration of why Black need not fear this endgame. 20 Rg1 BeS 21 Ke2 Ke6 22 Rg4 Bd6 23 Bd2 Bc7 24 Re4+ Kd7 2S Ra4 Bxh2 26 Rh4 Bc7 27 Rxh7 Rf8 28 Rh4 RfS 29 Rd4+ RdS 30 Rd3 Ke6 31 Re3+ Kd7 32 Rd3 Ke6 33 a4 Bb6 34 RxdS cxds 3S b4 axb4 36 Bxb4 Kd7 37 Kd3 es 38 c4 dxc4+ 39 Kxc4 Ke6! (Diagram 48)

In the time it takes for White to promote his a-pawn, Black will be able to score a touchdown of his own. 40 KbS Bd8 41 BcS gS 42 Be3 KfS 43 Bb6 Bxb6 44 Kxb6 g4 4S fxg4+ Kxg4 46 aS

Yz-Yz

The whole system with 8 ... a5!? and 9 ... d5! is an active and aggressive choice, so it is slightly unfortunate that White is able to steer the game towards endgames in which Black has so few winning chances. However, compared to the ending in Svidler-Lutz (see Sideline 1), White's winning chances are considerably slimmer and if Black knows what he is doing he should not be in any real trouble. I suppose the endgame is one with which both sides may be content: White can play on at little risk, while Black should be able to hold the draw. Game15

D J.Geller •

B.Savchenko

Russian Junior Championship 2006

1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 s Nc3 Bg7 6 Be3 Nf6 7 Bc4 o-o 8 Bb3 as 9 f3 dS 10

124

Yugoslav Accelerated Dragon: Main Line Bxds Nxds 11 exds This is the main line and is viewed as the sternest test of Black's play. u ... Nb4 12 Nde2 Bfs 13 Rc1 bs 14 a3 14 0-0 is the subject of the next game. The text is a principled try, but the game seems to indicate that Black can hold his own after the following forced sequence: 14... Bxc2! 15 Rxc2 Nxc2+ 16 Qxc2 b4 Black will regain the d5-pawn, leading to a material imbalance of two knights vs. rook and pawn. If White can organize his position and find secure squares for the knights, then he will stand clearly better. Unfortunately his uncastled king and uncoordinated pieces make this difficult, so Black seems to be able to maintain the balance. 17 Na4 QxdS 18 Nb6 Qe6! (Diagram 49)

Diagram 49 (W)

Diagram so (W)

Another well-studied variation

A hard pawn to stop

This is the tactical justification of Black's play. It looks as though he is walking a tightrope, but this is all well-established theory and has been seen in plenty of games. A lot of the lines have been analysed out very deeply, so a good theoretical knowledge is beneficial. 19 Kf2 19 NxaB?! is not at all dangerous; after 19... Qxe3 20 Nc7 Black has a pleasant choice: a) 20 ... RcB 21 Nd5 Rxc2 22 Nxe3 Rxb2 23 axb4 Rb1+ 24 Nd1 a4! (Diagram 50) 25 0-0 a3 26 :-Jdc3 Rxb4 27 Rd1 h5 2B RdB+ Kh7 29 RaB Rb3 30 ReB Rb2 31 RaB Rb3 is an equalizing line mentioned by Martin, which he attributes to Bagirov. Going back to Black's 25th, it may be interesting to consider 25 ... Rb3!?, attempting to restrict the white knights before advancing the a-pawn, though I suspect the final result should be a draw anyway. b) 20 ... Bd4!? was seen in F.Dannhauer-H.Kuntermann, Correspondence 19B2, which continued 21 Rfl ReB 22 f4 Ba7 23 axb4 axb4 (Diagram 51) [23 ...Qb6!? 24 bxa5 Qxa5+ 25 Qc3 Qxc7 26 Qxc7 Rxc7 is about equal]24 Qc6? (24 Qd1! looks unclear) 24 ... Qd3! 25 Rf3 Qb1 + 26 Qcl Qxcl + 27 Nxcl Rxc7 and Black eventually converted his advantage. 19 ... Rab8 20 Nf4

125

Starting Out: The Accelerated Dragon Usual, though White does have a significant alternative in 20 Qc5!?. E.Liss-E.Sutovsky, Tel Aviv 1999, continued 20 ... Rfd8 21 Re1 and now, rather than 2l...Rd6? which led to a big advantage for White after 22 Nf4, Tyomkin suggests 2l...Bxb2 22 axb4 axb4 23 Nf4 (23 Qxb4? leads to trouble after 23 ... Be5) 23 ... Qd6 24 Nfd5 Bc3 when the game is close to equality. 20... Qa2 21 Qb1!? (Diagram 52)

Diagram 51 (W)

White is under some pressure

Diagram 52 (B) Can White get the knights working?

An interesting novelty, though I doubt that it will cause anyone to abandon this line for Black. Instead 21 axb4 axb4 22 Qb1 b3 was about level in Y.Yakovich-S.Tiviakov, Port Erin 2001. It is likely that a pair of pawns will be exchanged on the queenside; the question is, does it make any difference whether the exchange takes place on b4 or a3? It seems to me that with accurate play on both sides the game is likely to end in equality either way. 21...Qxb1

2l...b3!? 22 Qxa2 bxa2 23 Ra1 Rxb6! 24 Bxb6 Rb8 25 Nd5 e6 26 Rxa2 exd5 27 Bxa5 Bxb2 looks like a draw, though White's a-pawn may give him slight chances to play for a win. Another interesting possibility is 2l...Qb3!? 22 axb4 Qxb4 23 Nfd5 Qb3 24 Nxe7+ (24 Rcl!?) 24 ... Kh8 when Black's activity and pressure against b2 make up for the slight material deficit. 22 Rxb1 bxa3 23 bxa3 Rfd8 (Diagram 53)

The remaining queenside pawns play a crucial role in these types of positions. If both of these pawns were to disappear from the board then a draw would be almost inevitable, as there are too few pawns remaining for either side to have any realistic chance of outplaying an opponent of similar strength. 24 Ne2

24 Rb5 is also likely to lead to a draw after, say, 24 ... Bd4 25 a4 Rxb6 26 Rxb6 Bxb6 27 Bxb6 Rd2+ 28 Ke3 Ra2 29 Bxa5 Rxa4. 24... Rd3 25 Bc5 e6 26 a4 Bf8 27 Bxf8 Rxf8 28 Nc4 ReS 29 Nxa5

Yz-Yz

A draw was agreed in view of 29 ... Ra3 30 Rb4 Ra8 winning the a-pawn.

126

Yugoslav Accelerated Dragon: Main line

Diagram 53 (W)

Diagram 54 (W)

Another drawish ending I'm afraid!

Enough for the exchange? Probably...

The above game and variations indicate that the direct 14 a3 does not set Black too many problems. That is why I consider 14 0-0, as featured in the next game, to be the most theoretically challenging move. Game16

0

B.lvanovic • M.Cebalo

Yugoslav Championship 1989

This is an older game, but many of the ideas remain relevant today. 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4 o-o 8 Bb3 as 9 f3 ds 10 Bxds Nxds 11 exds Nb4 12 Nde2 Bfs 13 Rc1 bs 14 o-o ReS 15 Nd4 Rxc3

Earlier I concluded that 15 ... Bxd4, while playable, did not look like too much fun for Black. The text is not without risk, but it should at least present White with more opportunities to go wrong. 16 bxc3 Nxa2 (Diagram 54)

16... Nxd5 does not give Black enough after 17 Qd2. 17 Ra1 Nxc3 18 Qd2 Qc7

18...Bxd4 is less advisable. Black is playing for an inferior, if technically drawn endgame which arises after 19 Bxd4 NxdS 20 RxaS Bxc2 21 RxbS Bd3! 22 Qxd3 Nf4 23 Qe4 Qxd4+ 24 Qxd4 Ne2+ 25 Kf2 Nxd4 (Diagram 55). Black should be able to hold the draw, and did so in D.Donchev-A.Fauland, Graz 1987, but he will certainly have to suffer for a while and I would second Martin's opinion that this is not an ending that Black should be aiming to reach. 19 d6!? Sacrificing a pawn to weaken the black pawn structure. Instead, in R.Ramesh-N.Wajih, Calcutta 2001, White tried 19 Nc6, but his idea did not have the desired effect after 19... e6

127

Starting Out: The Accelerated Dragon 20 Bh6 Bxh6 21 Qxc3 exd5 22 Qc5 ReS 23 Rfel Be6 24 Rxa5 BfS 25 Qxb5 ReS 26 Ne5 RbS 27 Qa6 Bb4 2S Raal Bxel 29 Rxe1 Qxc2, when Black had emerged from the complications with an extra pawn. 19 ... exd6 20 Nxf5 gxf5 21 Bh6 (Diagram 56)

Diagram 55 (W) A drawable, but unpleasant endgame

Diagram 56 (B) Black must take care on the kingside

In material terms Black is doing very nicely, but his weakened kingside is a cause of some concern (indeed, if Black's f5-pawn were back on g6 then he would enjoy a near-decisive advantage). 21 ...f4?

This looks wrong. Not only does it give White the option of grabbing the £-pawn immediately (see the next note), it also provides him with the opportunity to open the g-file with a well-timed g2-g3. 21...Bxh6!? 22 Qxh6 Qc5+ 23 Kh1 b4 24 Rfe1 a4 would have been stronger. True, White enjoys the luxury of being able to force a perpetual at will here, but I see no convincing way for him to aspire to anything greater. 22 Bxg7

White was successful with 22 Bxf4 in D.Neukirch-M.Bogorads, Sottrum 2005, which continued 22 ... RdS 23 Rae1 a4 24 Bh6 Qc5+ 25 Kh1 B£6 (25 ...Qd4!?) 26 Qf4 Qd4 27 Qg3+ KhS 2S Re3 RgS?? (2S ... Nd5 was essential) 29 QxgS+! 1-0. 22 ... Kxg7 23 Rfe1

Obviously 23 Qx£4?? Ne2+ was to be avoided, but now the threat has become real, hence Black's next. 23 ... d5 24 g3 b4

The computer suggests 24 ... £6!? allowing the king to hide on £7. This may well be a safer spot than hS, as subsequent annotations will reveal. 25 Kh1 Kh8 26 Qd4+ f6 27 gxf4 Qc4 (Diagram 57) 28 Qe3? 2S Qxc4looks to be winning for White after 2S ... dxc4 29 Rxa5 b3 30 cxb3 cxb3 31 Rc5 Na4 32 Rb5 b2 33 Re7 ReS 34 Kg2 etc.

128

Yugoslav Accelerated Dragon: Main line 28 ...Qc7 29 Rg1 a4 30 Qd4? This second error allows Black to escape from the jaws of defeat. 30 Rae1! would have rendered his existence perilous in the extreme; e.g. 30 ... a3?! 31 Qd4! (Diagram 58), threatening Qxf6+! while also attacking the b4-pawn, and the attempt to defend with 31...Qd6 loses immediately to 32 Qa7!.

Diagram 57 (W)

Diagram 58 (B)

White can win by swapping queens

Threatening a flashy mate

30... Ne2 31 Qxb4 Rb8 32 Qxa4 Nxg1 33 Rxg1 Qc3 Yz-Yz There is not much left for either side to play for. Black escaped by the skin of his teeth in this game. We have seen that the shattered kingside presents a constant hazard, but in the event of the improvement noted at move 21 it seems that the onus is on White to come up with a truly threatening alternative to the draw by perpetual.

Theoretical? It hardly needs stating that the 8 ... a5 variation is extremely theoretically demanding. A great number of variations have been analysed well into the endgame. The more complex lines require detailed preparation from both sides - unfortunately it is not enough to understand the general principles of the position.

Statistics 8... a5 scores 46% for Black. As for White's responses, 9 a4 scores 53%, 9 0-0 makes a surprisingly high 62%, while the main line of 9 f3 paradoxically scores less well with 52% (and 9... d5! results in an even 50/50 split).

Summary and Conclusions In this chapter we have examined a number of different possibilities that can occur after Black's most popular continuation of 1 ...0-0 against the Yugoslav formation. White must

129

Starting Out: The Accelerated Dragon begin with the prophylactic retreat 8 Bb3!, after which Black faces a major choice. Firstly, there are the two sidelines S...Qasl? and 8 ...e6!?. Both of these moves carry considerable surprise value and are theoretically sound, though there seems to be little doubt that White can obtain a slight advantage. The next possibility was 8... d6, reaching a position very close to a main line Dragon. This is probably not the best option for a 'pure' Accelerated player, but it would definitely be a logical choice for those whose repertoire also includes the traditional Dragon. For White, it is worth remembering that, after the standard moves 9 f3 Bd7, the immediate commencement of kingside operations with 10 h4! seems to be more promising than 10 Qd2. Black's final possibility is the fascinating and ingenious S... aSI?. This certainly has the power to bamboozle an unsuspecting opponent, especially in the complications resulting from 9 f3 dSI. Theoretically it seems to be holding up quite well, although there are two potential disadvantages which may deter some players. First and foremost, there is quite an intimidating body of theory which must be assimilated. Secondly, White has the irritating opportunity to reach a quiet endgame with a minimal advantage and almost no losing chances after 10 Bxd5 Nxd5 11 Nxd5 f5! 12 Nxc6 bxc6 13 Nb6 etc. Still, this is hardly a reason to abandon the line altogether. If you like the look of it in general, then my advice would be to learn and play it in most of your games, but also include in your repertoire a sideline such as 8...e6!? which can be used in 'must win' situations where you don't want to 'risk' the aforementioned endgame.

Repertoire Suggestions For readers looking for a theoretically sound repertoire against the Yugoslav Attack, I would suggest one of two main possibilities. If you enjoy dynamic play and relish the prospect of a theoretical battle then 8 ... a5 makes very good sense. If you prefer to avoid having to learn too much theory then 7...Qa5 (of the previous chapter) is a reliable choice. This leads to solid yet complex positions with chances for both sides. If the reader also happens to be versed in the Dragon proper, then 8... d6 could also be considered. Finally there are the two interesting sidelines covered in the opening part of this chapter. Personally I would probably not select either of these as the foundation of my repertoire, but I would certainly consider using them on an occasional basis.

130

Chapter Five

Lines with Nxc6

-

A Tame Sideline

-

The 'Important' Nxc6 Variation

-

A Dynamic Response

-

A Solid Alternative

-

A Better try for White

-


Starting Out: The Accelerated Dragon In this chapter we will examine variations involving an early exchange of knights on c6. Although these lines are not viewed as being theoretically challenging, the Accelerated Dragon player certainly ought to have some idea of how to meet them. NOTE: The general rule of thumb for the Dragon and Accelerated Dragon (as well as most other Sicilians!} is that an exchange of knights on c6 will usually help Black, because the recapture ... bxc6 will strengthen his control over the centre and provide him with an open b-file.

Therefore we will only consider those cases in which White may claim a particular justification for making the exchange in a specific position. We will begin with a brief examination of a deservedly rare sideline, which I include mainly for the sake of completeness. The second variation, which forms the bulk of the chapter, is much more important and needs to be studied carefully. Still, I believe that after absorbing the contents of this chapter the Accelerated Dragon player will have every reason to feel confident when facing the early knight exchange. That is not to say that it is a bad choice for White either, and the reader will also find some suggestions which I hope will prove useful should he wish to take the other side of the board.

A Tame Sideline After the standard opening moves: 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6

White occasionally opts for the immediate exchange ... 5 Nxc6 bxc6

...with the point that 6 Qd4 (Diagram 1)

Diagram 1 (B) A harmless sideline

Diagram 2 (W) Black leads in development

... creates an immediate threat to the h8-rook (compare the Hyper-Accelerated Dragon

132

lines with Nxc6 with 2... g6 3 d4 cxd4 4 Qxd4!? in Chapter One). As I said before, the text is harmless at best, and I would not advise the reader to try it as White. At the same time I must admit that I once suffered an embarrassing defeat against this variation. Furthermore, a quick database search reveals that it was once played by the great Russian GM Alexander Beliavsky, so it is obviously not completely without merit. Nevertheless, the following brief analysis should enable help the reader to avoid any pitfalls and reach a fully satisfactory position. 6... Nf6

6.. .f6!? is also possible and was my own choice in the aforementioned game. Black can complete development with moves like ... Nh6-f7, ... Bg7, ... 0-0 with a reasonable game, but the text looks more straightforward to me. 7 es Nds 8 e6

8 c4 is comfortably met by 8 ... Nc7 or 8 ... Qb6!?. 8...f6

8 ... Nf6 9 exf7+ Kxf7 is also quite satisfactory. 9 exd7+ Bxd7 (Diagram 2)

9...Qxd7!? is also playable, perhaps intending to reposition the queen somewhere along the second rank. 10 Bc4 and now 10...e5 or 10...Qb6 are both fine for Black. His queenside pawns are split, but his pieces are active and he controls the centre.

Theoretical? Not really; Black need only possess a minimal amount of knowledge to reach comfortable equality.

Statistics 5 Nxc6 bxc6 6 Qd4 scores 47% for White, a figure which drops to 42% after 6... Nf6. The following variation is of far greater significance:

The 'Important' Nxc6 Variation 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6

So far we are following the course of the previous three chapters. Most of the time White will continue with one of the traditional systems we have already examined. However, it is also possible for White to attempt to 'exploit' Black's move order with ... 7 Nxc6!? bxc6 8 eS (Diagram 3) This disruptive move provides the justification for White's decision to exchange on c6. Black must now choose between two contrasting options: the aggressive 8 ... Nd5!?, sacrificing a pawn for active play, or the more solid retreat 8... Ng8, which is nowhere near as passive as it may appear! We will examine both moves in turn, after briefly noting that 8... Nh5?? would obviously not be advisable, in view of 9 g4 when the knight is trapped.

133


Diagram 3 (B) A difficult choice for the knight

Diagram 4 (W) Black has compensation

A Dynamic Response 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Nxc6 bxc6 8 e5 Nd5!?

Black is not interested in retreating and instead prefers to fight for the initiative by sacrificing a pawn. 9 Nxd5 cxd5 10 Qxd5

White has to accept the offer, otherwise his previous moves make no sense. 10... Rb8 (Diagram 4)

At the cost of a pawn, Black obtains rapid development and active piece play. His king will be extremely safe once castled, and he will typically aim for a quick ...d7-d6, undermining the e5-pawn and opening the long diagonal for the g7-bishop. At the same time, the variation should not be viewed as bad for White; he is only slightly behind in development, he already has one extra pawn and can sometimes double his gain by taking on a7. One thing is certain: neither side should play this variation without a good theoretical knowledge, as a single inaccuracy can lead to serious difficulties. White has tried a number of different approaches here. 11 Bc4 is the main line, and will be considered in Game 17. Alternatively: a) 11 0-0-0 is possible, though the dangers of such an approach are self-evident. Black should prefer 11...Bb7! (after 11...0-0?! White can probably get away with the greedy 12 Bxa7!), when play may continue 12 Qd4 0-0 (Diagram 5). While the theoretical verdict is not completely clear, I think it fair to say that White's potential king worries make his position slightly more precarious. M.Yeo-J.Bentley, British League 2000, continued with the greedy 13 Qxd7?! (13 f4 is better, though 13... d6! 14 Qxa7 Qc8 or 14 Qa4 Qc8 gives Black ongoing compensation) 13...Qa514 Qxe7 Bxe5 (the calm 14... Rfc8!? also looks quite dangerous) 15 Qa3 Bxb2+! 16 Qxb2 Be4 gave Black a dangerous initiative. Also possible was 16... Bxg2 17 Qf6 Qa3+ (but not 17... Bxh1 ?? 18 Bd4 winning) 18 Kd2 Qb4+ 19 Kcl with an easy draw by perpetual.

134

Lines with Nxc6 Summing up, by parking his king on the queenside White is walking a tightrope. Black is certainly not winning by force, but I would not recommend any human player to play like this in an over-the-board encounter. The following approach is rather different: b) 11 Bxa7!? (Diagram 6)

Diagram 5 (W)

Diagram 6 (B)

White's king may be uncomfortable

Hoping to queen the a-pawn!

This may look like the kind of overindulgent pawn-grabbing expected from a beginner, but we will see that usually the opposite is true. In most cases White intends to return the material by allowing the capture of his b- and c-pawns, hoping that, after he catches up in development, his passed a-pawn will decide the game. To begin with this approach scored well for White, but the correct defensive methods have since been worked out. In fact the main line has been more or less analysed out to a drawn ending! 11... Rxb2 12 Bd4 12 Bc4?! e613 Qc5? Rb7! wins a piece, e.g. 14 0-0 Qc7!. 12 ... Rxc2 12 ...Rb8!? has been played by a few strong players, presumably as a means of avoiding the drawing continuation of the main line. Nevertheless, the text must surely be the objectively stronger move. 13 Bd3 e6 14 Qa8 Rc6 (Diagram 7) 15 o-o This is an important juncture, as White has an apparently tempting option in 15 Bb5?!. Unfortunately this move turns out to be fraught with danger after the spectacular reply 15... Ra6! 16 Bxa6 Qa5+ 17 Kfl 0-0 as seen in many encounters. After regaining the bishop on a6, Black will enjoy massive compensation for the exchange. Instead, 15 0-0 leads to a long and relatively forcing sequence. Well, it is not completely forced, but I doubt that either side has any worthwhile improvements.

15 ...0-0 16 Bb5 Ba6! 17 Qxd8 Rxd8 18 Bxc6 Bxf119 Kxf1 dxc6 20 Bc3 Rd3 21 Rc1 Rds (Diagram 8) This endgame is equal; White's outside passed pawn is balanced by the weakness of the e5-

135

Starting Out: The Accelerated Dragon pawn. 22 f4 g5 23 fxg5 Bxe5 24 Bxe5 Rxe5 was agreed drawn in F.Ribeiro-C.Lopez, Ciego de Avila 1996, while 22 Re1 ReS 23 Re3 Rc4 was also dead equal in V.Kupreichik-M.Petursson, Reykjavik 1980.

Diagram 7 (W}

Diagram 8 (W}

Complex, but leading to equality

A level ending is reached

It is a fact of modern chess that certain opening variations have been more or less analysed

out to a conclusion, but this should not necessarily dissuade you from including such variations in your repertoire. After all, it is unlikely that many of your opponents will know all of the correct moves; and even if they do, it is hardly disastrous to make a comfortable draw with the black pieces. Finally, if someone surprises you with this variation in a 'must win' game, you always have the option of 12... Rb8!?, although this approach obviously carries some risks as well.


D R.Aschenbrenner •

J.Krebs

Austrian League 2006

In this game we will consider White's most natural and popular choice on the eleventh move. 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 1 Nxc6 bxc6 8 es Nds 9 Nxds cxds 10 Qxds RbS 11 Bc4 {Diagram 9)

The most common move; what could be more natural than developing a piece while threatening checkmate? 11... 0-0

The main line and probably best. l l ... e6!? can also be considered, when 12 Qc5 Bb713 0-0-0! {Diagram 10) is critical, and if Black is to survive, he will need an improvement on the following: 13 ... Rc8 14 Qb4 Bxg2 (14 ... Qc7?! 15 Bg5! Bf8 16 Qd2 Bxg2 17 Rhg1 was very good

136

Lines with Nxc6 for White in V.Muratov-G.Veresov, Novgorod 1961) 15 Rhg1 Bf3 16 Bg5! f6 17 Bxe6! and now 17... Bxd118 Rxd1 Rc719 exf6 Bxf6 20 Bf4 (20 Bxf6!? also wins after 20 ...Qxf6 21 Qb8+ Qd8 22 Bxd7+ Rxd7 23 Qe5+ Kf7 24 Rxd7+ Qxd7 25 Qxh8) 20 ... Bg5 is recommended as leading to equality by Nielsen and Hansen, but 21 Bb3 seems to win immediately.

Diagram 9 (B)

Diagram 10 (B)

Spot the threat

The critical test of 11... e6!?

12 0-0-0 After 12 f4 d613 Bb3 Bb714 Qd2 dxe5 Black is at least equal, or if 12 0-0 Qc713 f4 d614 exd6 exd6 15 Bb3 ReS 16 Rae1 (L.Hajdu-A.Horvath, Hungarian Team Ch. 1994), then 16... Bxb2 looks simplest, again with at least equality. 12 ... d6! (Diagram 11)

Diagram 11 (W)

Diagram 12 (W)

Black must get the bishops working

Enough compensation?

A typical move for this variation, opening lines for both black bishops.

137

Starting Out: The Accelerated Dragon 13 Bxa7 Be6

Depending on how the reader evaluates the main line, he may also wish to consider 13... Rb4!? as in A.Frolov-A.Shabalov, USSR Ch., Moscow 1991, which continued 14 Bb3 (the threat was 14 ... Be6) 14... Qc715 exd6 (15 Be3 Bxe5 gives decent compensation) 15 ... Qxa716 d7 Rxb3 17 dxc8Q Bxb2+ 18 Kb1 Rb8 19 Qh3 Bg7+ 20 Kcl Bb2+ with perpetual. If White is looking for an improvement he could investigate 17 axb3!? Bxd718 Qxd7 Qa1+ 19 Kd2 Qxb2 20 Qc7, when the computer likes White, although strong nerves may be required for a human! 14Qd3 Bxc4

14 ... Ra8!? could also be considered. 15 Qxc4 Rc8 16 Qb3 Bxes (Diagram 12)

White has an extra pawn, but Black's central pawn majority and open queenside files provide decent practical compensation. 17 Bb6

Hoping to stifle Black's activity by blocking the b-file. The alternative was to exchange Black's powerful bishop with 17 Bd4 Rb8 18 Qe3 Bxd4 19 Rxd4 Qa5 20 a3 Rfc8 with a tough position to evaluate. White has an extra pawn, but the question is whether his a-, b- and cpawns are really more valuable than the black e- and d-pawns. Black controls the centre and can exert pressure on the queenside, whereas it is hard to imagine White being able to advance his queenside pawns any time soon. On the other hand, if White can consolidate and gradually exchange pieces, he may be able to exploit the extra pawn in the endgame. 17 ... Qd718 a4

18 Bd4 Bxd4 19 Rxd4 Rb8 20 Qd5 (if 20 Rb4 Qa7 regains the pawn with advantage) 20 ... Rb5, R.Helmreich-W.Hobusch, Correspondence 1992, was similar to the previous note. 18 ... Ra8

18 ... Rc6 has been seen in a few games and also looks pretty good for Black. 19 as Rfb8 20 Qb4 Qc6 21 f3

Now 21...d5 was a natural move, intending ... Bc7 or ... Bd6 according to circumstances, with a good position (note that 22 Qxe7? Bxb2+! 23 Kxb2 Rxa5leaves White defenceless). It seems to me that White will be unable to avoid the exchange of his bishop on b6, after which his advanced a-pawn may prove to be a weakness. Therefore, if you are thinking of playing this variation as White, 17 Bd4looks like the way to go, reaching a similar sort of position but with the pawn back on a2. In the game Black opts for something considerably more spectacular ... 21... Rxas!? (Diagram 13) 22 Bxas?

A bad decision, perhaps induced by the shock of Black's last move. Necessary was 22 Qxa5 Rxb6 23 c3, when 23 ... Ra6 (if 23 ...Qb7 24 Qa2 seems to survive) 24 Qb4 (24 Qd8+? Kg7 leaves the white king defenceless) 24 ... d5 gives Black a decent initiative for the sacrificed exchange, though no immediate breakthrough. 22 ... Rxb4 23 Bxb4

Despite White's slight material advantage of two rooks for a queen, he is clearly worse, as the quality of Black's pieces is much higher. 23 ...Qc4

138

Lines with Nxc6 Stronger was 23 ...Qb6! 24 Ba3 Qf2, forcing the immediate win of a pawn. 24 Bd2 Qa2

Once again the f2-square was beckoning: 24 ... Qd4! 25 Bc3 (or 25 c3 Qf2) 25 ...Qe3+ 26 Bd2 Qf2 with big problems for White. 25 c3 ds 26 f4?!

26 Rhel was better. 26 ...Qa1+ 27 Kc2 Qa4+ 28 Kb1 Bxf4 29 Bxf4 Qxf4 (Diagram 14)

Diagram 13 (W)

Diagram 14 (W)

A shock for White!

Black has good winning chances

Despite missing a couple of opportunities, Black has managed to win a pawn. 30 Rd4 QgS 31 Rhd1 e6 32 g3 Kg7 33 b4 Qe3 34 Kb2 Qe2+ 35 R1d2 QbS 36 Kb3 fS 37 c4 dxc4+ 38 Rxc4 Kf6 39 Res Qb6 40 Rdc2 gs 41 bs f4 42 Kb4 Qd8 43 R5c4 Kfs 44 Rb2 es 45 Ka4 Qd1+ 46 Rb3 f3 47 g4+ Ke6 48 b6 f2 49 Rc6+ Kd7 (Diagram 15)

Diagram 15 (W)

Diagram 16 (W)

The f-pawn decides the game

Not as passive as it may appear

139

Starting Out: The Accelerated Dragon White's resistance is almost at its end. A few more accurate moves and it is all over. 50 b7 Qa1+ 51 Kb4 Qd4+ 52 Ka3 Qa7+ 53 Kb2 Kxc6 0-1

Theoretically speaking, the jury is still deliberating over the soundness of 8 ... Nd5. But in practical terms it provides decent chances for Black to take over the initiative, especially against an unprepared opponent.

Theoretical? I'm afraid the answer is yes. This variation can produce a variety of long and forcing lines, some of them stretching far into the endgame.

Statistics The pawn sacrifice scores 48% for Black. White has three main options on the eleventh move, all of which score between 51-54%, with 11 Bxa7 being the narrow victor. Still, the margin is so small that I doubt it means very much.

A Solid Alternative 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Nxc6!? bxc6 8 e5 Ng8 (Diagram 16}

Although the sacrificial 8... Nd5 can be fun to play and analyse, it also carries certain drawbacks, including a disproportionately large body of theory for a rarely encountered variation. 8 ... Ng8 is less theoretically demanding and entails the sacrificing of time instead of a pawn. Black's retarded development may appear to render his position dubious, but subsequent analysis will reveal that underneath the arid surface there lies an oasis of hidden dynamism. In Diagram 16 White's first decision concerns the defence of his e5-pawn. Depending on how he chooses to guard it, he may still have to reckon with the undermining moves ... d7d6 or ... f7-f6. The black knight can reappear via h6 and f5, harassing the e3-bishop in the process. We will now take a look at some possible continuations: 9 f4

The e-pawn needs to be defended and this is the most natural way to do it. 9 Bd4 is the alternative and will be considered in due course. 9... Nh6!

9 ... d6, 9 ... d5 and 9 .. .f6 have all been tried, but there is little doubt in my mind that the text is the most promising. Black plays a useful developing move while keeping all of his other options open. 10Qd2

Planning queenside castling. White can also go short, for example with 10 Bc4 0-0 11 0-0, but after ll...d6! 12 exd6 exd613 Bd4 Nf514 Bxg7 Kxg7, intending ...Qb6(+) and ... d6-d5, Black had absolutely no problems in J.Blokhuis-S.Van Blitterswijk, Haarlem 1997. 10... 0-0 11 0-0-0 d6!

This strong pseudo-sacrifice enables Black to free his position by eliminating the e5-pawn.

140

Lines with Nxc6 12 exd6

12 h3 Nf5 13 Bf2 c5 was fine for Black in N.Faulks-J.Donaldson, Bermuda 1995. A future g2g4 could always be met by ... Nd4, when White would be ill-advised to 'sacrifice' his darksquared bishop for the sake of a mere pawn. 12 ...exd6 (Diagram 17)

Diagram 17 (W}

A balanced position

Diagram 18 (B) A tougher challenge

13 Bd4

The justification of Black's play can be seen in the variation 13 Qxd6?! Qxd6 14 Rxd6 Nf5 15 Rd3 Ba6. This is not the end of the story, as White can avoid the loss of material with 16 Bc5, but 16... Bxd317 Bxf8 Bxf118 Bxg7 Bxg2 19 Rg1 Kxg7 20 Rxg2 still left Black with better chances in the endgame in Comp Fritz 2-G.Kasparov, Cologne 1992. ADP mention a couple of other weak options for White in 13 Bc5?! Bg4 and 13 h3? Nf514 Bf2 Qa5 15 g4 Rb8! 16 gxf5? Rxb2 with a decisive attack. 13 ... Bxd4 14 Qxd4 Nf5 15 Qd2 d5

This position was reached in J.Koch-E.Birmingham, Val Thorens 1988 (and others). White has managed to exchange dark-squared bishops, but Black has a slightly superior control over the central squares, while the open b-file might also prove useful. I would evaluate the position as approximately equal, though slightly easier to play for Black. To summarize, 9 f4 does not achieve the desired effect of maintaining the e5-pawn. In just about all variations Black succeeds in achieving the undermining ... d7-d6, and after an exchange on d6, the white f-pawn is not ideally placed on f4.

A Better try for White 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Nxc6!? bxc6 8 e5 NgS 9 Bd4! (Diagram 18)

This looks more flexible and challenging than 9 f4. White improves the position of his bishop and may follow up with e5-e6. Black has tried several moves here. My personal

141

Starting Out: The Accelerated Dragon preference would probably be for 9...Qa5!?, which is the subject of Game 18. Alternatively: a) 9... Nh6 can be met by 10 e6!, when White can probably claim a slight edge. b) 9... cs is a principled move, exchanging a flank pawn for a central pawn. The drawback is that Black is opening up the game while behind in development, and 10 Bxc5 Qc711 Bd4 Bxe5 12 f4 Bxd413 Qxd4 Nf6 14 0-0-0 0-0 15 h4 d616 h5 gave White the makings of a strong attack in W.Muir-P.Arnaudov, Correspondence 1968. c) 9...f6 is the main line and has been more popular than 9 ...Qa5, though in my opinion the latter is a more efficient method for Black. Now 10 exf6?! Nxf6 just helps Black to develop, so the first player should prefer 10 f4!, which is stronger than on the previous move.

TIP: After 9 f4 Black had an effective response in 9 ... Nh61 and later ... d7-d6, whereas in the present variation White has waited for ...f7-f6 before committing himself to the advance of his own f-pawn, thus denying Black the aforementioned plan. After 10 f4 Black usually continues with the principled move 10... Qa5 (Diagram 19), putting immediate pressure on the e5-pawn.

Diagram 19 (W) A complex variation

Diagram 20 (W) lvanchuk's novelty should equalize

This complicated variation has been analysed extensively, and in these pages I will have to restrict myself to a concise summary. 11Qd2 11 exf6?! Nxf612 Bc4 d5 13 Bd3 Rb814 Qd2 0-0 was fine for Black in M.Tseitlin-M.Mikac, Ostrava 1991.

TIP: For positions in which White has advanced his pawns toes and f4, a basic 'rule of thumb' is as follows: "If, after the undermining ...f7-f6, White can find nothing better than an immediate e5xf6, then Black should be fine." 11 Qe2!? is more challenging, when 11...fxe5 (ll ... c5!? could also be considered) 12 Bxe5 (12 fxe5? was seen in R.Kennedy-J.Szabo, Varna 1958, and now 12... Rb8! would have been strong, e.g. 13 b3?? c5- ADP) 12 ... Nf613 0-0-0 (13 Qc4 Qb6 14 0-0-0 d5 was ok for Black in

142

Lines with Nxc6 M.Ulibin-G.Serper, Thilisi 1989) 13...0-0 led to a complex and double-edged middlegame in A.Kovchan-S.Vokarev, Swidnica 1999. 11.. .fxe5 12 fxe5 c5

12... Bxe5 13 0-0-0 Nf6 14 Re1 Bxd4 15 Qxd4 0-0 16 Rxe7 Qg5+ 17 Kb1 d5 18 Bd3 gave White a slight edge in V.Varavin-A.Khasin, Russian Ch. 1994. 13 Be3 Bxe5 14 Bc4 Nf6 15 o-o Ba6! (Diagram 20)

This looks like the best solution to Black's problems. Now 16 Bxa6 Qxa6 17 Bxc5 d6 18 Rae1 (18 Bd4 Bxd4+ 19 Qxd4 Qb6 is also equal- Tsesarsky) 18...0-0 19 Bd4 Bxd4+ 20 Qxd4 e5 21 Qd3 Qxd3 22 cxd3led to equality in the high level encounter G.Kasparov-V.Ivanchuk, Prague (rapid) 2002. So 9 .. .f6 seems to be holding up fairly well, and would probably be my recommendation were it not for the fact that 9 ...Qa5!? seems to achieve the same results with even fewer difficulties. The following game will shed some light on what I believe to be a very promising variation for Black.


0

5.Ghisi • D.Giadis

Correspondence 2004

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Nxc6!? bxc6 8 e5 Ng8 9 Bd4 Qa5!? (Diagram 21)

Diagram 21 (W)

Diagram 22 (W)

Possibly Black's best

White had better avoid this

10e6

This has been the most popular move, but if this is the best White can do then perhaps he should avoid 7 Nxc6 altogether. 10 f4?! looks inaccurate in view of the powerful reply 10... Rb8! (Diagram 22). [10.. .f6?! would transpose to the line 9.. .f6 10 f4 QaS discussed earlier.]

143

Starting Out: The Accelerated Dragon Now 11 e6 Nf6 12 exf7+ Kxf713 Bc4+ d5 14 Bb3 reaches the main game, while 11 Bc4 Rxb2 12 e6 f6 13 Qf3 Rb4 14 Qd3?! d5 15 Bb3 Bxe6 gave White nowhere near enough compensation in R.Woller-A.Zschalich, Correspondence 1984. In view of the problems encountered by White below, it may turn out that 10 Bc4!? is his most promising option. This type of pawn sacrifice can often prove dangerous, but in this case Black can reach a safe position if he develops calmly and returns the material at the correct moment. Play continues 10 ...Bxe511 0-0 Nf6! (1l...f6!? is an unrefuted attempt to hang on to the pawn) 12 Re1 d6 13 Bxe5 dxe5 14 Qe2 Bf5 (Diagram 23).

Diagram 23 (W)

Diagram 24 (W)

Leading to rough equality

Beware of the d-file!

In the time it takes for White to regain his pawn, Black will complete development, thus reaching an approximately equal position; e.g. 15 Bb3 (15 Qxe5 Qxe516 Rxe5 Bxc2 17 Rae1 e6led to an eventual win for White in M.Golubev-M.Fiodorov, Odessa 2005, though this was probably more to do with the 350+ gap in rating points than the superiority of his opening play) 15 ... e4 16 Qc4 0-0 17 Qxc6 has been seen in several games. Black's e-pawns may be doubled, but he has excellent central control, whereas White's queenside pawns are not exerting much of an influence. Therefore we can evaluate the position as dynamically equal, with the potential for either side to outplay the opponent. 10... Nf6 11 exf7+ 11 exd7+ Bxd7 should be about level. Black's active pieces should provide ample compensation for his damaged pawn structure, and it is White who has to be careful here; e.g. 12 Bc4 Rd8!? (Diagram 24). [This seems more incisive than the immediate 12... 0-0; however, the computer's suggestion of 12... Bg4!? 13 f3 Rd8! 14 fxg4 c5 deserves serious attention.) Now White had better play 13 Qe2, after which 13 ... Bg4 14 Qe5 Qxe5+ 15 Bxe5 0-0 16 0-0 Nd517 Bxg7 Kxg718 Ne4 was agreed drawn in T.Heinemann-C.Lutz, German Ch. 1999. Instead 13 0-0? is asking for trouble, and 13 ... Bg4 14 Be2 (14 f3 Be6! 15 Bxe6 fxe6 also leaves White unable to prevent the loss of a piece) 14... c5 was winning for Black in B.PerruchoudT.Markowski, Biel2001. 11... Kxf7

144

Lines with Nxc6 Planning ... e7-e5 followed by ... d7-d5. 12f4 12 Bc4+ d513 Bb3 Re814 Be5 Ng415 Bxg7 Kxg716 Qd2 e5 was clearly better for Black in N .Cherkasov-S.Pekacki, Czestochowa 1992. 12 ... Rb8! (Diagram 25)

Diagram 25 (W)

Diagram 26 (W)

White must fight for equality

White's opening is a failure

13 Bc4+ 13 Be2 was tried in L.Sobolevsky-G.Pfitzenreuter, Miihlhausen 2005, but if Black had gone ahead and called White's bluff with 13 ...Rxb2 I am sceptical about his compensation. Perhaps White should settle for 13 Rb1!? with some chances for equality, though this would hardly be a very inspiring way to conduct the opening. 13 ... d5 14 Bb3 Bg41 Amazingly this natural move was a novelty. 14... Rd8 had been played in a few other games, but the text just brings Black a clear advantage. 15 Qd3 BfS 16 Qd1 White again wishes to avoid the endgame after 16 Qd2 Ne4, but if he already has to resort to moving back and forth like this it is clear that his opening has been a failure. 16... Rhd8! (Diagram 26) Black could repeat with 16... Bg4, but he is unsurprisingly uninterested in a draw. Now he threatens ... c6-c5 followed by ... c4, embarrassing the b3-bishop. 17Qd2 17 0-0 would also be met by 17...c518 Be5 c4. Instead, 17 Be5looks best, but even here Black can obtain a clear advantage with 17... Ng4! (17 ... Ne4 also looks good) 18 Bxg7 Kxg719 Qd4+ e5! 20 fxe5 c5 21 Qa4 Qc7 followed by ... c5-c4, when White is in a dreadful mess. 17 ... cs 18 Bes c419 Bxb8 Rxb8 20 o-o cxb3 21 axb3 QcS+ (Diagram 27) Black's material advantage (two bishops for rook and pawn) is only slight, but his minor pieces are perfectly placed to wreak havoc upon White's fragile position. 22 Qf2?!

145

Starting Out: The Accelerated Dragon White would have done better to remove his king from its vulnerable position with 22 Khl, although this would only postpone the inevitable defeat after 22 ... d4. 22 ... d4 23 Rac1 Ng4 24 Na4 Qb4 25 Qe2 Ne3 26 Rf2 Nxc2 (Diagram 28} 27 Qf3 ReS 28 Rcf1 Ne3 29 Ra1 Qxb3 30 h3 Rc4 31 Ra3 Qb5 32 g4 Bc8 33 Nc3 dxc3 34 bxc3 Qb1+ 35 Kh2 Bb7 36 Ra1 Qxa1 37 Qxb7 Rxc3 38 g5 Bd4 0-1

Diagram 27 {W)

Diagram 28 (W)

A disaster for White

Game over

Theoretical? 8 ... Ng8 requires far less memorization than 8 ... Nd5. Just a few of the key lines after 9 f4 (ll...d6!) and 9 Bd4 (9 ... Qa5! and a few supporting variations) and away you go!

Statistics 8 ... Ng8 scores exactly 50%. After 9 f4, the strength of 9 ... Nh6! is clearly reflected in its statistical score of 61%. White does better with 9 Bd4 (54%), but the pendulum swings back to 53% in favour of Black after my recommendation of 9 ... Qa5!.

Summary and Conclusions I think it is safe to conclude that Black should not be unduly worried by the prospect of an early knight exchange. 5 Nxc6 followed by 6 Qd4 is hardly worth a second thought. However, 5 Nc3 Bg7 6 Be3 Nf6 7 Nxc6 bxc6 8 e5 needs to be taken seriously. Then both black responses are playable, but my personal preference would be for 8 ... Ng8 rather than 8 ... Nd5. Not only does the former require less memorization, I also believe it to be more reliable in theoretical terms. Despite the outwardly passive appearance of the knight retreat Black's position is highly conducive to counterattacking ideas, involving the open b-file and the timely undermining of the e5-pawn by ... d7-d6 or ... f7-f6 to open the long diagonal for the g7-bishop. If Black truly relishes the prospect of sacrificing a pawn for the initiative, and is not afraid of a heavy dose of theory, then 8... Nd5 is a perfectly valid option. But on balance, I think that 8... Ng8 is likely to represent the more attractive choice for the majority of players.

146

Chapter Six

Mar6czy Bind: Classical Variation ~

A Good Introduction to the Mar6czy

~

Basic Theory and Early Deviations

~

An Alternative Move Order for Black

~

A Natural, but Sub-standard Option

~

Alternatives after g ••• Bd7

~

The Main Line

~

White Exchanges the Dark-squared Bishops

~

White Avoids the Exchange of Dark-squared Bishops

~


Starting Out: The Accelerated Dragon 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 (Diagram

1)

Diagram 1 (B)

Diagram 2 (W)

The Classical Mar6czy

Black should avoid this sort of thing

Diagram 1 shows a common tabiya of one of the main lines of the Mar6czy Bind. As far as I am aware this system does not have a universally recognized name, but I am happy to follow the nomenclatural example of Nielsen and Hansen in referring to it as the Classical Mar6czy.

A Good Introduction to the Maroczy It is not by chance that I have decided to make this the first chapter on the Mar6czy Bind.

Later chapters will examine various deviations, refinements and move-order finesses for both sides, but before we concern ourselves with any of that our first priority should be to cover the fundamentals. The Classical Variation can be viewed as the 'vanilla' flavour of the Mar6czy Bind. In other words, neither side is trying to do anything particularly fancy in the early stages. Instead, both are content to develop their pieces and simply play the position.

Strategy for White White's main asset is his space advantage. The e4/c4 pawn duo, combined with his beautifully centralized pieces, will make it difficult for Black to achieve any notable counterplay in the immediate future. White hopes to restrain the traditional freeing advances ... d6-d5 and ... b7-b5 while gradually improving his position. Depending on how the game develops, he may attack on either flank, perhaps even switching between targets on each side of the board, as is often the privilege of the player controlling more space.

Minor piece exchanges In Experts vs. the Sicilian, the Danish GM Peter Heine Nielsen offers some highly instructive

148

Mar6czy Bind: Classical Variation advice on this topic. First he mentions that Black's life can become quite difficult if all four sets of minor pieces remain on the board (his recommendation for White involves an early Nd4-c2, with the specific intent of avoiding its exchange; see Chapter Nine for full coverage of this system). This is hardly surprising as it is easy to see that Black controls less territory than White. However, the really important point which Nielsen goes on to articulate is that Black also tends to have a hard time if all of the minor pieces are exchanged. This might sound counter-intuitive, but generally I think he is correct (though it all depends on the specific placement of the remaining pieces). Why should this be the case? The following example will illustrate the point better than my words ever could.

D B.Larsen •

J.Sorensen

Danish Championship 1989 (From Diagram 2) 21 NdS Bxds 22 RxdS!

A very instructive moment! One might easily be tempted to recapture with the e-pawn in order to create a weakness on e7. Indeed, White would still enjoy a slight advantage in that case, and we will encounter other situations in the course of this chapter when White does well to open thee-file in this way. But in the absence of minor pieces, White can maintain a larger advantage by recapturing on dS with a piece. The main reason is that having a rook on the fifth rank enables him to switch between targets on either side of the board. It also maintains the useful option of a pawn break with e4-e5. The further course of the game will demonstrate these points. 22 ... Red8 23 h4 f6

In case of 23 ... h5 Black would have to worry about a timely breakthrough with g2-g4 and h4-h5. 24 a4 e6 25 RbS Rb8 26 hS (Diagram 3)

Diagram 3 (B) Stretching the black defences

Diagram 4 (B) The h-file decides the game

26 ... Rb7

149

Starting Out: The Accelerated Dragon Staying passive. Perhaps the lesser evil would have been to close the kingside with 26 ... g5, although White still dominates the board and could pick his moment for a suitable breakthrough. 27 Qh3 Re7 28 Qg3 Rd7 29 Qg4 Re7 30 Rd3 Qb8 31 hxg6 hxg6 32 Rh5! (Diagram 4)

Larsen has skilfully switched between targets on both flanks and, with two of Black's pieces marooned on the queenside, the time has come for the final assault on the king. Black's defence has not been perfect, but in all fairness to Sorensen it will always be hard for a non-silicon based player to defend such a position. The game lasts just a few more moves: 32 ...Qg8 33 Rg3 Kf7 34 Rh6 Ke8 35 Rxg6 Qf7 36 Qf3 Kd7 37 e5 1-0

Now we have a fairly good idea of what Nielsen was talking about! It is certainly easy to imagine how the presence of minor pieces might have eased Black's defensive task. For example, a knight or light-squared bishop might have covered the sensitive b5- and d5squares, while a dark-squared bishop would have helped to defend the kingside.

Strategy for Black As we have already seen, knowing which pieces to exchange is very important. Generally speaking, Black will look to exchange knights on d4 at an early stage. Naturally White can avoid this with an early knight retreat. (In view of the earlier comments about avoiding exchanges, the reader may be forgiven for wondering why everyone on the white side doesn't play an early Nc2 or Nb3! Suffice to say that matters are not quite this simple, and I refer you to Chapter Nine for full details.) Once the knights have been exchanged, Black will usually organize his position in the following way (see Diagram 5):

Diagram 5

Diagram 6 (B)

A typical set-up for Black

Black would love a bishop exchange

The remaining knight can be redeployed to c5 via d7, following the pawn advance ... a7-a5 to safeguard its position from b2-b4 (though White can aim to prepare the advance with b2-

150

Mar6czy Bind: Classical Variation b3, a2-a3 and later b3-b4- see Game 20).

Q

Vj{

TIP: An important point here is that by playing ... Nd7 Black is actively encouraging the exchange of dark-squared bishops.

Readers who are familiar with the traditional Dragon may find this counter-intuitive, because in that opening it is usually White who goes to great pains to exchange these bishops. However, in the Maroczy Bind there are some important positional differences. The presence of a pawn on c4 renders the dark squares in White's position a little more sensitive than they would be in the Dragon, while also restricting the scope of his remaining bishop. Another important point is that White is in no position here to execute a checkmating attack along the h-file. Generally speaking, if Black is able to exchange the dark-squared bishops he will be very happy. His ultimate strategic goal is to reach a favourable 'good knight vs. bad bishop' position. The following example is typical and should be remembered: B.Annakov-P.H.Nielsen

World Junior Ch., Buenos Aires 1992 (From Diagram 6) 22 ... Bf6!

Black does not mind compromising his pawn structure in order to achieve the desired exchange of bishops. 23 Bxf6?

White is unable to resist the temptation of damaging Black's structure, although the subsequent course of the game demonstrates that the black pawns are in little danger of becoming truly weak. A lot of strong players have made the same mistake in similar positions, so it is worth making a mental note of this theme. Instead 23 Be3! would have maintained equality. 23 ... exf6 24 Qd4 Qe7 25 Rb2 Qes 26 Qxes fxes (Diagram 7)

Diagram 7 (W)

Diagram 8 (W}

A dream position for Black

The most common move order

151

Starting Out: The Accelerated Dragon Just look at the difference in quality between the minor pieces! Black is clearly better and went on to win a nice ending. We have seen some examples of what both sides must look to avoid at all costs. Obviously most games end up somewhere in between these two extremes, and we will now begin to look at some specific variations.

Basic Theory and Early Deviations We will take as our starting position that shown in Diagram 1 after White's ninth move: 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 At this point the vast majority of games continue with the natural developing move 9... Bd7 (Diagram 8}

Now the usual move is 10 Qd2, but there is a school of thought which states that 10 Nc2!? may be a strong option for White here. The idea is that, by comparison with the lines involving a very early Nc2, Black no longer has the thematic ... Nd7-cS manoeuvre at his disposal. This variation will be considered in Chapter Nine (page 281). My advice for Black is to compare the relevant section of that chapter with the following alternative move order and decide which would be best for you.

An Alternative Move Order for Black In view of the above, a strong argument can be made in favour of an immediate knight exchange with ... 9... Nxd4!? 10 Bxd4 Bd7

.. .intending to transpose to the main lines after, say, 11 Qd2 Bc6 12 f3 aS. It is worth noting that Tiviakov has always preferred this sequence when defending this variation. In many cases the desired transposition will indeed be reached, but after highlighting the positive it would be negligent of me not to point out the possible negative impact of this move order, namely the more active development of the white queen with... 11 Qd3!? (Diagram 9)

In the 9 ... Bd7 variation such a queen development would expose Her Majesty to possible attacks by ... Nc6-eS, but after the early knight exchange it could be argued that the queen exerts a slightly greater influence from d3 than d2. For example, the additional support to the e4-pawn means that White may not have to waste time with f2-f3 (see line 'a' below). Possible continuations include: a) 11... Bc6?! 12 b41 (Diagram 10).

In volume three of his splendid Opening for White according to Kramnik series, Alexander Khalifman imparts an instructive piece of wisdom concerning the importance of the d7square for Black: that his light-squared bishop and f6-knight often both require the use of this square (a theme we will see again shortly). In Diagram 10, 12... Nd7?? would cost Black his bishop after 13 bS. Therefore in Y.Pelletier-K.Lie, Gothenburg 200S, Black settled for 12 ... b6, but after 13 a3 e614 Rfd1 Qe71S a4! White had the makings of a strong queenside initiative. The game continued 1S ... Rfd8 16 aS bxaS, and now Pelletier suggests 17 bS! Bb7

152

Mar6czy Bind: Classical Variation 18 Rxa5 with a clear advantage to White.

Diagram 9 (B)

Diagram 10 (B)

An extra option for White

Accurate play from White

b) In view of the above, Black should definitely prefer 11 ...as!. Then if White is to make sense of the queen's position he should continue 12 Qe3 (sidestepping possible attacks by ... Nd7-c5 or -e5) 12 ... Bc6 13 Rfd1 Nd714 Bxg7 Kxg7 (Diagram 11).

Diagram 11 (W)

Diagram 12 (W}

A critical position for 11 Qd3!?

An unsatisfactory line for Black

This position should be compared with Game 19. It seems to me that Black should be ok, although Y.Kosashvili-Y.Afek, Tel Aviv 1992, saw White obtain an edge after 15 Rab1 Qb6 16 Qg5 Rfe8 (16 .. .f6!?) 17 Rd3 h6 18 Qh4 Qc5 19 Nd5 Bxd5 20 exd5, which he eventually converted to victory after combining threats on both sides of the board. My conclusion is that 9 ... Nxd4!? 10 Bxd4 Bd7 is a perfectly valid move order for Black. Both

153

Starting Out: The Accelerated Dragon this and the alternative 9... Bd7 permit White to steer the game along independent channels. The former allows him the option of 11 Qd3!?, while the latter permits 10 Nc2!? which can be found in Chapter Nine. From Black's perspective, it is really a personal choice, although the 9 ... Nxd4 move order has the benefit of guaranteeing a characteristic middlegame for this variation, whereas 9... Bd7 10 Nc2!? steers the game into something markedly different.

A Natural, but Sub-standard Option After 9 ... Nxd4 10 Bxd4, perhaps the most natural developing move is 10... Be6?!, which can be followed by the natural plan of development ...Qa5, ... Rfc8, ... a7-a6 and ... b7-b5- as in the Gurgenidze System found in the following chapter. It is definitely worth examining this plan a little more closely to see why Black is unable to solve his problems in this way. Play continues 11 Qd2 Qa512 Rad1! (threatening Nd5 when ... Qxd2 can be met by Nxe7+ followed by Rxd2 retaining an extra pawn) 12... Rfc8 13 b3 a614 f4! b5 15 f5 Bd7 (Diagram 12), when 16 Bxf6! Bxf6 (Khalifman points out that 16... exf6 does not help Black after 17 Nd5 Qd818 Qd4!) 17 Nd5 Qxd2 18 Nxf6+ Kg7!? (18 ...exf619 Rxd2 was very good for White in L.Ftacnik-P.Haba, Czechoslovakian Ch. 1986) 19 Rxd2 Kxf6 20 fxg6+ Kxg6 21 e5! bxc4 22 Bxc4 Bb5 23 exd6 Bxc4 24 bxc4 Rxc4 25 dxe7 ReS 26 Rd6+ Kg7 27 Re1 Rc7 28 Rxa6 Rcxe7 29 Rxe7 Rxe7left White a very good ending in O.Salmensuu-T.Verkasalo, Jarvanpaa 1998.

Alternatives after 9 ... Bd7 After the most common move 9... Bd7 most games continue with 10 Qd2, and I see no particular reason for White to deviate. Nevertheless, alternatives are occasionally tried, and the following brief analysis is mainly intended to show how to meet them as Black. a) 10 f3 is seen from time to time. Personally I see nothing at all wrong with 10... Nxd4 11 Bxd4 a5, when White will usually transpose to the main line by playing Qd2 in the next couple of moves. Instead, Black can attempt to punish White's move order with 10... Qb6!? (Diagram 13),

154

Diagram 13 (W)

Diagram 14 (W)

A double-edged position

Black's defences are solid

Mar6czy Bind: Classical Variation but the consequences are by no means clear, and pragmatists may prefer to settle for normal positions with 10... Nxd4. Perhaps it depends on the opponent; if you are facing a Grandmaster then there is a pretty good chance that he has played 10 f3 deliberately and will have something planned against 10... Qb6. If, on the other hand, you sense that your opponent probably doesn't know much theory (many people don't!) then it might be a good way to rattle him.

TIP: As the late Simon Webb said in his classic work Chess For Tigers: "Play the man, not the board!" After 10...Qb6!? play can become highly tactical, with ... Nxe4, ... Ng4 and ... Qxb2 all possible depending on how White plays. It would take several pages to analyse all the different moves to have been tried here, so I will limit myself to a brief mention of the more interesting possibilities: in the rare but interesting 11 c5!? and the more common 11 Kh1!? (considered best by Nielsen and Hansen) 11... Qxb2 12 Na4, when Black will be forced to sacrifice his queen, obtaining approximate material equality either with 12 ... Qxa1!? 13 Qxat Nxd4 14 Bxd4 Bxa4, or 12 ... Qa3 13 Bc1 Qxa4 (forced, as 13...Qb4? 14 Bd2 wins the queen in much more favourable circumstances after 14... Qa315 NbS, 14...Qxa415 Qxa4 Nxd4 16 Qd1, or 14...Qxd2 15 Qxd2 Nxd4 16 Nc3) 14 Qxa4 Nxd4. It all looks very murky, so unless you are desperate to hurl the game into complications then a transposition to the main line via 10... Nxd4 would seem to be the most prudent choice. b) White can also try an immediate kingside expansion with 10 f4!?. This is seldom seen, but is not without venom. The first point to note is that 10... Qb6? would be pointless in view of 11 Nc2! when 11...Qxb2?? 12 Na4 traps the queen (in the analogous position after 10 f3 the queen could escape to e5). A much better example for Black was shown in A.Kapetanovic-M.Petursson, New York 1987: 10... Nxd411 Bxd4 Bc612 Bf3 a5 13 Rf2 a4! 14 Qd2 Nd715 Bxg7 Kxg716 Rd1 f6! (Diagram 14), fortifying the dark squares. After 17 Bg4 (trying to exchange his bad bishop) 17... Nc518 Qe3 Qa519 Rfd2 Qa7! 20 Bf3 Ne6 21 Qxa7 Rxa7 22 g3 Ra5 Black had succeeded in reaching a promising endgame in which White had to struggle to hold the draw. c) Finally I should mention that 10 Rc1 has been played many times. I will not cover it in any detail here because it usually leads, after 10... Nxd4 11 Bxd4 Bc6, to positions resembling the main line after a subsequent Qd2. And as we will later discover, most experts now agree that the best way for White to arrange his rooks is on b1 and cl, rather than c1 and dl.

Theoretical? None of these early deviations require much theoretical knowledge, unless you desperately want to play the 10 f3 Qb6!? variation. The most important thing for both sides is to be familiar with the two principal move orders of 9 ... Bd7 and 9... Nxd4!?. As White you need to decide whether to head for the main lines (which can be found in the following section) or attempt to 'punish' Black's move order with 10 Nc2!? (page 281, Chapter Nine) or 10 Bxd4 Bd711 Qd3!? respectively.

Statistics 9... Nxd4 10 Bxd4 Bd7 11 Qd3!? scores a massive 77% for White, though this may flatter his

155

Starting Out: The Accelerated Dragon position. After 9 ... Bd7, 10 f3 scores 55% on average, but this drops to only 44% after 10 ...Qb6!?, indicating that this could work well against an unprepared opponent who has 'accidentally' played 10 f3 instead of 10 Qd2. Now that we have familiarized ourselves with these sidelines, let us move on to the most popular move:

The Main Line 10Qd2 We have already established that the correct plan for Black involves exchanging a pair of knights. 10... Nxd4 This helps to relieve the congestion in the black position, and there is little point in delaying it any longer. For instance, 10... a5?! can be met strongly by 11 Ndb5!, preventing the desired exchange while occupying the newly created hole on b5. Black occasionally tries 10... Ng4!? which gains the bishop pair after the obligatory 11 Bxg4 Bxg4, but White should maintain a slight edge after 12 f3. The light-squared bishop can often end up being a problem piece for White in the main lines, so he should not mind exchanging it. 11 Bxd4 Bc6 (Diagram 15)

Diagram 15 (W}

Diagram 16 (W}

... e7-e5 is threatened

Gaining queens ide territory

Now White must protect the e4-pawn. Actually if Black were to capture this pawn immediately he would lose a piece, as the reader can quickly verify for himself. But the real threat is 12... e5! followed by ... Nxe4 when the loss of a pawn is genuine. Usually White plays the natural12 f3 here, reaching something resembling Diagram 5 after Black's typical reaction of 12 ... a5 followed by ... Nd7-c5. White's next major decision concerns the dark-squared bishops. Most of the time he prefers to avoid the exchange and play for a slight positional advantage, but it is also quite feasible for him to swap on g7 and tar-

156

Mar6czy Bind: Classical Variation get the black kingside. It is to this plan that we will first turn our attention.

White Exchanges the Dark-squared Bishops If White intends to exchange on g7 then the most promising way of preparing it is with 12 Bd3!?, instead of the more common 12 f3 (which is usually played with the intention of meeting ... Nd7 with Be3 or Bf2).

0

NOTE: The main point of playing 12 Bd3 instead of 12 f3 is to keep the third rank clear in order to facilitate a rook lift from e3 to h3.

12 Bd3 will be examined in Game 19, but in spite of the above comment, White does occasionally play ... 12f3

... with a view to exchanging on g7. This can typically be seen after... 12... a5 13 Kh1!?

Probably the best way of preparing for kingside activity. A move like 13 Radl could be met by 13... a4!. Further explanation concerning the timing of ...a5-a4 can be found in the note to White's 13th on page 162. 13 ... Nd7

Now 13 ... a4?! is well met by 14 b4! axb3 15 axb3, when White can switch plans and play to take over the newly opened a-file. 14 Bxg7 Kxg7 15 f4 a4!? (Diagram 16)

15... f6 is an alternative, but the text is more active. 16Rf3

16 Rael would also be met by 16...Qa5. 16...Qa5

Black seems to have adequate play. A.Sokolov-A.Haik, World Team Ch., Lucerne 1985, continued 17 Rh3 Kg818 Qel f5! 19 Bf3 Rae8 20 Qh4 Rf7 21 exf5 Qxf5 when White's kingside play had come to nothing, and Black went on to take over the initiative and score a fine win. Naturally White has different ways of positioning his pieces, but in general I don't think that the plan of f2-f3 followed shortly by f3-f4 should be too dangerous. The main defensive resource for Black involves a timely advance of the f-pawn. Sometimes it will go to f6, controlling the dark squares and preparing to fortify the kingside with ... Rf7, or otherwise strike at the centre with ... f7-f5 as in the above example.


0

H.Hamdouchi • S.Tiviakov

Wijk aan Zee 2004

1 e4 C5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Nf6 6 Nc3 d6 7 Be2 Bg7 8 Be3 0-0 9 0-0 Nxd4 10

157


Bxd4 Bd7 11 Qd2 Bc612 Bd3!? (Diagram 17)

Diagram 17 (B)

Diagram 18 (W)

An aggressive system

Ignoring the knight on dS

12 ... as This is the most flexible move, gaining some useful queenside territory without yet committing any of the other pieces. 13 Rae1 Other moves have been tried here, but the basic idea of attacking the kingside is almost always the same. It is worth mentioning that an immediate 13 f4 is well met by 13 ... e5! 14 dxe5 Ng4 when Black has no problems. 13 ... a4!? 13 ... Nd7 would be the routine move, but now that White has signified a clear intention to exchange the bishops it makes sense for Black to delay or even omit the knight retreat. Nevertheless, 13... Nd7 is a perfectly sound way of handling the position, and the choice between the two plans is largely a matter of taste. After 14 Bxg7 Kxg715 Re3 White threatens Rh3 and Qh6+, but Black seems to be fine after 15... Kg8!? 16 Rh3 f6 17 Qh6 R£7 or 15 ... h6!? 16 Rh3 Rh8 intending ... a5-a4 and ... Qa5. 14Nds This looks like the most natural move. 14 f4 e5!? (14 ... Qa5 is also quite playable) 15 Be3 (15 fxe5 Ng4 followed by ... Nxe5 should be fine for Black) 15 ... Ng4 (15 ... exf4!?) 16 f5 Bf6 17 Kh1 was agreed drawn in K.Bryzgalin-B.Savchenko, Krasnodar 2003. 14... Nd7! (Diagram 18) TIP: Despite the apparent power of a white knight on dS, Black is not always obliged to exchange or evict it immediately. In many cases he does better to 'play around' the knight, postponing direct countermeasures until the necessary preparations have been carried out. 15 Bxg7 Kxg7 16 Re3 16 f4!? is suggested by Tiviakov, without further comment.

158

Mar6czy Bind: Classical Variation 16... es!

It is important for Black to gain some space in the centre and kingside. At the same time he

creates a positional threat of ... Bxd5 leading to a typically favourable knight vs. bishop position, similar to that which occurs in Game 22.

0

NOTE: It would have been a mistake to exchange on dS without the preliminary ... e7-e5, because in that case the reply e4xds would have given White pressure along the e-file.

17 Rh3

17 Nc3 Nc5 may transpose to the following note after 18 Rh3 h5. Instead 17 Bc2?! Bxd5 18 Qxd5 Nc5 19 Rd1 Ra6 is given as equal by Tiviakov, but I would definitely prefer Black on account of his superior minor piece. 11 ... hs 18 Ne3!?

Aiming for a kingside attack, but Black's defences seem to hold up. Tiviakov calls this move dubious, and strictly speaking he may be right, although White can certainly conjure up some dangerous threats. The less committal18 Nc3 led to approximately balanced chances after the natural reply 18 ... Nc5 in G.Milos-A.Hoffman, Santos 2001. 18... Ncs! (Diagram 19)

Diagram 19 (W}

Diagram 20 (B)

Can White sacrifice on hs?

Where should the king go?

Black is not 'afraid of ghosts'! 18 ... Rh8 had been played previously, but Tiviakov is confident that the sacrifices don't work. 19 f4!?

Nevertheless, White is more or less forced to sacrifice something. A feeble defensive move such as 19 Qc2? would leave him in a mess after 19... Bd7 20 Rg3 h4 21 Rf3 Ne6 as pointed out by the victor. 19 Rxh5!? is much more interesting though. Tiviakov dismisses this on the basis of 19... gxh5 20 Nf5+ Kg8 21 Qh6 Ne6 when he says that Black is winning, but White can continue the attack with 22 f4! threatening Rf3, while another possible method of attack is c4-c5 followed by Bc4. Black has an extra rook but his pieces are rather tied up for the

159

Starting Out: The Accelerated Dragon moment, and even Fritz says that White should be able to obtain at least a draw by perpetual. To be fair, Tiviakov does mention a secondary possibility in 19... Nxd3, which I believe to be stronger than 19...gxh5. After the forced continuation 20 N£5+ K£6 21 Qxd3 gxh5 Black probably should be able to defend, perhaps by returning a bit of material (he currently has an extra rook), though he may have to endure some nervous moments after 22 £4. 19... Bxe4

Now the game enters a sharp, tactical phase. The remaining notes are based on Tiviakov's excellent ChessBase annotations. 20 fs!?

The most ambitious and certainly the most interesting move, sacrificing a piece in order to thrust the £-pawn into the heart of the black kingside. After 20 Bxe4 Nxe4 21 Qd5 Tiviakov mentions a possible continuation 21...Nf6 22 Qxb7 Rb8 23 Qa7 Rxb2 24 fxe5 dxe5 25 Rh£3 Qd4 26 Qe7 Rb6 27 Rd1 Re6! with a slight edge to Black. 20 ... Nxd3 21 f6+ {Diagram 20) 21 ... Kh7?!

Surprisingly 21...Kg8! would have been more accurate, though it looks terribly counterintuitive to leave the h6-square unguarded. After 22 Nd5 (22 N£5? Bx£5 23 Rx£5 N£4! and 22 Ng4? Qb6+ 23 Kh1 Qxb2 both win easily for Black) 22 ... Nf4 23 Rx£4 (23 Nx£4 ex£4 24 Qx£4 B£5 25 g4 Qx£6 26 gx£5 leaves Black a pawn up for a bit of compensation) 23 ...exf4 24 Qx£4 Qa5 (after 24 ... Bxd5 25 Rxh5! Qb6+ 26 Kh1 Bxg2+ 27 Kxg2 Black must take a perpetual) 25 Qxe4 Qd2 {Diagram 21) Black is clearly better, despite the rather fragile appearance of his kingside.

Diagram 21 (W}

Diagram 22 (B)

Black defends successfully

A spectacular drawing combination

22 Ng4 Qb6+ 23 Kh1 Bxg2+ 24 Qxg2

24 Kxg2 Qxb2 25 Rxd3 Qxd2+ 26 Rxd2 hxg4 27 Rxd6 Rac8 should win for Black. 24... Nf4 25 Rxf4 exf4

Up to this point both sides have done a fine job of negotiating the complications, but now Hamdouchi slips up. 26 RxhS+?

160

Mar6czy Bind: Classical Variation Tiviakov mentions a incredible drawing continuation in 26 Ne5!! (Diagram 22), the main line of which runs 26 ...Qd4 (26 ...dxe5 27 Rxh5+ Kg8 28 Qg5 also forces Black to take a perpetual) 27 Rxh5+ Kg8 28 Rh8+! Kxh8 29 Qh3+ Kg8 30 Qh6 when Black has nothing better than a draw by perpetual check. Unfortunately for Hamdouchi, after his chosen move the attack soon runs out of steam. 26...gxhs 27 Qe4+ (Diagram 23)

Diagram 23 (B)

Diagram 24 (B)

Black must choose the right square!

A patient build-up

27 ... Kg8!

Avoiding a final pitfall: 27 ... Kh8?? would lead to a dramatic reversal of the result after 28 Qxf4! Qc6+ 29 Kg1 when the king escapes the checks, e.g. 29 ... Qc5+ 30 Kg2 Qc6+ 31 Kg3 etc. 28 Ne3 28 Qxf4 would not change the result after 28 ... Qc6+ 29 Kg1 Qc5+ 30 Kfl Ra5. 28 ...Qxb2 0-1 White's attack evaporates after 29 Qxf4 Qe5.

Theoretical? There is not too much theory on the lines where White intends to exchange the darksquared bishops. Instead, the result is more likely to be decided by knowledge of the different attacking and defensive/counterattacking methods available to both sides.

Statistics The kingside attacking plans have proven to be effective practical weapons, with 12 Bd3!? and 12 f3 a5 13 Kh1!? both scoring around the 65% mark. We now move on to the main lines in which White avoids the exchange of bishops.

161


White Avoids the Exchange of Dark-squared Bishops 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Bg7 8 Be3 0-0 9 0-0 Nxd4 10 Bxd4 Bd7 11 Qd2 Bc6 12 f3

This is the most natural and common method of defending the e-pawn. 12 ... a5

Preventing b2-b4 and preparing ... Nd7-c5. 13 b3! (Diagram 24)

Practice has shown the text to be the most accurate move, ensuring that ... a5-a4 can be met by b3-b4. It should be noted that the immediate 13 ... a4 was not an immediate positional threat, as after 14 b4 axb3 15 axb3 White would be well placed to contest the newly opened a-file. The problem comes when White tries to move this rook somewhere else. For example 13 Rad1 Nd714 Be3 a4! and 13 Racl a4! 14 Rfd1 Nd7 are both quite comfortable for Black. It is far less attractive for White to allow the opening of the a-file once that the rook has vacated a1, and in the meantime Black may follow up with moves like ...Qa5 and perhaps ... Nc5, with healthy play on the queenside. A final possibility is 13 Nd5, but this is well met by 13 ... Nd7!. Once again we see the theme encountered in Game 19 of playing around the strong knight. As usual Black would be happy to exchange the dark-squared bishops, and can take his time in deciding what to do about the knight on d5. All things considered, 13 b3! seems to be White's most useful and flexible move at this juncture. 13 ... Nd7!

This is the thematic knight manoeuvre that we first witnessed in Diagram 5. As I explained there Black is more than happy to exchange the dark-squared bishops. 14Be3

Usual, although 14 Bf2!? leads to some minor differences and will be discussed in Game 23. 14... Nc5 15 Rab1 Qb6 It may seem strange to station the queen opposite an enemy bishop like this, but her

placement is only temporary. Black is planning ... Rfc8, after which the queen will drop back to d8. Then the black rooks will be ready to anticipate any opening of the queenside (typically involving a2-a3 and b3-b4), while the queen may find a useful role on the kingside. Alternatively, 15... e6!? (Diagram 25) is a solid move, which also avoids the passivity of Black's play in Game 21. Here I'd like to dispel a common misconception, namely that ... e7-e6 automatically represents a concession from Black in these Maroczy structures.

0

NOTE: In Diagram 25 Black can easily defend the weak d6-pawn with ... Be5. The bishop is relatively stable on this square, because f3-f4 can always be met by ... Bxc3 followed by ... Nxe4.

Of course there could be positions in which ...e7-e6leads to major problems for Black, but ... e7-e6 is not necessarily undesirable, and the present position happens to be one in which the move has some definite pluses. Black's queen will usually settle on e7, and he definitely won't have to worry about Nd5 ideas.

162

Mar6czy Bind: Classical Variation Game 23 provides a great illustration of Black's dynamic possibilities after 1S...e6 (with the small difference that 14 Bf2 was played instead of 14 Be3). For the moment we will consider what can happen after the natural16 Rfdl.

Diagram 25 (W)

Diagram 26 (W)

A promising idea for Black

An early disaster for White!

Black should obviously defend the d-pawn with 16...BeS. Then Khalifman's 17 g3!? may be White's strongest move, when 17... Qe718 NbS Rfd8leads to a position in which Black is very solid, even if objectively White's position must be viewed as slightly favourable. Instead, in P.H.Nielsen-B.Larsen, Danish Ch. 1997, White played the disastrous 17 NbS?? (with the bishop on f2 instead of e3 this would be a good move, but here it loses by force!), allowing 17...Qh4! 18 g3 Bxg319 hxg3 Qxg3+ (Diagram 26) when Black had a raging attack. The game continued 20 Kh1 (if 20 Kfl Nxe4! 21 fxe4 fS! wins- Nielsen) 20 ... Nxe4 21 Qd3, and now the most clinical move was 21. .. Nf6! intending ... Ng4 with an unstoppable attack, as pointed out by Nielsen. Returning to the main line of 1S...Qb6, White should play ... 16 Rfc1 .. .increasing his influence over the queenside.

TIP: Practice has shown that the white rooks are optimally placed for contesting the queenside on b1 and c1, rather than the more 'natural' c1 and d1. 16... Rfc8 Black continues with his plan. The attempt to seize the initiative with 16...Qb4 17 Rc2 a4? led to nothing but grief in F.Sigalas-T.Sideris, Nikaia 200S, after 18 NdS Qxd2 19 Nxe7+ Kh8 20Rxd2. 17 Rc2! An extremely important moment. Both sides must be aware that 17 a3? is a serious mistake due to 17... Nxb3! 18 Bxb6 Nxd219 Rb2 (Diagram 27). It looks as if Black is losing a piece, but the tactical justification is revealed after 19... Nxc4!

163

Starting Out: The Accelerated Dragon 20 Bxc4 Bd7. Now White should settle for 21 Bd5 when he has reasonable chances to grovel a draw. Instead, in F.Holzke-A.Greet, Port Erin 2004, I was able to chalk up a relatively easy point after 21 Nd1? Bxb2 22 Nxb2 Be6 23 Be3 Bxc4 24 Nxc4 Ra6! 25 Nb2 Rxcl + 26 Bxcl Rc6 27 Bh6 f6 28 a4 Rc2 29 Nd1 Rc4 30 Nb2 Rb4 31 Bel b5 32 axb5 a4 33 K£2 a3 34 Nd3 Rb3 35 Bxa3 Rxd3 0-1. It is worth mentioning that my opponent was a strong IM with a rating of almost 2500. In other words, if a player of this quality can mix up his move orders with 17 a3?, then a typical dub/weekend-congress player can certainly do the same.

Diagram 27 (B)

Diagram 28 (W)

Black has a trick

Preparing for the queens ide expansion

WARNING: This whole line only works for Black because of the positioning of the white bishop on e3. If White has played 14 Bf2 instead of 14 Be3, then 17... Nxb3?? would be a blunder due to the simple retreat 18 Qd1!.

Game 23 will explain some of the other differences between the two bishop retreats. After seventeen moves, the middlegame is really just getting started. The following games demonstrate the different strategies available to both sides.


D V.Filippov •

P.H.Nielsen

Minsk 1996

This game will show why White's most natural and obvious plan of a2-a3 and b2-b3-b4 may not be as strong as was previously thought. 1 Nf3 cs 2 c4 g6 3 d4 cxd4 4 Nxd4 Bg7 5 e4 Nc6 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd7 10 Qd2 Nxd411 Bxd4 Bc612 f3 aS 13 b3 Nd714 Be3 Nc5 15 Rab1 Qb616 Rfc1 Rfc817 Rc2 Qd8

Removing Her Majesty from the firing line of the e3-bishop. 18 a3

In Beating the Sicilian 3, a classic work of its time, John Nunn stated that "In order to make

164

Mar6czy Bind: Classical Variation progress White must expel the knight from c5, and so he needs to play a3 and b4". Twelve years' worth of subsequent games have painted a somewhat different picture, and nowadays most experts prefer to delay this expansion. The main reason is that the opening of the a-file may enable Black to obtain strong counterplay, the present game being a perfect example. Games 21 and 22 will illustrate a more patient approach in which White does not hurry to open the queenside. 18 ...Qf8!? (Diagram 28} Another possible way to organize the kingside is 18 ... h5!?, intending ... Kh7 and ...Qh8 or even ... Qf8 and ... Bh6 to achieve the desired bishop exchange. TIP: The move ... h7-h5 can be quite a useful resource for Black, as it enables him to improve his kings ide pieces in the manner described above. The slight weakening of the kingside will rarely prove harmful.

However, the text also has its plus points, one of which will become clear when we reach Black's 27th move. 19 b4?! This does not have the desired effect, so in my opinion White would be better off playing more slowly with something like 19 Bfl or 19 Khl. In which case we can safely say that 18 a3 was probably not the most best choice. 19... axb4 20 axb4 Ne6! 20... Na4 is also possible, but I prefer Nielsen's treatment. The knight is very nicely centralized on e6, whereas on a4 it may hamper Black's activity on the a-file. 21 Nds Ra3! Parrying the threat of Nb6 by tactical means. Note that 2l...Bxd5? would be a bad idea here, as after 22 exd5 the black knight has to retreat to a passive square. 22 Bf2 The immediate 22 Nb6?? is refuted by 22 ...Rxe3!. 22 ... Rca8 23 Nb6 Ra1! (Diagram 29)

Diagram 29 (W) Perfect use of the open a-file

Diagram 30 (W) Commencing the counterattack

165

Starting Out: The Accelerated Dragon Just in time! If it were not for this resource then Black would suffer from a passive rook, and White would arrange to take over the a-file. Looking at Diagram 29, I think it is safe to assume that this is not the sort of thing that Filippov (or indeed Dr. Nunn) envisaged when deciding to open the queenside. 24Rd1

24 Reel is well met by 24 ... R8a2! - Nielsen. 24... Rxd1+ 25 Bxd1 Ra126 Be3

26 Rcl would allow Black to execute a favourable trade of dark-squared bishops with 26 ... Bh6 27 Be3 Bxe3+ 28 Qxe3 Qg7. 26 ... Ra3 27 Bf2 f5!? (Diagram 30)

There was a second promising line in 27 ... Bh6!? 28 Qel Nf4, with ideas of ... Nd3 and/or ... Qg7. Either way, it should be clear that White's strategy has not been a success. 28 b5 Be8

It may look as though Black is being driven back, but in reality his position possesses the

qualities of a coiled spring: compact, but full of potential energy. 29 exf5

29 Qd5? is no good due to 29 ... 8£7 30 Qxb7? Ral (Nielsen), as 31 Rd2 Bc3 32 Rd3 Nf4 wins for Black. 29 ... Qxf5 30 Nd5 KfS 31 Rc1 Qe5 32 Rb1

Fritz suggests 32 f4, but after 32 ...Qb2 33 Rc2 Qal 34 Qe2 Bd4! White will have a hard time covering the numerous holes in his overextended position. 32 ... Ra133 f4 Qe4 34 Rxa1 Bxa1 (Diagram 31)

Diagram 31 (W}

Diagram 32 (W}

White is under pressure

Time to resign

Black is doing very nicely. The white position looks rather overextended, and it is easy to imagine his pawns becoming weak in an endgame. 35 Bf3?

White finally cracks under the pressure. 35 Qc2 or 35 Qe2 would have been better.

166

Mar6czy Bind: Classical Variation 35 ... Qxc4 36 fs

Desperately trying to open a path towards the black king, but to no avail. Filippov's original intention may have been 36 Qe1 forking a1 and e6, but 36 ... Bd4! is a powerful rejoinder as 37 Qxe6?? Qb1+ wins immediately. 36...gxf5 37 h4

37 Qh6+ Ng7 keeps everything covered. 37 ... Bxb5!

Grabbing another pawn, while activating the bishop for counterattacking purposes. 38 Qh6+ Ng7 39 Bg3 Bc6 40 Nxe7

A desperate tactical shot in a hopeless position. 40... Bxf3 41 Bxd6

If 41 Nxf5 then 41...Kg8! is simplest. 41... Kf7 42 gxf3

Now it is White who will suffer a checkmating attack. 42 ... Bd4+ 43 Kg2 Qe2+ 44 Kh3 Qf1+ 45 Kh2 Bg1+ 0-1 (Diagram 32) It is mate after 46 Kh1 Bf2+ 47 Kh2 Qg1 + 48 Kh3 Qh1 + 49 Bh2 Qfl, or 46 Kg3 Qf2+ 47 K£4 (47

Kh3 Qxf3+ 48 Bg3 Qfl) 47... Qd2+ 48 Ke5 Qd4. This was a very finely played game by Nielsen, and a perfect illustration of how Black can counter the obvious a2-a3/b3-b4 plan. Game21

0

M.Pavlovic • A.Greet

Port Erin 2004

1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd7 10 Qd2 Nxd411 Bxd4 Bc6 12 f3 as 13 b3 Nd7 14 Be3 NcS 15 Rab1 Qb616 Rfc1 Rfc8 17 Rc2! Qd8 18 Bf1

In my opinion this is more promising than 18 a3. White takes a few moves to tidy up his position before embarking on a definite plan. 18... Be5

Intending ... Qf8-g7, increasing Black's influence over the long diagonal. If I had this position again I would probably prefer 18 ...h5!? intending ... Kh7 and ... Qh8. The basic idea is the same, but the more precise method would enable Black to control more space on the kingside. 19 Kh1 Qf8 20 Nd51 (Diagram 33) Good timing. Now Nb6 is an irritating threat, so I felt it appropriate to eliminate the knight. 20... Bxd5 21 exds!

This is usually the best way to recapture in such positions. Now as long as White is careful, it will be almost impossible for Black to arrange any queenside play with ...b7-b5. His only other realistic pawn break is ... e7-e6, but this could easily just create weaknesses and open things up for the white bishops. Nevertheless, although the black position may be passive, things are not all bad; he has no significant weaknesses and his position is extremely tough to break down.

167

Starting Out: The Accelerated Dragon 21 ... Qg7 22 f4 Bf6 23 g3 Ne4 24 Qg2 Nc5

I couldn't think of any particularly active plan, so resorted to shuffling around for a few moves while waiting to see what White intended. 25 Rd1!

Preventing ... Bd4. 25 ... Rc7 26 Qf3 Rb8 27 Bf2 b6 28 g4! (Diagram 34)

Diagram 33 (B) An irritating knight move

Diagram 34 (B) White finally starts attacking

Natural and strong. After organizing his pieces in a harmonious way, Pavlovic begins to expand on the kingside. 28 ... h6 29 h4 g5 30 hxg5

White could obtain a passed h-pawn with 30 fxg5 hxg5 31 h5, but then Black's kingside would be much more secure. 30... hxg5 31 Bxc5!?

A very committal decision. In positional terms the text is a major concession, but Pavlovic is investing his faith in the subsequent rapid transfer of forces to the kingside. 31 Be3 would be a promising and less radical alternative. 31... bxc5 32 Bd3 a4!

Naturally Black must break open the queenside before he is slaughtered on the opposite flank. 33 Rh2 axb3 34 axb3

On 34 Rh7 I had planned 34 ...bxa2! 35 Rxg7+ Kxg7 when Black should be ok; e.g. 36 fxg5 a1Q 37 Rxa1 Bxal 38 g6 f6 or 38 ... Rf8 with an unclear endgame. 34... Rxb3 (Diagram 35) 35 fxg5

35 Rh7 Qg6 36 Rh3 Rxd3 was also possible. There are different ways for White to force the win of an exchange, but they all offer Black some drawing chances in a worse position. 35 ... Be5?

35 ... Bxg5 was necessary. White is still better after 36 Bh7+ Qxh7 37 Qxb3 Qe4+ 38 Rg2 Kg7,

168

Mar6czy Bind: Classical Variation but Black would have reasonable chances of a draw and this would certainly be an improvement over the game. 36 Rh71 Qg6! 37 Rh31

This was the detail I had missed. 37 ... Rxd3

There is nothing to be gained from delaying this. 37...Qg7 is no good due to 38 Rfl, when the threat of 39 Rh7 probably forces 38 ... Rxd3 anyway. 38 Qxd3 Qxd3 39 Rdxd3 (Diagram 36) 39... Rb7?

Diagram 35 (W)

Diagram 36 (B)

Heading for an endgame

Chances to draw?

A serious error, based on a misjudgement of the position after the subsequent rook exchange. 39 ...Kg7 would have kept the game alive. 40 Rb3 Rb4? 41 Rxb4 cxb4

I was hoping that the white rook's prospects would be limited by the need to cover the bpawn, but the game shows how wrong I was. The main problem of the text is that the white c-pawn now becomes mobile. 42 Kg2 Kg7 43 Re3 Bc3

Forced, due to the threat of 44 c5. 44 Rxe7 b3 45 Rb7 b2 46 Kf3 Kg6 47 Kf4 BeS+ 48 Ke4 (Diagram 37) 48 ... Bg7

After 48 ... Kxg5? 49 c5 Black can resign. 49 Kd3 Kxgs so Kc2 Kxg4 51 Rxf7 Bes 52 Rf1

52 Rb7 K£5 53 Rxb2 Bxb2 54 Kxb2 would be a winning king and pawn ending. Pavlovic opts for an alternative route to victory. 52 ... Kg5 53 Rd1 Kf6 54 cS! dxcs 55 d6 Bd4 56 Re11-0

This game was not a very pleasant experience. Even though I missed some opportunities near the end, I certainly did not enjoy the phase in which I had no active plan and could do nothing but wait. In the following game I attempted a small refinement.

169


Diagram 37 (B) Not any more

Diagram 38 (W) An underrated refinement

Game22

0

D.Ciuksyte • A.Greet

Southend 2006

1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd7 10 Qd2 Nxd411 Bxd4 Bc612 f3 aS 13 b3 Nd7 14 Be3 NcS 15 Rab1 Qb616 Rfc1 Rfc817 Rc2 Qb4!? (Diagram 38)

I believe this to be a useful refinement, for reasons that will be explained. Due to the positional threat of ... a5-a4 (with the king's rook on c8, the response Nd5 can be safely met by ... Qxd2 as capturing on e7 would lose a piece), White is more or less forced to play: 18Qc11

Now White threatens to win a piece with 19 a3 followed by b3-b4, and Black's only good defence is to retreat immediately with ... 18 ... Qb6

So what on earth does Black gain from playing like this? The main idea is to try and provoke a2-a3 which, as we have seen, is probably less than useful to White in these positions. It looks strange to move the queen back and forth, but the white queen is far from ideally placed on c1 and very often she returns to d2. 19 Bf1!

A good choice, postponing a2-a3 and making a useful waiting move. In V.KotroniasA.Greet, Hastings 2004/05, White played 19 a3 Qd8 20 Qd2 Qf8 21 Bfl (21 b4?! would transpose to Game 20) 2l...h5 22 Nd5 Bxd5 23 exd5 b6, leading to a very similar position to Game 21, the principal difference being that the white a-pawn is on a3 instead of a2. This may not seem like a big deal, but it does force White to maintain a constant defence of the b3-pawn, making his task of building a kingside initiative more troublesome. After the further moves 24 g3 Rc7 25 h4 Qc8! 26 Kh2 Qf5 (Diagram 39), I had managed to conjure up

170

Mar6czy Bind: Classical Variation some activity, and following 27 Qdl Rb8 28 Rd2 Bc3 29 Bh3 Qf6 30 Bg5 Qg7 31 Re2 Bf6 32 Bf4 b5! Black was slightly better and only lost due to a subsequent blunder.

Diagram 39 (W)

Diagram 40 (W)

Black is doing well

Black would love to trade bishops

19... hS!? 20 Kh1 Kh7

The point of Black's last few moves is to facilitate manoeuvres such as ... Qd8-h8 controlling the long diagonal, or the positional threat of ... Qd8-f8 and ... Bh6 to exchange the darksquared bishops. 21 BgS! Qd8

Ciuksyte has manoeuvred well so far. She has resisted the temptation to play a2-a3, and with her last move she created an awkward attack on the e7-pawn. I couldn't see an ideal way to defend it, so decided to settle for a queen retreat. 22 NdS Bxds 23 exds Bf6! (Diagram 40)

We first learned about this theme in the introductory game fragment Annakov-Nielsen. An exchange of bishops would leave Black with a classic 'good knight vs. bad bishop' position, while the doubled pawns would not be at all weak and could even prove useful for gaining kingside territory later in the game. 24 a3?!

The wrong plan. 24 Be3! was best, aiming for a slow kingside build-up in the style of Pavlovic-Greet. 24... b6

Preventing any ideas of b3-b4 and c4-c5. Now White should retreat the bishop, but she had clearly overlooked Black's 26th. 25 Qf4? Bxgs 26 Qxgs es! (Diagram 41)

All of a sudden Black has achieved a 'dream position' for this variation: a superior minor piece and the more mobile pawn majority. 27 Qxd8

On 27 Qe3 Black can conquer further kingside territory with 27 ... h4.

171

Starting Out: The Accelerated Dragon 27 ... Rxd8 28 h4!

Ciuksyte defends well; it is important not to allow Black to gain too much space on the kingside. Though Black is still clearly better with virtually no losing chances, it is difficult to find a definitive winning plan. 28 ... Kg7 29 Kg1 Rde8 30 g3

This move was accompanied by a draw offer. I would have settled for this at the start of the game (a draw guaranteed a minimum of shared first place in a reasonably strong open tournament), but in this position I was able to play on without taking any risks.

Diagram 41 (W}

Diagram 42 (W}

An ideal scenario for Black

White must defend patiently

30... Re7 31 Reel Nd7 32 Kf2 It would be a mistake for White to try and activate her queenside majority with b3-b4, as

Black could then invade along the newly opened a-file, while his total control over the cSsquare would make it practically impossible for White to advance her pawns further. 32 ...f5 33 Be2 Kf6 (Diagram 42) 34 Ke3 Rae8 35 Rg1 Ke7 36 Rgd1 Nf6

Perhaps 36 ... Rf8!? would have been better, intending .. .f5-f4+. 37 Kf2 e4

This doesn't seem to lead anywhere, but I was having trouble finding a convincing plan. 38 f4 a4 39 bxa4 Ng4+ 40 Bxg4 hxg4 41 Rdet! It is important for White to exchange a pair of rooks. After 41 Rxb6? Rxc4 Black has chances

to develop a mating attack. 41... Rxe4 42 Rxe4 Rxe4 43 Rxb6 (Diagram 43) 43 ... Rd4

43 ... Rxa4 also leads nowhere after 44 Rb7+. 44 Rb7+ Kf6 45 Rb5 Rxa4 46 Rb6 Ke7 47 Rb7+ Ke8 48 Rb6 Ke7 49 Rb7+ Ke8

Yz-Yz

The last two games have demonstrated a solid, but slightly passive method of handling the position. The following game features the more dynamic plan with ... e7-e6.

172

Mar6czy Bind: Classical Variation

Diagram 43 (B) A drawn endgame

Diagram 44 (W) A solid yet dynamic line

Game23

D A.Ashton • S.Williams Hastings 2006/07

1 e4 cs 2 Nf3 g6 3 d4 cxd4 4 Nxd4 Bg7 5 c4 Nc6 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd7 10 Qd2 Nxd411 Bxd4 Bc6 12 f3 aS 13 b3 Nd7 14 Bf2 White sometimes prefers this extended retreat of the bishop. The basic features of the position remain the same, but there are one or two concrete differences and I advise the reader to pay close attention to the following annotations.

14... Nc5 15 Rab1 e6!? (Diagram 44) Black can also opt for the standard queenside regrouping seen in the previous three games, although the details may vary slightly. For example, in the event of the immediate 15 ... Qb6!? Black must reckon with 16 a3, when the b-pawn is immune from capture (as noted on page 163). Black's 15th may therefore appear to be a serious error, as there is no time for the thematic regrouping with ... Rfc8 because b3-b4 would win a piece. However, some specialists, such as Larsen, have happily continued with 16...Qd8!, claiming that it was worth sacrificing two whole tempi in order to provoke a2-a3. For instance, 17 Rfcl? could be strongly met by 17... a4!, while 17 b4 axb418 axb4 Na4 is nothing to worry about either. White's best is probably 17 Bdl!?, controlling b3 and maintaining a slight edge. Although 15 ... Qb6!? is quite playable, my personal preference would be for 15 ...Qc7!?. That way Black can follow up with the usual ...Rfc8 and ...Qd8, reaching positions similar to the previous two games. Note also that after a subsequent ...Qf8 or ... h7-h5/... Kh7, the possibility of ... Bh6 in certain positions may force White to put his bishop back on e3, thus losing a tempo in comparison with 14 Be3line. 16 a3?1

173

Starting Out: The Accelerated Dragon 16 Rfd1 is the main alternative, when Black must defend d6 with 16 ... Be5!. Then 17 Bd4 should be met by 17... Qe7! (Diagram 45), after which Black can consider capturing the bishop, perhaps followed by ...e6-e5 and ... Ne6 to target d4.

Diagram 45 (W)

Diagram 46 (W)

Is 18 Bxes a good move?

Blasting open the centre

TIP: One should try to avoid thinking stereotypically about doubled pawns. It would actually be a bad idea for White to exchange on e5 here. Not only are the doubled e-pawns safe from attack, they also control important squares such as d5 and d4. Instead of 17 Bd4, White may do better to consider 17 NbS!?, then after 17... Bxb5 18 cxb5 he is slightly better due to the long-term potential of his bishop pair (another example in which the theoretical weakness of doubled pawns is outweighed by other factors).

WARNING: This plan is only playable with the bishop on f2! A glance at page 163 will remind us of the game fragment Nielsen-Larsen, in which White (with his bishop on e3) confused these two lines and suffered a swift demise after ...Qh4!. In the main game White is hoping to expel the knight, but an unpleasant surprise awaits him. 16 ... a4! 17 b4 Nb3 Black should always consider this type of action very carefully. In a slightly different situation the knight might end up being stranded, but Williams has correctly judged that the dynamic benefits will outweigh any strategic risks here. 18 Qd3 Qf6 19 Na2 Hoping to exchange with Ncl, so it is important for Black not to rest on his laurels. 19 ... d5! (Diagram 46) This is an excellent move, exploiting White's temporary lack of coordination. 20 exd5 exd5 21 c5?! Hoping to keep the position relatively closed, but this does not seem to be an adequate so-

174

Mar6czy Bind: Classical Variation lution to White's problems. 21 cxd5looks better, when after 21...Rad8 Ward points out the excellent defence 22 Nc3! Qxc3 23 Qxc3 Bxc3 24 dxc6 bxc6 with approximate equality. Perhaps 21...Bd7!? would be worth considering, making it into a genuine pawn sacrifice. Black would certainly enjoy a great deal of activity here. 21... Bd7!

Anticipating b4-b5 while threatening ... Bf5. 22g4?

The desire to prevent ... Bf5 is understandable (for instance, 22 Qxd5 is met by 22 ... Bf5! intending 23 Rbel Qb2! when the a2-knight is in trouble), but ultimately this causes too many weaknesses. 22 Qdl!? may have been the best chance. 22 ... Rfe8 23 Rfe1 hS!

Having occupied the open e-file on the previous move, Williams wastes no time in softening up the white kingside. Needless to say, if the white g-pawn can be diverted from g4 then ... Bf5 will come with great force. 24Nc1

24 h3 would be strongly met by 24 ...Qf4!, threatening ... Be5; e.g. 25 Kg2 Be5 26 Rhl d4 intending ... Bc6 with an overwhelming position. 24... Nd4! (Diagram 47)

Diagram 47 (W)

Diagram 48 (W)

White is in a mess

The end is nigh

The black knight is far more powerful than its feeble counterpart! Williams decides that if it has to be exchanged, it will be for the important f2-bishop rather than a fellow knight. 25 Bxd4 It is hard to suggest an alternative. Ward offers the sample variation 25 Bdl? Rxel + 26 Bxel Bb5! 27 Qe3 hxg4 28 fxg4? Qfl mate.

2S ...Qxd4+ 26 Qxd4 Bxd4+ 27 Kf1 hxg4 28 fxg4 Re4 (Diagram 48)

Intending to double rooks, while incidentally threatening the g-pawn. 29 Bf3 BbS+ 30 Be2

175

Starting Out: The Accelerated Dragon In case of 30 Ne2 there is 30 ... Rae8! 31 Bxe4 Rxe4 (Ward), when the duel threats of ... Rf4-f2+ and ... Rxg4-g1 mate make White's position hopeless. 30... Rae8 0-1

Theoretical? Yes and no- both sides will do well to familiarize themselves with around 15-20 moves of theory, which sounds like a lot but is actually not too complicated. Each side needs to know how to avoid certain pitfalls to reach the desired middlegame. For example, White players should remember that 17 a3? as in Holzke-Greet is a serious mistake and that 17 Rc2! is the correct way to proceed. Beyond that, positional understanding is far more important than knowledge of variations.

Statistics The main line of 14 Be3 Nc5 15 Rab1 scores 59% for White. Out of Black's major responses, 15 ... Qb6 scores 42% while 15 ...e6 scores 40%.

Summary and Conclusions The Classical Mar6czy is solid and reliable for Black. Usually the game will develop quietly in the early stages as both sides develop their forces. Black's first important choice comes at move nine when he must decide between the standard developing move 9... Bd7 or the more refined sequence of 9... Nxd4 10 Bxd4 Bd7. Statistically, it is highly probable that the two lines will converge on the same position, though we have seen that White can attempt to 'exploit' either move order and steer the game along an independent path, by means of 10 Nc2!? and 11 Qd3!? respectively. In my own games I have always played 9...Bd7, though having compared the pros and cons of the two lines more thoroughly, I would probably prefer 9 ... Nxd4 10 Bxd4 Bd7 today. Following an exchange of knights and the developing manoeuvre ... Bd7-c6, Black will be ready to increase his influence over the dark squares with ... a7-a5 and ... Nd7-c5, while offering the exchange of dark-squared bishops. In the middle part of the chapter we examined White's plan of exchanging the bishops with the intention of attacking on the kingside. I feel quite optimistic about Black's theoretical status in these lines, though one cannot ignore the fact that the exchange of bishops has scored rather well for White in practice. In spite of the above, the main line and most respected plan for White has always been to avoid the bishop exchange with 14 Be3 (or 14 Bf2). Now it is Black's turn to decide on a plan, and the traditional solution has involved the regrouping ... Qb6, ... Rfc8, ... Qd8 to be followed either by ... Qf8, or in some cases ...h7-h5, ... Kh7 and ...Qh8. This enables Black to achieve an extremely solid, if slightly passive position. The typical strategy for White used to involve the preparation of the b3-b4 advance, but Game 20 provided a perfect illustration of the shortcomings of this approach. Therefore White does better to adopt a more patient strategy as in Game 21. Finally, the move 15 ... e6!?, as in Game 23, deserves serious consideration. This is a solid approach, and we have seen that the d6-pawn only appears weak as the dark-squared bishop will find a relatively secure post on e5. At the same time Black prevents any Nd5 ideas once and for all, while maintaining reasonable prospects for active counterplay.

176

Chapter Seven

Mar6czy Bind: Gurgenidze System

-----

Black's Model Position A Typical Example An Early Deviation 6 ••• d6: 7 f3 (and Other Deviations) 7 Be2 - Introduction and Lines with an Early o-o 9 Be3 with an Early 0-0 9 BgS -the Main Line The Endgame Summary and Conclusions

Starting Out: The Accelerated Dragon In Chapter Six we examined what can happen after straightforward play from both sides. Here we will see one of the principal ways in which Black can modify his development to challenge his opponent with a different set of opening problems. Following the usual moves ... 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4

... the most obvious continuation is 5 ... Bg7, developing the bishop while attacking the knight on d4, as we saw in the previous chapter. But in the Gurgenidze System Black has something else in mind. s... Nf6 6 Nc3 d61? (Diagram 1) Once again Black delays ... Bg7. This may seem a peculiar decision, but the idea will soon become clear.

Diagram 1 (W) A refined move order

Diagram 2 Black's typical set-up

This particular method of development is named after the Georgian Grandmaster Bukhuti Gurgenidze who used it with some success in the 1950s and 60s.

What does Black have to gain by altering the move order? In the previous chapter we saw that Black's natural plan involved an exchange of knights on d4 at which point White would recapture with the e3-bishop. The distinguishing feature of the Gurgenidze System is that Black will play ... Nxd4 at an earlier stage, thus forcing White to recapture with the queen. As we will see, the earlier knight exchange leads to some major differences in the nature of the subsequent battle. WARNING: It is important to note that White should not try to prevent this plan with 7 Be3? on account of the disruptive move 7... Ng4!.

This is the reason why Black played an early ... d7-d6 instead of 6... Bg7, which might have transposed to the previous chapter after 7 Be3.

178


Black's Model Position In many ways the Gurgenidze is an ideal 'starting out' system for anyone new to the Accelerated Dragon. Black's play is fundamentally sound, he develops quickly and will usually follow a step-by-step recipe of development (see Diagram 2). This method of development occurs in the overwhelming majority of games in the Gurgenidze. Black has developed quickly and actively, and both of his bishops have their sights trained on the white queenside. He will increase the pressure on that side of the board with moves like ...Qa5 and ... Rfc8, before blasting it open with ... a7-a6 and ... b7-b5 (which can be further prepared with ... Rab8 if necessary). One of the great things about this system is that it is so easy to learn for Black; in the majority of games he can more or less follow this plan on autopilot. Naturally there are some ways for White to attempt to neutralize Black's activity, and we will learn about them over the course of the chapter.

A Typical Example I can imagine cynical readers thinking: "it's all very well to show an abstract diagram and suggest a plan of action for Black, but I bet it wouldn't work in a real game- after all, White gets to make some moves as well!" Well, the following example provides the proof. I have deliberately selected a highly plausible line which has been seen in a vast number of games, including many GM encounters. 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2

This is the main line and most popular continuation. 7... Nxd4

This early capture, forcing the reply Qxd4, is the defining characteristic of the Gurgenidze System. 8 Qxd4 Bg7 9 Be3

9 Bg5 is an important alternative here, though the two moves can often lead to similar positions. 9...0-0

Now that the g7-bishop is protected, ... Ng4 has become a serious positional threat. Therefore White sensibly removes the queen from her vulnerable position. 10 Qd2 Be6 11 o-o

White can also delay castling, with the aim of exchanging queens and keeping the king in the centre where it could become useful in an endgame. This plan will be examined later in the chapter, but the text is likely to be the choice of the average player. 11...Qas 12 Rac1

White fortifies his queenside, removes his rook from the dangerous long diagonal and supports the c3-knight in preparation for b2-b3. I think you will agree that every one of White's moves has been highly logical, and the entire sequence is exactly the kind of thing you could expect from a typical opponent. 12... Rfc8

179

Starting Out: The Accelerated Dragon I will use this, the most common move order for illustrative purposes, although later on in the chapter I will be recommending the more refined 12... a6!, for reasons that will be explained at the bottom of page 195. Then a typical continuation might be 13 f3 Rfc8 14 b3, reaching the same position as below. 13 b3 a614f3

Here the more aggressive 14 f4!? is a major alternative. (The move order with 12... a6! is specifically designed to counter this possibility.) Remember, the present sequence is being used purely for illustrative purposes. The Mar6czy Bind has a reputation for solidity and the restriction of Black's counterplay, so it is instructive to see how easily Black can counter White's most natural attempts to implement his strategy. 14... bs! (Diagram 3)

Diagram 3 (W)

Diagram 4

Easy equality after 14 moves

An endgame for Black to avoid

White's set-up may not be the most theoretically challenging, but every one of his moves has been natural and logical. In fact, starting as far back as move five, my reference database indicates that all but one of White's moves have been the most popular choice at each turn (the only exception being 12 Racl, which was played in just a few less games than the main alternative 12 Rab1). And yet Black has still managed to follow, move for move, the exact formula as laid out above! From Diagram 3 White had better bail out to an equal endgame with ... 15 Nds

Instead, 15 cxb5?! axb5 would leave him struggling to equalize. 16 Bxb5? is definitely a bad idea due to 16... Rxc3! 17 Rxc3 Qxb5, while 16 Nxb5 (R.Panjwani-J.Yoos, Toronto 2004) leads by force to a better endgame for Black after 16... Qxd2 17 Rxc8+ Bxc8 18 Bxd2 Rxa2 19 Rd1 Be6, when the b-pawn is weak. 15 ...Qxd2 16 Bxd2 Nxds

and now 17 cxds or 17 exds should be met by 17 ... Bd4+! 18 Kh1 Bd7 with equal chances, as seen in several games.

180

Mar6czy Bind: Gurgenidze System Before moving on, I wish to make the reader aware of a type of endgame that Black should generally try to avoid. The endgame in Diagram 4 could arise in certain variations where White delays castling and meets ...Qa5 with Nd5, after which Black has exchanged queens and captured a white knight on d5 with his e6-bishop (instead of the f6-knight). These seemingly minor differences can turn an equal endgame into a troublesome one for Black.

0

NOTE: If the black b-pawn had already advanced to bS then his problems would be even greater, as it would be very easy for White to undermine the queenside with a timely a2-a4.

Even with the b-pawn back on b7, White has the simple and strong plan of placing his light-squared bishop on h3, threatening to harass the black queenside pawns from c8.

0

NOTE: The bishop manoeuvre to h3 is even more powerful when there are rooks are on the board, as the bishop's attack on c8 will force Black to relinquish the open c-file.

In order to counter this plan Black may have to play ...e7-e6, but after the response d5xe6 ...f7xe6, the game is starting to open up for the white bishops. In this type of position Black will generally have to suffer for a long time before he can hope for equality.

An Early Deviation Before looking at any main lines, let us first briefly examine the consequences of an 'Accelerated Gurgenidze' system. Fans of the Nc2 systems for White (and those who wish to avoid it as Black) should be aware that Black can exchange knights immediately with ... 6... Nxd4!? 7 Qxd4 d6 (Diagram 5)

Diagram 5 (W}

Diagram 6 (B)

The 'Accelerated Gurgenidze'

The bishop is better on d3 than e2

This move order was frequently used by Gurgenidze himself. Nowadays it is considered somewhat inaccurate, with good reason as we shall see. However, it is not mentioned by

181

Starting Out: The Accelerated Dragon Nielsen in Experts vs. the Sicilian and, despite its dubious reputation, may prove a worthwhile move order if your opponent happens to be a dedicated Nc2 adherent. Likewise, if the reader is a fan of the Nc2 system then the following section will show you how to play against this variation. If White wants to highlight the drawback of Black's move order then he should develop accurately and actively with ...

s Bgs! 8 Be2 Bg7 would transpose to the main line of this chapter, with Black having achieved his objective of preventing the Nc2 system. 8... Bg7 9 Qd2 0-0

There is not much to be gained from delaying castling. 9... Be6 is possible, but the typical prophylactic move 10 Rcl! ensures that White will be able to stabilize the queenside, as 10... Rc8 can now be safely met by 11 b3. After the text, 10 Be2 would reach the main line, while 10 f3 is met by 10...Be6 intending the standard plan of ... Qa5, ... Rfc8, ... a7-a6 and ... b7-b5. But White has something better: 10 Bd3! (Diagram 6)

This is the best way for White to exploit the early knight exchange. The bishop is more useful on d3 than e2 as it lends extra support to the e4-pawn, enabling ideas such as a timely Nd5 without wasting time with f2-f3. Nevertheless, Black's disadvantage is hardly a fatal one. True, he has a slightly inferior version of a normal line, but if you happen to know that your opponent wants to play Nc2 then the trade-off may be a profitable one. Conversely, if you like the Nc2 lines as White, it may be worth familiarizing yourself with some of the ideas in the present chapter, in order to be ready for an opponent who attempts to confuse you with this move order.

Theoretical? The only really important thing for White to remember is that the king's bishop should go to d3. Apart from that detail, a good general understanding of the position will be more important than theory.

Statistics Black scores 41% after 7 ... d6, with White's score rising to 66% if he gets as far as 10 Bd3!. We will now begin to examine some of the important variations after the usual6 ... d6.

6... d6: 7 f3 (and Other Deviations) 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 f3!? (Diagram 7)

Let us briefly note that 7 Nc2 reaches Chapter Nine, while we have already learned that 7 Be3? is a bad idea in view of 7... Ng4!. The text is considered by most sources to be slightly less critical than 7 Be2, but still demands accurate play from Black. While it may appear slightly cumbersome to move the fpawn at such an early stage, it does bring two potential advantages over the bishop development:

182

Mar6czy Bind: Gurgenidze System 1) The immediate fortification of the e4-pawn may facilitate an early Nc3-d5 jump at a time when Black is not fully prepared. 2) In certain types of endgames, such as that shown in Diagram 4 above, the bishop may actually be better off on f1 than e2, due to the strong possibility of g2-g3 and Bh3. 7... Nxd4

As is customary for the Gurgenidze, Black exchanges at a time when White is forced to recapture with the queen. 8 Qxd4 Bg7 9 Be3 9 BgS is seen only occasionally. There is nothing wrong with this move, but it looks more logical to put this piece on e3 to cover White's dark squares which have become slightly more sensitive after f2-f3. 9 ... 0-0 10 Qd2 (Diagram 8)

Diagram 7 (B) A solid and respectable line

Diagram 8 (B) Black faces a major crossroads

Now Black can choose between two contrasting plans. The most common option involves the traditional Gurgenidze scheme with ... Be6 and ...QaS. Black seems to be holding his own in this line, though he needs to demonstrate a certain degree of theoretical knowledge, and in many variations White does not really risk losing. Therefore some strong players have preferred a more modern approach involving ...a7-a5-a4, gaining some valuable queenside territory and leading a complex middlegame with more chances to play for a win. Nowadays 7 f3 is only seldom encountered, so I have decided only to examine the more modern plan in detail. I believe this will be a better choice for most players as it requires far less theoretical knowledge without compromising on soundness. Before moving on to my main recommendation, here is a brief summary of the traditional approach. 10... Be6 11 Rc1!

This essential developing move is the most accurate choice in the position.

183

Starting Out: The Accelerated Dragon TIP: In practically all variations of the Gurgenidze the prophylactic move Ra1-c1 (or occasionally -b1) will be indispensable for White. The reason is that after the inevitable ... Rfc8 he will need to defend his c4-pawn with b2b3. If he were to do this with a rook on a1, his position would quickly be destroyed by ... Nxe4. It is also useful to guard the tactically vulnerable knight on c3. Finally, in the event of Nd5

followed by an exchange on d5, the rook can contest the open file after the possible recapture c4xd5. 11...Qa5 12 NdSI?

White can transpose to a 7 Be2 main line with 12 Be2 Rfc8 13 b3, but it is the text move, exploiting the early defence of the e4-pawn, which truly defines the 7 f3 system. The main idea is to threaten the e7-pawn before Black can arrange a convenient defence. 12... Qxa2! ~

-w

~n

WARNING: 12... Qxd2+? is a serious mistake here and should be avoided at all costs.

The is that after 13 Kxd2 Black has no good way of dealing with the threat of Nxe7+. 13 ... Nxd5leads to big problems after 14 cxd5 Bd715 Rc7 or 14 ... Bc815 b3. In L.Polugaevsky-P.Ostojic, Belgrade 1969, Black tried 13 ... Bxd5 14 cxd5 Rfc8, but after 15 Rxc8+ Rxc8 16 g3! he faced a highly unpleasant endgame. 13 Nxe7+ Kh8 (Diagram 9)

Diagram 9 (W)

Diagram 10 (W)

Structure vs. development

Pawn rather than queen to aS

A very interesting position has been reached and the theory from this point runs well into the endgame. The ultimate question is whether Black's lead in development will prove sufficient to offset his structural inferiority. The answer to that question seems to be yes, but he will certainly need to play accurately and in many variations the most that he can hope for is a draw. 14 Be2 is usually considered the most challenging move here; 14 Bd4 is also quite playable. I will not delve any deeper into the matter in these pages. Detailed analysis of

184

Mar6czy Bind: Gurgenidze System these moves can be found in numerous other sources. In this book I wish to recommend what I believe to be a more promising and pragmatic approach for Black.

The Modern Interpretation 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 f3!? Nxd4 8 Qxd4 Bg7 9 Be3 o-o 10 Qd2 aS!? (Diagram 10)

This variation demonstrates the flexibility of the Accelerated Dragon. Now we have a position resembling the Classical Mar6czy, with some small yet significant differences. The first thing to note is that if Black naively aims for an identical set-up to the Classical Mar6czy (i.e. with ... Bd7-c6 and ... Nd7-cS) then he will simply end up a tempo down, as here White has played the moves Qxd4, Be3 and Qd2 compared with Be3-xd4-e3 and Qd2 in the Classical. Therefore Black must look to maximize his chances in some other way. The important points can be summarized as follows: 1) The light-squared bishop will usually be deployed on e6 instead of c6. NOTE: One must always remember that a small change in the position can turn a strong piece manoeuvre into a weak one, and vice versa.

If we cast our minds back to the Classical Mar6czy, let's stop and think for a moment about why the bishop manoeuvre to c6 was a good idea. Well, the first point was that it actually created a threat against the e4-pawn, which White would usually answer with f2-f3, whereas in the present position White already has a pawn on f3. So how about the e6square? A glance at page 1S4 will remind us that set-ups with ... Be6 can be strongly countered by f2-f4-fS. Here that plan will cost White extra time since he has already played f2-f3; nor does he enjoy the benefit of the strong bishop on d4. The final conclusion is obvious: in the present variation, the bishop is likely to be better off on e6 than c6. 2) Black is well placed here to obtain counterplay with a pawn advance to a4. His queen may come to an active post on aS, and combined with moves like ... Be6 and ... Rfc8 he can exert strong queenside pressure. 3) Finally, we should note that the typical knight manoeuvre ... Nd7-cS will almost always feature somewhere in Black's plans. 11Be2

This is nearly always played, although my first reaction was to wonder whether 11 b3!? Diagram 11) might be a good idea (compare the previous chapter in which 12... aS was met by 13 b3! to ensure that ... aS-a4 could be answered by b3-b4). The present situation is slightly different, and we should investigate whether either side is in a position to benefit from the specific features of the position. Tsesarsky calls 11 b3 dubious, offering 1l...a4 12 b4 Be6 13 Rd1 Qc8 14 NdS BxdS1S exdS (15 cxdS? loses to 1S... Nxe4!) 1S ... e6! 16 dxe6 Qxe6 with decent counterplay. However, in E.Andreev-E.Sazhin, Cherepovets 200S, White improved with 13 Rcl! and obtained a clear advantage after 13... a314 Be2 Rc8 1S NbS Ra8 16 0-0. A short while after this game Sazhin faced 11 b3 again. On this occasion he preferred 1l...Be6!? and obtained a fully satisfactory position after 12 Rcl Nd713 Bd3 NcS14 Bb1 Qb6 1S 0-0 Qb4 in S.Dovliatov-E.Sazhin, Cherepovets 200S. So it appears that Black does better to keep the ... aS-a4 advance in reserve, in accordance with the old teaching: "The threat is stronger than the execution".

185


Diagram 11 (B)

Diagram 12 (B)

Is ... as-a4 the right move?

A variation on the Classical

11 ... a4!

This pawn thrust forms an integral part of Black's strategy. From here it disrupts the harmony of the white queenside and makes room for the queen to take up an active post on aS.

0

NOTE: It is important for both sides to remember that the queen will be defended on as, thus eliminating certain NdS tricks.

12 o-o (Diagram 12)

12 Rcl is sometimes seen, but it seems to me that White might as well perform the essential task of castling before deciding where to place his rooks. One top level encounter continued 12 ... Be613 Nd5 Nd7! 14 0-0 Nc5 with nice play for Black in J.Polgar-S.Tiviakov, Chalkidiki 2002.

0

NOTE: With this in mind, it may be worth considering a move order with 10... Be6, waiting for 11 Rc1 before reverting back to the modern plan with 11... as!?. This is virtually untested but may be helpful in restricting White's options.

12 ... Qas

An important moment. Although the text is not a bad move, my personal view is that it is time for Black to attend to the development of his pieces, and to that end 12...Be6 looks best. This can be found in Game 24. However, the text will also help to illustrate some of the positional and tactical features of this position. 13 Rac1 Bd7

13 ... Be6 is possible, but if Black wishes to develop the bishop here he should do it on the previous move, as White can gain the advantage with 14 Nd5!. Now either exchange on d5 would be met by c4xd5 leading to a comfortable endgame for White, so Black probably has to play 14... Rfe8 to defend e7 (one reason why I would prefer not to rush with ...QaS). Following 15 QxaS Rxa516 Nc7 Rc817 Nxe6 fxe618 Rfd1 Kf719 Rc2 White has a slight but stable advantage, as pointed out by Tsesarsky, due to his superior structure and pair of bishops.

186

Mar6czy Bind: Gurgenidze System 14 cS! (Diagram 13)

Diagram 13 (B)

Diagram 14 (B)

A strong tactical device

Black is under some pressure

This accurate move exposes another drawback of Black's early queen development. We will follow the game G.Kasparov-B.Alterman, Tel Aviv (clock simul) 1998. 14... Bc6

14... dxc5? drops a piece to 15 e5. 15 cxd6 Rfd816 Nb11 Qxd217 Nxd2 exd6

17... Rxd618 Nc4 Rdd8 19 Nb6 Ra5 20 Rfd1 is better for White- Tsesarsky. 18 Nc4 d5 19 Nb6 Ra5 20 e5 Nd7 21 f4 (Diagram 14) 21... Nf8

21...Bb5 22 Bxb5 Rxb5 23 Nxd7 Rxd7 24 ReB+ Bf8 25 f5 wins for White- Tsesarsky. After Alterman's move Kasparov decided to force a draw with ... 22 Bd2 Ra7 23 Be3 Ra5 24 Bd2 Ra7 25 Be3

Yz-Yz.

Without intending any disrespect to Alterman, this decision probably had something to do with the fact that this game was part of a simultaneous exhibition. In a normal game I suspect that the then World Champion would have continued the fight with 22 Rfd1 or 23 Rfd1, with a slight advantage in either case. Black was slightly worse here, but in Game 24 we will investigate what I believe to be a more accurate move order in 12... Be6.

Theoretical? The traditional lines involving ... Be6, ...Qa5 and ... Qxa2 require a considerable amount of theoretical preparation, especially from Black's side. By contrast, the modern plan with ... a7-a5 requires minimal knowledge of variations.

Statistics 7 f3 Nxd4 8 Qxd4 Bg7 scores 59% for White. The old main line with 12...Qxa2 produces the same score, whilst 10... a5 yields a slightly better 45% for Black.

187



D A.Volokitin •

D.Jakovenko

Russian Team Championship 2007

1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7 9 Be3 0-0 10 Qd2 as!? 11 f3 a412 o-o Be6! (Diagram 15)

Diagram 15 (W)

Diagram 16 (B)

Probably Black's best

A balanced semi-endgame

As I explained earlier, I believe this to be more flexible and accurate than 12...Qa5. 13 Rab1!?

Preparing b2-b4. White is positioning his rooks in the same way as the previous chapter, although the opening of the a-file does not combine well with the removal of White's rook from al. However, the immediate 13 b4 would be met by 13 ... Rc8! attacking c4, rather than 13 ... axb314 axb3 when White is well placed to contest the open file. 13 ... Qas

13 ... Nd714 b4!? axb3 15 axb3looks slightly better for White. 14 Rfc1

Reinforcing c3. White probably rejected the immediate 14 b4 on account of 14... axb315 axb3 Ng4! 16 Bd4 Bxd4+ 17 Qxd4 Ne5, when Black has achieved the desirable exchange of bishops, while the open a-file is a useful source of activity. 14 Nd5 is nothing special either after 14 ... Qxd2 15 Bxd2 (15 Nxe7+ Kh8 16 Bxd2 Rfe8 17 Nd5 Nxd5 18 cxd5 Bxd5 is more than ok for Black) 15 ... Nxd5 16 cxd5 (16 exd5? Bf5 is no good) and now 16... Bd4+ 17 Kh1 Bd7looks about level. 14... Rfc8 15 b4

15 Nd5 Qxd2 16 Bxd2 (16 Nxe7+?? loses a piece after 16... Kf8) 16... Nxd517 cxd5 Bd7 is equal again.

188

Mar6czy Bind: Gurgenidze System 15 ... axb3 16 axb3 Qb4! 17 Rc2 Nd7

Black p~oceeds with the thematic knight manoeuvre. 18 Nd5 Qxd2 19 Rxd2 (Diagram 16) 19... Kf8

19... Bxd5 also seems playable; e.g. 20 Rxd5 (20 exdS? Bc3! followed by ... Bb4 creates a nice dark square blockade, while 20 cxd5 is met by 20 ... Rc3) 20 ... Ra2 21 Kf1 Nc5, though it is possible that White's pair of bishops may count for a little something. 20b4

20 Nf4?! Bc3! followed by ... Bb4looks good for Black; the potential doubled e-pawns are nothing to worry about here. 2o... Ra3 21 Kf2 f5!?

An ambitious approach, leading to unclear complications. 22 exf5

Also possible was 22 Bg5 Bxd5 23 exd5 Nb6 24 c5 Na4 25 cxd6 exd6 26 Bf4 BeS 27 Bxe5 dxe5 28 d6 Nc3 29 Rcl with a difficult ending to evaluate. 22 ... Bxf5 23 Rbd1 Nf6 24 Nb6 Rc6 25 c5 (Diagram 17)

Diagram 17 (B)

Diagram 18 (W}

An exchange sac beckons

Black emerges on top

The threat was ... Rxe3 followed by ...Rxb6. But even with the knight defended, the exchange sacrifice turns out to be rather powerful. 25 ... Rxe3! 26 Kxe3 Bh6+ 27 f4 dxc5 28 Rd8+

28 bxc5 could have met by 28 ... e5 29 g3 Rxc5. 28... Kg7 29 bxc5 Re6+ 30 Kf2 Bxf4 31 Nd5 Nxd5 32 R1xd5 Be3+ 33 Kf1 Be4 (Diagram 18)

Black's aggressive policy which started with 21...f5!? has paid off. He will emerge with two pawns for the exchange and a strong initiative, while the white king is beginning to look uncomfortable. 34 R5d7 Bc6 35 Bf3 Bb5+ 36 Be2 Rf6+ 37 Ke1 Bf2+ 38 Kd1 Ba4+ 39 Kc1 Bxc5!?

39 ...Be3+ 40 Kbl Bxd7 41 Rxd7 Bxc5 42 Rxb7leads to a position in which the oppositecoloured bishops give White good drawing chances.

189

Starting Out: The Accelerated Dragon 40 Rxb7 Bc6 41 Rdd7!?

An amusing self-fork, hoping to obtain an opposite bishop ending as in the above note. 41... Ba3+ 42 Kdl Ba4+! 43 Kel Bb4+! 44 Rxb4 Bxd7 (Diagram 19)

Thanks to Black's clever play, White has been forced to give up his rook for the 'wrong' bishop.

Diagram 19 (W)

Diagram 20 {W)

Can White draw?

Black's advantage is decisive

When I first saw this endgame I assumed that White ought to have some way of drawing, given the extremely limited material remaining, but having examined it more closely I am not at all sure. Here I will just state that White faces a difficult and unpleasant defensive task. Fascinating as it would be to search for a final evaluation, this is not an endgame book and so I will leave it to the reader to delve into the mystery if he so wishes. 45 g3 Bc6 46 Rf4 Rd6 47 Rb4 gS 48 h4 h6 49 Kf2 Rd2 50 Rb6 BdS 51 Ke3 Ra2 52 hxg5 hxgS 53 Rbl Ra3+ 54 Kf2 Kf6 55 Rdl e6 56 Rd2 Rc3 57 Rdl KeS 58 Rel Kd4 59 Rdl+ KcS 60 Ral eS 61 RaS+ Kd4 62 Ra4+ Bc4 63 Bxc4 Rxc4 64 Ra2 Kd3 65 Ra3+ Rc3 66 Ras Rc2+ {Diagram 20) 67 Kf3?

67 Kfl would have put up more resistance, though Black still wins after 67 ... Ke4! 68 Ra4+ Kf3. 67 ... e4+ 68 Kg4 e3 69 ReS Rcl 70 Kxgs e2 71 g4 e1Q 72 Rxel Rxel

Thanks to the magic of tablebases, I can report that it is now mate in 20 with best play. 73 Kf6 Rfl+! 74 Ke6 Rgl 75 KfS Kd4 76 gS KdS 77 Kf6 Kd6 78 g6 Rfl+ 79 Kg7 Ke7 80 Kh7 Rgl 0-1

This was a superb game; not only was it relevant to our study of the opening, but Black's fighting spirit and endgame technique were also first rate. 7 f3 is a valid choice for White and provides a clear, no-nonsense method of countering the typical Gurgenidze plan. Most of the traditional lines lead to near-equality with a slight hint of an advantage for White. Theoretically Black has no major difficulties in any case, but I hope to have convinced the reader of the benefits of the modern ... a7-a5, which leads to

190

Mar6czy Bind: Gurgenidze System more interesting positions with less emphasis on memory and more chances to play for the win.

7 Be2 - Introduction and Lines with an Early 0-0 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2

Although 7 f3 has its merits, the text is justifiably recognized as the main line. White develops a piece while taking control of the g4-square, thereby enabling Be3. 7... Nxd4

This is the characteristic move for the Gurgenidze System. 7... Bg7 will almost certainly transpose to a different variation after 8 Be3 or 8 Nc2!?. 8 Qxd4 Bg7 (Diagram 21)

Diagram 21 (W} Starting on the main line

Diagram 22 (W)

White usually moves the queen

An important moment. White's most popular strategy here involves postponing castling in order to arrange his pieces in an optimal fashion. He can do this by developing his darksquared bishop on either e3 or g5. The idea is that after 9 ...0-0 he will be able to retreat his queen to its best square on d2 without blocking the bishop in on cl. Only then will he make a decision concerning castling. In this section we will consider the ways in which the game may develop should White castle immediately. 9 o-o o-o (Diagram 22)

Even though Black has no immediate threats, the queen is obviously not going to be able to remain on d4 for long, so White usually retreats it to e3 or d3. a) After 10 Qe3 Black should continue developing with 10... Be6 (Diagram 23). In the event of an exchange on b6 the doubled b-pawns are nowhere near as weak as they appear, and if White is not careful Black's lead in development and open queenside files can provide him with a powerful initiative. This point is well illustrated in the following variations: a1) 11 Rb1 Qb6! and now 12 Qxb6 axb6 13 Be3 Rfc814 b3 b5! (Diagram 24) is very good for

191

Starting Out: The Accelerated Dragon Black (ADP), while 12 Qd3 Nd7 13 NdS Bxd514 exdS Rfc8 15 Qh3 NcS 16 BgS Ne4 17 Be3 Bd4 18 Qh4 Bxe3 19 Qxe4 Bd4 20 b4 Bf6 21 Rfcl Qd4 22 Qxd4 Bxd4 23 Rc2 b6 was completely equal in F.Nijboer-S.Tiviakov, Mondariz 2000.

Diagram 23 (W) Black intends the strong ... Qb6!

t:') 'i{

Diagram 24 (W) A dream position for Black

TIP: When the white queen retreats to e3, Black's most important resource is the odd-looking ...Qb6, offering to trade queens.

a2) 11 Bd2 is also well met by 11...Qb6! (Diagram 25).

Diagram 25 (W) Black shouldn't fear doubled pawns

Diagram 26 (W) White is under pressure

Once again this typical move solves all of Black's opening problems. Now 12 b3 is probably best, when 12... Qxe3 13 Bxe3 Nd714 Racl NcS (or 14... Rfc8) has been seen in a number of games, and leads to a position reminiscent of the previous chapter. In the top class encounter A.Morozevich-V.Topalov, Wijk aan Zee 2001, White varied with

192

Mar6czy Bind: Gurgenidze System 12 Qxb6 axb613 a4, but after 13 ... Nd7! Black became very active, and following 14 Ra3 Nc5 (Diagram 26) 15 Rfa1 Bxc3 Topalov was winning a pawn, though Morozevich managed to steer the game towards an opposite bishop ending after 16 Bxc3 Nxe4 17 Bd4 Rfc8 18 Re3 Bxc4 19 Rxe4 Bxe2 20 Bc3 Bb5 21 Rxe7 Bxa4 22 Rxb7 Bc6 23 Rba7 Rxa7 24 Rxa7, when a draw was soon agreed. b) 10 Qd3 (Diagram 27)

Diagram 27 (B)

Diagram 28 (W)

Another non-critical line

Equality has been achieved

This avoids running into the powerful ...Qb6 idea from line 'a'. 10... Be6 Black almost always develops the bishop here in the Gurgenidze System, and the present variation is no exception. 11 Bd2!? The idea of this move is to discourage the standard development of the black queen on a5. White is playing cautiously, attempting to pre-empt any potential counterplay before gradually creeping forward. 11 ... Nd7!

TIP: One of the disadvantages of the queen's retreat to d3 is that Black can execute the thematic transfer of a knight to cs with gain of tempo. This looks like the most logical move, as in the game Z.Almasi-S.Tiviakov, Cacak 1996. 12 b3 as 13 Rac1 Ncs 14 Qe3 Bd71? Tiviakov adjusts his plan slightly. 15 f4 Bc6 16 Bf3 e6! (Diagram 28) A good move, preventing a knight hop to d5, while preparing to complete development by ... Qe7 and ... Rfc8. After the further moves 17 Kh1 Qe718 NbS a draw was agreed.

193


Theoretical? These lines tend not to lead to heavy complications and come with only minimal theoretical baggage. Indeed, White often plays them with the express intention of drawing his opponent away from the main lines. This approach is perfectly sound, but Black only needs to remember a few key ideas (e.g. meeting Qe3 with ... Qb6!) to obtain a comfortable position.

Statistics After 9 0-0 0-0, 10 Qd3 scores 56% while 10 Qe3 has managed a surprisingly high 60%. We will now begin to investigate some of the most popular lines of the Gurgenidze System.

9 Be3 with an Early 0-0 After the opening moves ... 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7

... the most harmonious way for White to arrange his pieces is to develop his dark-squared bishop on e3 or g5, followed by a queen retreat to d2. 9 Be3

9 Bg5 will considered later in the chapter. There is not much to choose between the two bishop moves, but a summary of the main differences between them can be found at the start of the 9 Bg5 section on page 203. 9 ...0-o 10 Qd2 Be6

As usual Black has little to gain by deviating from the usual plan. 10... Ng4 11 Bxg4 Bxg4 12 Bd4 Bxd4 13 Qxd4 looks comfortable for White, while 10... a6 11 £3 Bd712 0-0 Rb8 13 a4 also left Black somewhat passively placed in M.Peretz-Y.Richter, Netanya 1971. 110-0

An important moment; White's other main plan involves delaying castling, with the intention of offering a queen exchange on d2 and hoping that the centralized king will be an asset in the endgame. He can do this with 11 Rcl Qa5 12 f3 Rfc8 13 b3, which leads to very similar positions (and sometimes a direct transposition) to the 'endgame variation' covered under 9 Bg5. Please refer to the relevant section beginning on page 206 for a full discussion of this important line. At present we will investigate the consequences of White's immediate castling. This natural move has been a popular choice at all levels, and Black certainly needs to treat it with respect. First of all, he should play the standard move: 11... Qas (Diagram 29) At this point White must make a far-reaching strategic decision. We already know that Black's typical plan will consist of ... Rfc8, ... a7-a6 and ... b7-b5, so White needs to decide how best to deal with that. Essentially he has two main options and we will examine each approach in turn. The first strategy involves the aggressive plan of launching the £-pawn up the board. This can lead to complex play, but we will see that with the aid of a precise move order Black

194

Mar6czy Bind: Gurgenidze System can conveniently avoid the most critical lines while maintaining a full share of the chances.

Diagram 29 (W)

Diagram 30 (W)

A strategic crossroads for White

A very complicated position

The aggressive plan If White is intending to attack in the centre and/or on the kingside, he will usually begin with the prophylactic move 12 Rac1. By keeping the other rook on f1 for the time being, White hints at a possible advance of the £-pawn. To get an idea of what he is planning, let's see how the game might develop after the typical move: 12 ... Rfc8 This is not forced; in fact this is one of the rare occasions on which I recommend that Black deviates slightly from the standard scheme. More accurate in my opinion is 12... a6!, as we will soon see. 13 b3 a6 14 f4!? 14£3 b5! was mentioned at the start of the chapter. 14... bs 15 fs Bd7 16 fxg6 hxg6 (Diagram 30) The position is extremely complex, with 17 e5!? and 17 c5!? being the main possibilities. This whole variation used to be considered too risky for Black, though the modern consensus is that the second player is probably holding his own. :"Jevertheless there is a far more straightforward and economical solution available in ... 12 ... a6! (Diagram 31) ... preparing an immediate queenside expansion. The general rule of thumb in the Gurgenidze is that ... Rfc8 should come before ... a7-a6, but the present position is an exception. This small refinement completely negates the £4-£5 plan without making any other concessions.

TIP: Whenever Black suspects that White may be intending f4-f5, he should almost always delay the standard ... Rfc8 in favour of a preliminary ... a7-a6.

195

Starting Out: The Accelerated Dragon This enables him to meet f2-f4 with the immediate ... b7-b5 break, while keeping the rook on f8 just in case it may be needed to defend the kingside. We will encounter an analogous situation in the 9 Bg5 variation.

Diagram 31 (W) An exception to the rule

Diagram 32 (W) Black is absolutely fine

Let's see what happens if White proceeds with the plan of kingside aggression. 13 f4 13 f3 Rfc8 14 b3 b5! was dealt with at the start of the chapter. The attempt to exploit the omission of ... Rfc8 with 13 Nd5? turns out badly after 13 ...Qxd2 14 Nxe7+ Kh8 15 Bxd2 Nxe4, while 13 b3 b5! also sees Black achieve his desired counterplay. 13 ... bs! 14 cxbs 14 f5? Bxc4 15 Bxc4 bxc4 was clearly better for Black in A.Riccetto-R.Silva Nazzari, Montevideo 1976. 14... axbs 15 fs 15 Bxb5? Nxe4 16 Nxe4 Qxb5 is very bad for White. After 15 f5, the simple 15...Bc4! (Diagram 32) looks best to me, when Black is absolutely fine. To summarize, from an objective theoretical perspective 12 ... Rfc8 is probably a valid option. However, the variations are very complicated and it makes far more sense to play 12 ... a6! which avoids the whole issue and guarantees Black a fair share of the chances by relatively simple means.

The positional plan White's alternative plan is perhaps more typical for the Mar6czy Bind, and involves stabilizing the position with the intention of gradually expanding on the queenside while inhibiting any counterplay. If White wishes to play this way then he should place his rooks on b1 and cl. 12 Rab1!? may be the most accurate way to implement this plan, and will be discussed in Game 25. Here is a summary of the other possibilities:

196

Mar6czy Bind: Gurgenidze System a) 12 Rfc1 is another way of aiming for the 'positional' set-up, but it is slightly inaccurate due to the surprising tactical shot 12 ... Nxe4! (Diagram 33), which could well come as an unpleasant surprise to your opponents.

Diagram 33 (W)

Diagram 34 (B)

An effective combination

A complex, but equal endgame

There now follows the forcing sequence 13 Nxe4 Qxd2 14 Nxd2 Bxb2 15 Bf3 Rab8! (15 ... Rfb8 was suggested by Psakhis, but this leaves Black's queenside pieces very passive and it actually looks preferable to give up the a-pawn to open the a-file) 16 Bxa7 Ra8 17 Be3 (Diagram 34). This position has been reached a couple of times. P.Popovic-D.Velimirovic, Ulcinj 1998, which soon led to a draw by repetition after 17... Rfb8 18 Re1 Bxa1 19 Rxa1 Ra3 20 h3 Bd7 21 Nb1 Ra4 22 Be2 Be6 23 Nc3 Ra3 24 Nb5 Ra4. In K.Van der Weide-S.Tiviakov, Hoogeveen 1999, Black tried 17... Bxc118 Rxcl Rxa2, but the final result was the same after 19 Bxb7 Rb8 20 Be4 Rc8 21 Bd3 d5 22 c5 d4 23 Bh6 Ra3 24 Bfl Rc3 25 Nf3 R3xc5 26 Rxc5 Rxc5 27 Nxd4 Bc4 28 Be3 112-1/2. White may not be losing in either of these lines, but he certainly has no trace of an advantage. Therefore 12 Rfcl must be viewed as somewhat inaccurate for White. b) 12 f3 is a typical stabilizing move in the Maroczy Bind, but in this position a potential drawback can be seen after 12 ... Rfc8 13 b3 Ng4!? (13 ... a6 is also quite playable). Now 14 Bd4 Bxd4+ 15 Qxd4 Qe5! 16 Qxe5 Nxe517 f4 Nc6 should be quite alright for Black, who has reached a safe endgame having achieved the desirable bishop exchange. Instead, some players have experimented with the temporary pawn sacrifice 13 Rfc1!? (Diagram 35). This may be viewed as an attempt to improve on 12 Rfc1, aiming for the same type of position without allowing the 12... Nxe4! trick. Now after the obvious 13... Bxc414 Nd5 Qxd2 15 Nxe7+ Kf8 16 Bxd2 Kxe7 17 Bxc4 Nd718 Rab1 White is a little better thanks to his pair of bishops. However, Black can postpone the capture with 13 ... Nd7!?. A.RodriguezS.Tiviakov, Ubeda 1998, continued 14 Rab1 (White has nothing better; 14 b3 b5! was already difficult for him in M.Hawelko-A.Pekarek, Polanica Zdroj 1986) 14... Bxc4 15 Nd5 Qxd2 16 Nxe7+ Kf8 17 Bxd2 Bd4+! 18 Kh1 Kxe7 when Black had obtained an improved version of the aforementioned ending, as the white king has been forced away from the centre,

197

Starting Out: The Accelerated Dragon while the black knight may prove useful on the queenside. The game went on 19 Bxc4 Nb6 20 Bb3 (other bishop retreats would allow 20... Na4) 20 ... Rc6!? (20 ... Rxcl + 21 Rxcl Rc8 22 Rxc8 Nxc8 23 Bd5 was a little better for White in K.Aseev-S.Vokarev, Russian Ch., Elista 1996) 21 Rd1 Rac8 (Diagram 36).

Diagram 35 (B) An interesting pseudo-sacrifice

Diagram 36 (W} Chances are balanced

The position is approximately equal; White has the bishop pair and sounder pawn structure, but all of Black's pieces are active. If White is able to manoeuvre very skilfully then he may be able to claim some sort of an advantage, but generally I would not consider Black to be in any great danger.

Theoretical? The lines with 9 Be3 and a later 0-0 do not require extensive preparation by either colour, although a certain amount of specific knowledge will always prove useful.

Statistics 11 0-0 scores a total of 55% for White. Following the usual11...Qa5, the aggressive 12 Racl scores 52%, while the more positional12 Rab1 yields a more promising 63%. This option will be considered in the following game.


D J.Rowson • V.Malakhov Selfoss 2003 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7 9 Be3 Qd2 Be611 o-o Qa5 12 Rab1 Rfc8 (Diagram 37)

198

o-o 10


Diagram 37 (W)

Diagram 38 (W)

Black follows the standard recipe

Black has no problems

13 b3 On 13 b4 the queen can retreat tidily with 13 ...Qd8, when Black can feel quite happy with the king's rook already having neatly slotted over to c8. R.Hanel-H.Stefansson, Vienna 1991, continued 14 c5 a515 a3 axb416 axb4 Ra317 NbS Ra2 18 Qd3, and now 18 ... dxc5 19 bxc5 Qxd3 20 Bxd3 Ng4 21 Bd4 Rd2looks nice for Black. Occasionally White tries 13 Rfcl ?!, but if he wishes to play like this he would have been better off with 12 f3 Rfc8 13 Rfcl as examined in line 'b' prior to this game. The vulnerability of the e4-pawn will prove significant after 13... Bxc4 14 Nd5 Qxd2 15 Nxe7+ Kf8 16 Bxd2 Nxe4! (Diagram 38). N.Sulava-N.Brunner, Metz 2003, continued 17 Nxc8 Nxd2 18 Bxc4, and now Khalifman's recommendation of 18 ... Rxc819 Bd5 Ra8! would have left White with some problems. 13 ... Ng4!? This is quite an important moment. At first glance the tactical blow 13 ...b5 looks very attractive, but this is probably insufficient for equality after 14 b4! Qc7 (14 ... Qd815 cxb5 Rxc3 does not work due to 16 Qxc3 Nxe4 17 Qe1) 15 e5! dxe516 Nxb5 Qb717 Rfcl, when White's powerful queenside pawns eventually proved decisive in V.lvanchuk-A.Kovacevic, Belgrade 1997. A better try is 13 ... a6, when White should continue to strengthen his queenside with 14 Rfcl (Diagram 39) and now: a) 14...b5?! 15 b4 Qd816 cxb5 axb517 Bxb5left Black with insufficient compensation in V.lvanchuk-A.Khalifman, Yalta (rapid) 1995. b) 14... Rab815 f3 (15 Ba7 Ra816 Be3 Rab8 repeats, but White must take care to avoid 16 Bd4? Nxe4!) 15... b5? (15 ... Nd7 16 b4 Qd8 17 Nd5 leaves White with a slight edge, though everything is still to play for) 16 b4! Qd8 17 cxb5 axb518 Nxb5 saw White emerge with a clear extra pawn in J.Rowson-R.Palliser, British Ch., Scarborough 1999. c) 14...Qb4!? 15 f3 b5 has been recommended for Black on the basis of R.Rodriguez LopezC.Nanu, Szeged 1998, which continued 16 cxb5 axb5 17 a4 Rxc3 18 Rxc3 Nd5 19 Rd3 Nc3 20

199

Starting Out: The Accelerated Dragon Rcl Nxe2+ 21 Qxe2 bxa4 22 bxa4 Bc4 23 Qd2 Qxa4, when Black regained the exchange to reach an endgame a pawn up. This is all well and good, but unfortunately 19 Reel! Nc3 20 Bxb5 seems to refute the combination. Instead, Black should prefer 17...bxa418 bxa4 Qa5, when the evaluation hinges on whether the passed a-pawn will be a strength or a weakness. For the moment I would tend to go with the former, though any result is possible.

Diagram 39 (B) White is ready for ... b7-b5

Diagram 40 (B) Both sides have chances

14Nd5

14 Bd4 has also been played, but the exchange of dark-squared bishops is a definite achievement for Black. In O.Korneev-S.Tiviakov, Linares 1998, the great Accelerated Dragon specialist won a nice game after 14 ... Bxd4 15 Qxd4 Qc5 16 Qd3 Nf617 Kh1 Bd718 f4 aS 19 a4 (a committal decision; instead Tiviakov suggests 19 Bf3 or 19 Rbe1, though Black should be fine in both cases after 19... Bc6 or the immediate 19... a4!?) 19... Bc6 20 Nd5 Re8!? 21 Rbd1 Rad8 22 h4? (too ambitious; necessary was 22 Nxf6+ exf6 23 Bf3 with equalityTiviakov) 22 ...Bxd5 23 cxd5 Qb4 24 Bf3 Rc8 25 Rcl Rxcl 26 Rxcl e5! 27 dxe6 Rxe6 28 Kh2 d5! 29 exd5 Qxf4+ 30 g3 Re3 31 gxf4 Rxd3 32 Kg3 Nxd5 and the extra pawn proved decisive. 14... Qxd2 15 Bxd2 (Diagram 40)

White enjoys a typical space advantage, though the exchange of queens represents a minor achievement for Black. 1S... Kf8 16 BgS

This is the most ambitious move, hoping to exploit the advantage of the bishop pair. At the same time it carries an element of strategic risk, as Black is a step closer to his dream of a good knight vs. bad bishop endgame. Instead, 16 Bxg4 is the safest move, but 16... Bxg4 17 Bg5 f6 should be fine for Black, who intends ... Be6 (or ... Bd7) and ... f7-f5, opening the game for his bishop pair. 17 Bc3looks better, though Black should not encounter too many problems here either. 16... Bxds

16... Nf6 17 Nxf6 Bxf6 18 Bxf6 exf619 Rfd1 Ke7left White with a slight but permanent edge in M.Adams-B.Alterman, Elista Olympiad 1998. The text leads to more interesting positions

200

Mar6czy Bind: Gurgenidze System in which both sides have prospects to outplay the opponent. 17 exds 17 cxd5 gives White nothing after 17... Nf6 18 Bd3 Rc3 (ADP), while 17 Bxg4 is equally toothless after 17... Be6 or 17... Bxe4!? 18 Bxc8 (or 18 Rbe1 f5) 18 ... Bxb119 Bxb7 Rb8 20 Rxb1 Rxb7 with easy equality. 17 ... Nf6 (Diagram 41)

Diagram 41 (W)

Diagram 42 (W)

Is the e2-bishop strong or weak?

Should White allow ...a5-a4?

The ending is an interesting one. White hopes that his bishop pair will prove to be a longterm asset, but for the time being it is hard for him to open the position and there is always a risk of ending up in a nightmare bad bishop endgame. 18 Bd3 The encounter J.Nunn-E.Pigusov, Reykjavik 1999, was quite entertaining and instructive. Play continued 18 Rfe1 h6 19 Bh4 Rc7 20 f3 g5 21 Bf2 b5! 22 Rbd1 (or 22 cxb5 Nxd5, aiming for c3) 22 ... b4 23 c5 (otherwise Black continues ... Nd7-c5 and ... a5-a4 with excellent prospects) 23 ... Rxc5!? (23 ... dxc5 24 d6 exd6 25 Rxd6 ReS also looks fine) 24 Bxc5 dxc5 25 d6 e6 26 Bb5 Rd8 27 Kfl Nh5, which gave Black decent compensation for the exchange and the game was eventually drawn. 18... h619 Bd2 19 Be3 Ng4 20 Rfe1 Nxe3 21 Rxe3 Rc7 was completely equal in B.Nikel-J.Jordan, Internet 2003, but White should have retained his bishop with 20 Bd2. 19 ... Nd7 20 Rfe1 as! (Diagram 42) This presents White with a difficult dilemma: does he allow Black to become active with ... a5-a4, or does he play a2-a4 himself, preventing immediate counterplay but permanently immobilizing his own queenside pawns? 21 Rbc1?! All things considered, the lesser evil looks to be 21 a4, preventing the opening of the a-file and hoping for an eventual breakthrough on the opposite flank. At the same time White would have to take care not to end up in a strategically unfavourable position, especially if

201

Starting Out: The Accelerated Dragon Black were ever able to exchange the dark-squared bishops. A possible continuation might be 2l...Nc5 22 Bc2 h5, setting up a strong defensive pawn chain while freeing up the g7bishop from defensive duties. The resulting position holds chances for both sides. 21 ... Rc7

Malakhov probably rejected the immediate 2l...a4 on account of 22 b4, hence this preparatory move. 22 f4?!

Missing the final opportunity to stabilize the queenside with 22 a4!. 22 ... Bd4+

This works out very nicely here, though it is worth mentioning that in similar positions it can be useful to consider the manoeuvre 22 ... Bb2, to be followed by ... Ba3-b4 attempting to exchange the dark-squared bishops. 23 Kf1 a4! 24 b4

A difficult decision. The text leaves the c4-pawn weak, while alternatives would grant

Black free reign over the a-file. Rowson must have anticipated this position, but probably underestimated the strength of Black's counterplay along the c-file. 24... Rac8 25 Re4

Postny calls this the decisive mistake, instead suggesting 25 Rc2, although it seems to me that 25 ... Nb6 26 Reel a3! will be hard to meet. 2S ... Bb2 26 Rc2 a3 (Diagram 43)

Diagram 43 (W} Problems for White

Diagram 44 (W} A strategic triumph for Black

27 Bc3

27 Be3? Nf6 traps the rook on e4, and after 28 Bb6 Nxe4 29 Bxc7 Nc3 Black wins the crucial a-pawn and with it the game (Postny). 27 ... Nb6! 28 Bxb2 axb2 29 Rxb2 Nxc4 30 Rf2

30 Bxc4 Rxc4leads to a rook ending in which White's numerous pawn weaknesses should prove decisive.

202

Mar6czy Bind: Gurgenidze System 30... Nb6 31 Rd4 Rc1+ 32 Ke2 Ra81 33 Ke3 Ra3 (Diagram 44)

Black's pieces are completely dominant, and it is only a matter of time before the positional advantage is converted to a material one. 34 g4 Nc4+ 35 Ke2

35 Kf3 bS leaves White completely tied up. 35 ... Rxa2+ 36 Kf3 Ra3 37 Kg2 Ne3+ 0-1 (Diagram 45)

Diagram 45 (W}

Diagram 46 (B)

Game over

The main line

White resigned on account of 38 Kf3 Nc2! leading to huge material gains. This was a model performance from one of the great modern exponents of the Accelerated Dragon. We will now turn our attention to the more active development of the bishop on g5.

9 BgS -the Main Line 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7 9 Bg5 (Diagram 46}

Compared to 9 Be3, the text has both positive and negative features. On gS the bishop is somewhat more active, and in certain situations it can be useful for White to have the option of Bxf6 followed by NdS. In some endgames the bishop may also help to create awkward pressure against the e7-pawn, especially if thee-file is opened after an exchange of knights on dS. The drawback is that the bishop is slightly more exposed, while in the event of short castling by White, Black can sometimes obtain tactical chances connected with the a7-gl diagonal. In some cases, most notably the endgame variation examined later, the two moves end up reaching to the same positions, though along the way there are various subtle points in favour of one move over the other. Nunn and Khalifman have both recommended 9 BgS in their respective repertoire books, while Tiviakov has argued in favour of 9 Be3.

203


TIP: If the reader is trying to decide which option to play with White, then the single biggest determining factor may be his assessment of the 14...Qd8 variation on page 209. If he is happy to allow that possibility then the generally more active 9 Bg5 should be preferred, while if he wishes to simplify his preparation by avoiding it then 9 Be3 would be the way togo.

9...0-0 9... h6?! is a bad idea. After the simple retreat 10 Be3 0-0 11 Qd2 Black is forced to waste time with 1l...Kh7, thus leaving White virtually a full tempo up on normal lines. 12 0-0!? deserves particular attention as the aggressive f4-f5 plan looks even more attractive after Black has weakened his kingside. 10Qd2 This is the standard move, although 10 Qe3!? is also possible, which has the advantage of lending extra support to the e4-pawn. After the natural continuation 10... Be6 11 0-0 Qa5 12 Racl Rfc813 b3 a614 f4 Qc515 Qxc5 Rxc516 Bf3 White is quite active, so Black should prefer 11...Qb6! (Diagram 47). Note the obvious parallel with Diagram 25.

Diagram 47 (W) Doubled b-pawns? No problem!

Diagram 48 (W) A trick to watch out for

Now 12 Qxb6?! axb6 is inadvisable for White, just as in the analogous variations on page 192. For example, 13 Racl Rfc8 14 b3 b5! is already very difficult for White. Instead, he should prefer the safer 12 b3, but Black should be fine in the endgame after 12 ... Qxe3 13 Bxe3 Nd7. S.Dolmatov-S.Tiviakov, Rostov 1993, continued 14 Racl Rfc8! (14 ... Nc5 would be met by 15 Nd5) 15 Nd5 Kf8 16 f4 aS! with good prospects on the queenside.

NOTE: In this game Tiviakov, a leading specialist in this opening, deliberately delayed the typical move ... Ncs, in order to reserve the option of the strong manoeuvre ... Bb2! intending ... Ba3-c5 to exchange the dark-squared bishops. 10... Be6

204

Mar6czy Bind: Gurgenidze System 10... a6 and 10...Qa5 have also been played, but Black has little to gain from delaying the essential bishop development. At this point White has can decide between two major strategies: castling immediately and battling in the middlegame, or delaying castling and offering a queen exchange.

Option 1 - White Castles 11 o-oQas There is an obvious comparison to be made with the 9 Be3 0-0 10 Qd2 Be6 11 0-0 Qa5 variation examined earlier. The presence of the bishop on g5 is more conducive to a plan involving kingside activity, whereas on e3 the bishop would exert more influence over the opposite flank. For this reason it makes most sense for White to aim for f4-f5 as in lines 'b' and 'c' below. Here are some possible developments: a) 12 f3?! Bxc4! (Diagram 48) provides a good illustration of the kind of tactic White must watch out for with his bishop on g5. After 13 Bxc4 Qc5+ 14 Kh1 Qxc4 15 Bxf6 Bxf6 16 Nd5 White had good chances to equalize in L.Nikolaev-Y.Dumansky, Alushta 2003, but this is hardly what he should be aiming for in the opening. b) The safest way to meet 12 Rad1 is with 12 ... a6!, when 13 f4leads nowhere after 13... b5! (but not 13...Qc5+ 14 Kh1 Bxc4?! 15 e5! with a dangerous initiative) 14 e5 b4 15 Nd5 Bxd5 16 exf6 exf6 etc. So White has nothing better than simplifying with 13 Bxf6 (13 b3 b5 14 Bxf6 only leads to a dead equal ending: 14 ... Bxf615 Nd5 Qxd2 16 Nxf6+ exf617 Rxd2 bxc4 18 Bxc4 Bxc419 bxc4 Rfc8, as seen in several games) 13 ... Bxf614 NdS Qxd2 (14 ... Qxa2? 15 Nxf6+ ex£616 Qc3! Qa4 17 Rxd6 was very bad for Black in S.Tiviakov-T.Polak, Oakham 1992) 15 Nxf6+ Kg7 16 NhS+ gxhS 17 Rxd2 h4 (Diagram 49).

Diagram 49 (W)

Diagram so (W)

The endgame is close to equal

Achieving the standard counterplay

Black is in no real danger here. White can claim a symbolic advantage due to the fractured kingside pawns, but the h4-pawn is hard to attack and Black can easily keep White distracted with his intended counterplay with ... b7-b5. The game F.Nijboer-S.Tiviakov, Wijk

205

Starting Out: The Accelerated Dragon aan Zee 2000, continued 18 b3 Rac8 19 f4 f5 20 exf5 Bxf5 21 Re1 Rc7, at which point White could find nothing better than 22 g3 exchanging off Black's weakness. There followed 22 ...hxg3 23 hxg3 Rf6 24 Kf2 Bg6 25 Red1 Rc5 26 Rd5 b6 27 Bf3 a5 28 Re1 Kf8 29 Be4 Bxe4 30 Rxe4 Kf7 31 Kf3 Rxd5 32 cxd5 Rf5 33 Rd4 Rh5 and a draw was agreed. c) On 12 Racl Black should once again play 12 ...a6113 b3 (13 Rfd1 Rfc814 Bxf6 Bxf615 Nd5 is harmless; the endgame arising after 15 ...Qxd2 16 Nxf6+ Kg717 Nh5+ gxh518 Rxd2 h4 should be compared with Nijboer-Tiviakov above) 13 ... bs! (Diagram 50). This typical break assures Black of a satisfactory game. After the ambitious 14 cxbs axb5 15 Bxb5 Rfc8! Black is very active and White will hardly be able to maintain his material advantage, e.g. 16 a4 Bxb3 or 16 Bd3?! Rxc3! 17 Rxc3 Nxe418 Bxe4 Bxc3 etc. Instead, 14 NdS is a natural reaction, with the typical sequence 14... Qxd2 15 Nxe7+ Kh8 16 Bxd2 Nxe417 Ba5. At this point most games have proceeded with one of the obvious rook moves to e8, but the simplest path to equality may be 17... bxc4 18 Bxc4 (].Fraga CastroX.March Moria, Oropesa del Mar 2001) 18 ... Rae8! 19 Nd5 Bxd5 20 Bxd5 Re5 21 Bxe4 (or 21 Rcd1 Nf6) 2l...Rxa5 with total equality, or alternatively 19 Nc6 Bxc4 20 Rxc4 d5 or 20 bxc4 ReB when White has no trace of an advantage.

Theoretical? Much the same story as before: neither side needs to know a lot, though a bit of specific knowledge is bound to be helpful.

Statistics 11 0-0 scores a total of 59% for White. After 1l...Qa5 12 Rad1 the endgame reached in Nijboer-Tiviakov has yielded an amazing 83% for White, but I believe this to be a statistical anomaly caused by a small sample size and some poor defensive play by Black. Despite the score I do not believe that Black has any reason to fear this line, unless he is playing to win at all costs. White has made a total of 59% with 12 Racl, a figure which falls to 54% after the most accurate response 12... a6!. Finally we move on to what is probably the sternest test of the Gurgenidze System.

Option 2 - the Endgame Variation 11 Rcl (Diagram 51) This is White's most popular choice here.

TIP: A strong case can be made for White delaying castling in this variation. A subsequent ...Qas may lead, after a timely Na4 or NdS, to a queen exchange on d2, after which Kxd2 will leave the white king well centralized. ll...Qas This standard move is probably the best, although Black does possess a tricky alternative in 1l...a6!? intending to force through a quick ...b7-b5. For instance, after 12 0-0 there is 12...b5! 13 cxb5 axb514 Bxb5 Bxa2, while 12 f3 is also met by 12 ... b5!, e.g. 13 cxb5 axb514 a3 Nd7! intending ... Nc5 with strong counterplay. Instead White should prefer the solid move 12 b3! and then: 206

Mar6czy Bind: Gurgenidze System a) In G.Serper-J.Donaldson, Kissimmee 1997, Black tried the tricky but probably unsound 12... b5?! which was refuted by White's accurate play: 13 cxb5 axb5 14 Bxb5 Qa515 Bc6! (15 0-0?! Rfc8) 15 ... Ra6 16 Na4! Rc8 17 Qxa5 Rxa5 18 Bd2 and White went on to win.

Diagram 51 (B)

Diagram 52 {W)

Delaying castling is a serious option

Preparing ... b7-b5

b) Possibly more challenging is the preparatory 12... Rc8 (Diagram 52), when White should play 13 0-0! (13 f3 allows Black to stir up complications with 13 ... b5! 14 cxb5 axb5 15 Nxb5 Rxcl+ 16 Qxcl Qa5+ 17 Qd2 Ra8, though even here White may be able to stay on top with 18 Nc3!- Nunn), and if 13 ... b5 14 cxb5 axb5 15 Bxb5 Qa5 16 Nd5! (Diagram 53) [but not 16 Bd3? Nxe4!]16 ... Qxb5 (16 ... Qxd217 Nxf6+ Bxf618 Bxd2 Bb2 19 Rb1 Rc2 20 Bh6 Bg7 21 Bxg7 Kxg7 22 a4 saw White go on to convert his advantage in V.Kotronias-H.Banikas, Kavala 1997) 17 Nxe7+ Kh818 Bxf6 Bxf619 Nxc8 Rxc8 20 Rxc8+ Bxc8 21 Qxd6 Kg7 22 Qd2, reaching an endgame in which White's rook and three pawns are clearly stronger than Black's two bishops, as pointed out by Khalifman.

0

NOTE: It is clear that Black would ideally prefer to bring his king's rook to c8. Therefore the text is highly committal; Black needs to force through ... b7-b5 quickly. If he cannot achieve this then the whole plan with ... a7-a6 and ... Rac8 must be considered faulty.

To summarize, 1l...a6!? is rather tricky and could easily prove effective against an unprepared opponent, but it is probably not quite sound if White plays accurately. 12 f3!

The strongest plan, keeping the king in the centre for the time being and angling for a queen exchange. 12 0-0 a6! transposes to line 'c' in the previous section - 11 0-0 Qa5 12 Racl a6!.

12 ... Rfc8 13 b3 (Diagram 54)

The standard move, defending c4. 13 Nd5 is not dangerous after 13 ... Qxd2+ 14 Kxd2 Nxd5 15 cxd5 (15 exd5 Bd7 is nothing) when Black can choose between 15 ... f6 and 15 ... Bd7!?, as in the line with 14 Nd5 below.

207

Starting Out: The Accelerated Dragon After 13 b3 White is ready to offer a queen exchange. This typically occurs after ... 13 ... a6

Diagram 53 (B) Forcing a better endgame

Diagram 54 (B) Delaying castling

This is the normal move, although it does leave a hole on b6, and Black occasionally tries a different move in order to try and reach a more favourable version of the endgame. These alternatives can be found in the notes to Game 26. 14 Na4! (Diagram 55)

Diagram 55 (B) Aiming for a better endgame

Diagram 56 (W) Offering to gambit a pawn

This is the point of White's play; he hopes that his centralized king will prove to be an asset after the likely exchange on d2. Before continuing any further, it is worth considering why White usually prefers the text

208

Mar6czy Bind: Gurgenidze System over the more natural-looking alternative ... 14 Nd5 Qxd2+ 15 Kxd2 Nxd5 Naturally Black should not cede the bishop pair with 15 ... Bxd5?! 16 cxd5, when White has a slight but stable advantage, with one possible plan involving Rhd1 (to maintain contact between the rooks) followed by g2-g3 and a bishop transfer to h3. 16 cxd5 Bd71? (Diagram 56} This is an interesting pawn sacrifice. Otherwise Black can avoid the whole issue with 16... f6 which should lead to near equal play. After 16... Bd7!? 17 Rxc8+ Rxc8 18 Bxe7 (White is not forced to accept the offer, but alternatives are unlikely to yield any advantage) 18 ... Bh6+ 19 Ke1 Black is very active, though it is not completely clear how he should proceed. 19 ...Rc2 and 19... Rcl+ have been played, while Wells' suggestion of 19... f5!? is also quite logical, preparing to open things up for the bishops without committing the rook just yet. We will now consider Black's options after the more challenging move 14 Na4!.

An Unusual and Sometimes Underestimated Possibility Given that the whole point of White's delayed castling was to be able to offer an advantageous queen exchange, it would certainly seem desirable for Black to cut across that plan with 14...Qd8!?.

NOTE: It may look like an inefficient use of tempi to move the queen back to its original square, but in the meantime the king's rook has been able to slide across to c8, so the queen moves have not been wasted. It is worth noting that in his DVD presentation Tiviakov asserts that 9 Bg5 is less accurate

than 9 Be3 specifically because of this queen retreat. His point is that in the analogous position with the bishop on e3, 14...Qd8?? would be a blunder due to 15 Nb6 winning the exchange. It seems to me that White can still obtain an edge with the bishop on g5, but I suppose one could argue that it would be simpler to prevent the possibility altogether. In this section I will mention some possibilities for both sides to enable the reader to decide for himself whether he wishes to play this variation with either colour. Firstly we will see how a popular and frequently recommended move can lead to serious problems for White. The drawback of the queen retreat has traditionally been considered to be... 15 cs?t ... which looks both logical and strong for White. The immediate capture with 15... dxc5?! is certainly no good after 16 Qxd8+ Rxd817 Nb6 Ra718 Rxc5, or 17... Rab818 Bf4 winning an exchange. Khalifman also mentions 15 ... Rc6, adding that White can remain on top after 16 Kf2!, e.g. 16... Qf817 Qb4! (M.Krasenkow-R.Hernandez, Palma de Mallorca 1989). This is all well and good, but the sting in the tail is found after the excellent response: 15 ... Nd7! (Diagram 57) I discovered the strength of this move through a combination of a database search and computer analysis, before noticing that Nielsen and Hansen had recommended the same course of action way back in 1998! Black really is threatening to take on c5 now, so if White is hoping for any kind of advantage he must try 16 cxd6 (in truth it is probably time to bail out with 16 Be3! dxc517 Nxc5 209

Starting Out: The Accelerated Dragon Nxc5 18 Qxd8+ Rxd8 19 Rxc5 with equality), but after 16... Rxc1+ 17 Qxc1 ReS Black intends to follow up with ... b7-b5, when the a4-knight will be short of squares. 18 Qb1looks forced (the alternatives are unacceptable for White, e.g. 18 Qd2? b519 dxe7 Qe8 20 Nb2?? Bc3, or 18 Qa3? b5 19 dxe7 Qe8 20 Nb2 b4, or 18 Qe3? b5 19 dxe7 Qa5+ all lead to the loss of the knight), when Black should continue his energetic play with 18 ... b5! (Diagram 58).

Diagram 57 (W}

Diagram 58 (W}

The refutation of 15 cs - ?

A difficult position for White

I found two games which reached this position. In G.Medda-V.Conti, Correspondence 1982, the continuation was 19 dxe7?! Qc7 20 Bd2? (this looks very limp, though 20 Nb2

Qc3+ wins the knight anyway) 20 ...bxa4 21 Bxa6 ReS 22 bxa4 Rxe7 and Black soon won. In G. Von BUlow -P.Marxen, German League 2003, White tried the slightly better 19 Nb2 Qa5+ 20 Bd2 Bc3 (also possible is 20 ...Qa3!? 21 Nd3 Qxd6 22 0-0 Ne5 with the initiative) 21 Qd3 Bxd2+ 22 Qxd2 Qxa2! (Black correctly judges that he has enough time to gobble the white queenside before returning for the e7-pawn) 23 dxe7 Qxb3 24 f4, and now the straightforward 24 ... Qa3 would have been best, intending to take on e7. Note that 25 f5 is nothing to worry about after 25 ... gxf5, e.g. 26 Qg5+ Kh8 27 exf5 Qxb2 28 fxe6 fxe6, or 26 exf5 Bxf5 27 0-0 Be6 when there is no perpetual and White does not have sufficient forces remaining to mount an effective attack.

How should White react to 14... Qd8 - ? I have not been able to find a convincing antidote to 15 c5 Nd7! -if anything White is struggling to equalize. Therefore I can only conclude that his frequently recommended 15th move is in fact dubious. So what should White do instead? The obvious alternative would be to catch up on development with... 15 o-o! (Diagram 59) Now that Black has avoided the queen exchange, there is little reason for White to postpone castling any longer. At this point Black's most natural plan is to prepare ...b7-b5. He is helped by the absence of the white knight from c3, as well as the open a7-gl diagonal which may at some point enable him to count on the useful resource of a queen check on b6.

210

Mar6czy Bind: Gurgenidze System 1S ... Bd7!? This may appear slightly strange, but the reason for it can be found in line 'b' below. Here are the alternatives:

Diagram 59 (B)

Diagram 60 (B)

The correct response

An improved pawn break

a) Despite the above comment about preparing ...b7-b5, Black can also switch plans with 15 ... Nd7, when 16 Kh1 Nc517 Nc3 a518 NdS Bxd519 exdS reaches a position very close to those found in the previous chapter. Following 19... Bf6! 20 Be3! (after 20 Bxf6?! exf6 Black's superior minor piece is more important than his pawn weakness) White was slightly better in Liang Jinrong-Wang Zili, Beijing 1996, though there was certainly no need for Black to react with 20 ...e6?, opening the game for the white bishops and leading Black to a crushing defeat. b) The most natural way of preparing ... b7-b5 is with 15... Rab8, which is certainly the move Black would like to play. Unfortunately it allows White to execute a much improved version of a familiar idea with 16 c5! (Diagram 60). Now this really is a strong move! V.Yemelin-S.Tiviakov, Krasnoyarsk 2003, continued 16... Rc617 Rfd1 Qf818 Bf4 Nh519 Be3 Rd8 20 Nb6 Nf6, and while Black eventually managed to draw, he was definitely under pressure and in this instance I would not advise the reader to follow the expert example. 16 Nc3 bs 17 Rfd11 (Diagram 61) 17 NdS Nxd518 cxdS Kf819 Be3 was a shade better for White in F.Nijboer-S.Tiviakov, Dutch Ch. 2001, due to Black's slightly loose queenside, but the text also maintains a pleasant advantage for the first player. Black has a few possibilities here, but none of them appear sufficient to equalize. a) 17 ...bxc4?! can be strongly met by 18 eS!. b) 17... b4 18 Bxf6! followed by NdS gave White an edge in D.Marciano-N.Spiridonov, French Team Ch. 1996. c) 17... Be8 18 e5!? dxe519 Qxd8 Rxd8 20 Rxd8 Rxd8 21 cxbS axb5 22 Nxb5led to an unbalanced endgame in A.Beliavsky-A.Shabalov, Manila Olympiad 1992. I would tend to prefer

211

Starting Out: The Accelerated Dragon White slightly at this point, though Black has some chances to activate his pieces and in the end Shabalov managed to win a fascinating struggle. The more solid 18 Qe3! looks like a good way of maintaining a stable advantage. Black may have achieved ...b7-b5 but his pieces are less active than usual, especially the queen's bishop. The natural move 18 ... Qa5 can be well met by 19 Bxf6 Bxf6 20 Nd5 Qxa2 21 Nxf6+ exf6 22 Rxd6 with a clear advantage.

Diagram 61 (B)

Diagram 62 (B)

White has a pleasant advantage

A key Gurgenidze position

Summing up, it seems to me that 14...Qd8!? is probably insufficient for equality, but for a different reason than that offered by most sources. In my view the immediate break with the c-pawn deserves a dubious mark, whereas the simple 15 0-0! leaves White with a comfortable position.

Theoretical? There are a few important details which both sides should know. Black must certainly be familiar with some lines after 15 c5? Nd7!, while White needs to show a certain amount of precision after 15 0-0! in order to maintain his advantage. Then again the reader may simply opt for 9 Be3 and bypass this line altogether.

Statistics 14... Qd8 scores a total of 26% for Black, but this does not tell the whole story. I have only been able to track down two games with 15 c5? Nd7!, with Black scoring one win and one draw. In the numerous games in which Black has played a different 15th move, his score has been pitiful. Finally, after the strongest move 15 0-0! White has scored 71%.

The Endgame The sequence of moves so far has been 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7 9 Bg5 o-o 10 Qd2 Be6 11 Rc1 Qa5 12 f3! Rfc8 13 b3 a6 14 Na4! and now, following the normal recapture 14...Qxd2+ 15 Kxd2 (Diagram 62}, an important

212

Mar6czy Bind: Gurgenidze System position is reached. The evaluation of this semi-endgame is central in determining the validity of Black's entire scheme of development. The queen exchange carries certain benefits to both players. We already know that the absence of queens will tend to ease Black's defensive task in the Maroczy Bind. At the same time, there are some specific advantages to White in this particular position, the main one being the excellent placement of his king.

0

NOTE: Just another reminder that Diagram 62 could equally occur with the white bishop on e3 instead of gS. The general strategy for both sides will be exactly the same, and there is quite often a direct transposition.

Strategy White's usual plan in these positions involves the advance of his kingside pawns. This strategy is seldom seen in the Maroczy Bind and may seem like an odd way to follow a queen exchange. The paradoxical truth is that the queen exchange has helped to facilitate this plan by reducing the potential danger of a possible black counter-strike with ... b7-b5. Furthermore, the white king should not be in too much danger in the centre, whereas with queens on the board any queenside counterplay would be much more venomous. From Black's point of view, his position is very solid, and while White is preparing a kingside offensive, Black can look to organize his own counterplay. Usually this will involve the pawn break .. .f7-f5, fighting for a share of the central and kingside territory. Following a possible pawn exchange with e4xf5, ...g6xf5 he may look to open the centre with ... d6-d5, hoping to expose the white king on d2. Finally, the queenside break with ...b7-b5 may still occasionally prove useful. These possibilities will be explored in detail in the following illustrative games and accompanying annotations.

Theoretical? By the standards of the Maroczy Bind the present variation is one of the most theoretically intensive. Both sides should possess a reasonably thorough knowledge of the different plans and move orders, although the total theoretical baggage is small compared to many other Sicilian variations.

Statistics After 14 Na4 Qxd2+ 15 Kxd2 White scores 59%.


0

A.Karpov • L.Kavalek

Nice 1974

Some readers may be familiar with this game (it is rather well known, being widely quoted in other works, most notably Nunn's Beating the Sicilian trilogy) and I apologize if this ap-

213

Starting Out: The Accelerated Dragon plies to you. The reason why I felt I had to include it was that it remains one of the best and most instructive games ever to have been played in this variation, and something which no prospective player for either side should miss out on. 1 c4 cs 2 Nf3 g6 3 d4 cxd4 4 Nxd4 Nc6 5 e4

The Maroczy Bind can occur via several different move orders; most notably, as here, from the English Opening. s... Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7 9 BgS 0-0 10 Qd2 Be6 11 Rc1 QaS 12 b3 Rfc8 13 f3 a6

So far we have been following the main line as described earlier. At this point it is worth examining a few of Black's attempted improvements. a) With 13 ... Rab8!? (Diagram 63)

Diagram 63 (W)

Diagram 64 (B)

A refined move order

The h4-pawn is a help to White

Black avoids weakening b6 for the moment and reasons that, when he finally does play ... a7-a6, the extra support provided by the rook will greatly improve his chances of executing the freeing ... b7-b5 thrust. Let us examine a couple of approaches for White: a1) Khalifman recommends a change of plan with 14 0-0, citing the game A.BeliavskyM.Cebalo, Slovenian Team Ch. 1998, which continued 14 ... a615 £4!? b5 16 f5 b4 17 fxe6 bxc3 18 ex£7+ K£8 (or 18... Kxf719 Rxc3) 19 Qe3 Qxa2 20 Rxc3 when White had a great position and went on to win convincingly. However, Black would be far better advised to play 15 ... Rc5! preventing the further advance of the £-pawn and leading to a tense middlegame with chances for both sides. a2) It seems to me that 14 h4! would be a better, not to mention more thematic, choice for White. We will see that this pawn move often plays a part in White's strategy in the queenless positions, and it can prove useful in other ways. Following the natural 14... a6 (14 ... h5 15 Nd5 was very nice for White in S.Vega Gutierrez-J.Gilbert, World Junior Ch. 2003) 15 Nd5! (15 Na4? would now be pointless in view of 15 ... Qxd2+ 16 Kxd2 b5) 15 ...Qxd2+ 16 Kxd2 (Diagram 64), White enjoys a considerably improved version of the 14 Nd5 line which was earlier considered harmless. The point is that Black can no longer play

214

Mar6czy Bind: Gurgenidze System 16... Nxd5? because of 17 exd5 Bd718 Bxe7!, as the reply 18... Bh6+ can safely be met by 19 Bg5, making full use of the pawn on h4. b) 13... Kf8!? is a different type of waiting move, providing defence of e7 which could prove useful in the event of a subsequent Nd5, as well as in the endgame generally. Again White has two main approaches. b1) 14 0-0 a615 £4 b516 f5looks murky after 16... b417 fxe6 bxc3 when, compared with the Beliavsky-Cebalo game, White is unable to take on f7 with check. However, 16 e5!? may be more promising, as after 16... dxe517 fxe5 the black king was looking a little uneasy on the £-file in A.Sulypa-M.Fiodorov, Swidnica 2000. b2) Nevertheless, 14 h4! is probably the strongest move, and is again likely to be useful in most circumstances. After 14 ... a6 (Black should begin counterplay, otherwise White may be tempted to forgo the queen exchange and play for a direct attack) 15 Na4! Qxd2+ 16 Kxd2 Nd7 (16 ... Rc6 17 Be3 was also pleasant for White in D.Goins-G.Tveit, Correspondence 2002) 17 h5 White is already making good progress on the kingside. Now the logical continuation 17... h618 Be3 g5 would transpose to the game L.Psakhis-E.Pigusov, Krasnoyarsk 1980, which continued 19 g3! (Diagram 65) 19... Rcb8 20 Nc3 b5 21 Nd5 Bxd5 22 cxd5 b4 23 Rc7 Nc5 24 Bxc5 dxc5 25 £4 with a clear advantage to White who went on to win convincingly.

Diagram 65 (B) Preparing to advance with f3-f4

Diagram 66 (B) White is slightly better

To summarize, both 13... Rab8!? and 13... Kf8!? are interesting and could set original problems to an unsuspecting opponent. However, in both cases White can maintain the advantage with 14 h4!, which is likely to lead to a slightly improved version of the endgame reached in the main line. 14 Na4 Qxd2+ 15 Kxd2 Rc6

Black needs to prevent the awkward knight invasion on b6. 15 ... Nd7 is the main alternative, and will be examined in the next game. 16 Nc3 (Diagram 66)

The knight returns to the centre, having done its duty on a4. 16... Rac8

215

Starting Out: The Accelerated Dragon The dubious 16...Re8?! has been suggested, intending 17 NdS Nd7? 18 Nxe7+ Rxe719 Bxe7 to trap the bishop with 19 .. .£6 (19 ...Bh6+ is no good after 20 Kc3 Bxc121 Rxcl f6 22 Rd1Khalifman). Unfortunately this idea can be refuted tactically by 20 Ke3 Kf7 21 Bd8 b6 22 c5! when there is no way to keep the bishop imprisoned. Black can also try 16... Rcc8!?, when White should execute the standard plan with 17 h4! (17 Na4 repeats), e.g. 17... Kf8 18 g4 with a slight advantage, as in Bu Xiangzhi-V.Malakhov, Ergun 2006. Note that in the analogous position with the white bishop on e3 the text would be pointless as the knight could still come to b6. 17 NdS KfS

After 17...Bxd5 White has a pleasant choice between 18 exd5 R6c7 (D.FleetwoodA.Tabernero Palacios, Correspondence 1995) and 18 cxd5 Rxc119 Rxcl Rxcl 20 Kxcl (T.Fogarasi-S.Deak, Zalakaros 1991), with a slight but stable advantage in both cases. 18 Be3!?

18 Nxf6 Bxf6 19 Bxf6 exf6 looks tempting, but it would not be so easy to exploit Black's pawn weaknesses. 1S ... Nd7 19 h4! Bxds

Black is also unable to solve his problems with 19... h5 20 Nf4, but 19 .. .£5!? was a possible alternative. After Khalifman's 20 Bg5 Black has the cheeky 20 ... h6!, when 21 Bxe7+ Kf7 leaves the bishop marooned on e7. Instead 20 h5!? fxe4 21 fxe4 may be better, when Black's kingside worries remain. 20exds

20 cxd5 also promises White an edge, but the text looks stronger here. The kingside advance is already underway, so it makes sense for White to keep the rooks on the board. 20... R6c7 21 hs (Diagram 67)

Diagram 67 (B)

Diagram 68 (B)

A thematic kingside advance

The kingside initiative is strong

21 ... Kg8

Nunn and Gallagher suggest 21...Re8!?, aiming for counterplay with ...e7-e6. It looks risky to open the game for White's bishop pair, but overall I would have to agree that an inferior

216

Mar6czy Bind: Gurgenidze System position with counterplay is preferable to an inferior position with no counterplay. 22f4!

An excellent move, making way for the bishop to come to g4. 22 ... Nc5 23 Bg4 Ne4+ 24 Kd3 fS It looks as though Black is becoming active, but the soon-to-be World Champion has seen further. As Nunn has commented, the clever point of Karpov's play is that he has been deliberately provoking Black into playing ... Ne4+ and ... f7-f5, in order to obtain the possibility of g2-g4 to open up the kingside, after which his unopposed light-squared bishop will turn into a monster.

2s Bf3 bs

The typical freeing move, but it does not solve Black's problems on this occasion. 26 g41 (Diagram 68) 26... bxc4+ 27 Rxc4

Karpov exchanges one pair of rooks to reduce Black's chances of obtaining meaningful counterplay. 27... Rxc4 28 bxc4 Ncs+ 29 Bxcsl

The opposite bishops are of little consequence here. 29 ... Rxcs

29 ... dxc5 30 h6 Bd4 31 Rb1 also leaves Black in serious difficulties. 30 h6! (Diagram 69)

Diagram 69 (B)

Diagram 70 (B)

Playing against the bishop

Is White's pressure enough to win?

30... Bf8

Burying the bishop is a depressing choice. Instead 30 ...Bf6 would have offered some chances to resist. Then 31 Rb1 allows 31...Ra5! with counterplay, while 31 g5 Bh8 32 Rb1 Kf7 33 Rb8 Ba1 34 Bd1 Ra5 35 Bb3 Bb2leaves Black very tied up, but it is not so easy to suggest a convincing plan for White. Perhaps the strongest continuation is 31 gxf5 gxf5 32 Rb1 Rc8 33 Rb6 (Diagram 70) 33 ...a5 (33 ...Ra8 34 cS! dxc5 35 d6 c4+ 36 Ke3 wins) 34 Rb5 Ra8 35 cS dxc5 36 d6 Rd8 37 BdS+ Kf8 38 d7! e5! 39 Be6 exf4 40 Bxf5 reaching a position in which the passed d-pawn

217

Starting Out: The Accelerated Dragon gives White good winning chances, though the game is still far from over. 31 Kc3

In a way this is a typical Karpovian move: before proceeding with his own plans he spends a tempo nullifying any potential counterplay with ... Ra5, which would now be met by Kb3 and the rook move has achieved nothing. The appeal of such an approach is self-evident, but technically speaking it would have been more accurate to solidify the cage of the f8bishop with 31 g5! Ra5 32 Rb1 (Diagram 71).

Diagram 71 (B)

Diagram 72 (B)

Burying the bishop

Black's last chance

Now 32 ... Rxa2loses after 33 c5! dxc5 34 d6. The most resilient defence is 32 ... Kf7!, but 33 Rb3 Rxa2 34 c5! dxc5 35 d6! still looks very difficult for Black, e.g. 35 ... c4+ (if 35 ... exd6? 36 Bd5+, or 35 ... Ke6 36 Rb8, or 35 ... e6 36 Rb7+ Kg8 37 d7) 36 Kxc4 e6 37 Rb7+ Kg8 38 d7 Rd2 39 Bc6 Be7 40 Rb8+ Kf7 41 Rh8 and wins. 31 ...fxg4

White's problem now is that the h6-pawn requires constant defence. The f8-bishop may still be imprisoned, but at least it is performing a useful function in keeping the white rook occupied. 32 Bxg4Kf7

Nunn and Gallagher criticize this move, but I do not believe it to be any worse than the alternatives. They suggest 32 ... Rc7 33 Be6+ Kh8 34 f5 Rb7, claiming that the pressure against h6 makes it hard for White to achieve anything. Note that 35 Bf7 is pointless in view of 35 ... Bxh6! 36 Rxh6 Kg7. A more promising plan is to prepare an assault against the a6pawn, particularly as the black bishop will find it hard to influence matters on the queenside. The immediate 35 Bc8 could be met by 35 ... Rb8 36 Bxa6 Ra8 37 Bb5 Rxa2, so perhaps 35 Rh2!? would be a good way to begin. If White can arrange to capture the a6-pawn at the expense of, say, the h6-pawn, then he would retain chances to win the game. 33 Be6+ Kf6 34 BgB Rc7

34 ... Bxh6? is no good in view of 35 Rxh6 Kg7 36 Rxh7+ Kxg8 37 Rxe7 and wins, as pointed out by Nunn and Gallagher. But now the pseudo-sacrifice is a genuine threat.

218

Mar6czy Bind: Gurgenidze System 35 Bxh7 e6 36 Bg8 exds 37 h7 (Diagram 72) 37 ... Bg7?

The decisive error. Black could probably still draw with 37 ... Rxc4+ 38 Kd3 Bg7 39 Bxd5 (39 h8Q achieves nothing after 39 ... Bxh8 40 Rxh8 Rc8!) 39 ... Rc5 40 Ke4 Bh8. White's position is still preferable, but the reduced material and opposite bishops would make the task of winning highly problematic. 38 Bxds

Now we reach essentially the same ending as in the previous note, except that White has an additional pawn on c4. This helps his cause enormously by stabilizing the bishop on d5, while shutting the black rook out of the game. A few more improving moves are all it takes for Karpov to wrap up the win. 38 ... Bh8 39 Kd3 KfS 40 Ke3 Re7+ 41 Kf3 as 42 a4 Rc7 43 Be4+ Kf6 44 Rh6 Rg7 45 Kg4 1-0 (Diagram 73)

Diagram 73 (B)

Zugzwang

Diagram 74 (W) Transpositions often occur here

Black is in zugzwang and will soon face further material losses. This was a fine positional masterclass by Karpov, notwithstanding the slight inaccuracy on his 31st. Game27

0

F.Nijboer •

S.Tiviakov

Dutch Championship, Leeuwarden 2003 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 Be2 Nxd4 8 Qxd4 Bg7 9 Be3

So far we have only looked at the endgame occurring after 9 Bg5, but it can occur in almost

identical form with the bishop on e3. 9 ... Be6 10 Rc1 o-o 11 Qd2 Qas 12 f3 Rfc8 13 b3 a6 14 Na4 Qxd2+ 15 Kxd2 Nd7

This is the main alternative to 15 ... Rc6. It is important to note that the same move is equally valid in the position with the bishop on g5. Let us consider for a moment the relevant position (Diagram 74}: WARNING: The e7-pawn is most definitely poisoned!

219

Starting Out: The Accelerated Dragon White must definitely not take on e7 here, as after 16 Bxe7? Bh6+ 17 Kd3 b5! (even better than the immediate capture on cl) Black is practically winning. N.Efentakis-M.Prevenios, Correspondence 2004, continued 18 Nb2 Ne5+ 19 Kc3 Bg7 20 Kc2 Nc6 21 Bxd6 Nd4+ and White resigned as he is losing a piece. Instead of this dreadful blunder White should proceed with 16 g4!, when 16.. .£617 Be3 f5 leads to a direct transposition to the present game except that both sides have played one additional move.

~

'VJ/

TIP: One of the main points of Black's ... Nd7 is the preparation of the freeing move ...f7-f5. Both sides should pay special attention to this possibility.

For example, the familiar move 16 h4?! would be inaccurate here, due to 16 .. .£617 Be3 f5! which has scored very highly for Black. After that brief diversion we will return to the game. I should mention that the alternative move orders may lead to some confusion over numbering, as the 9 Bg5 sequence reaches the same position after one additional move. To keep things simple I will henceforth treat all game references according to the move order of the present game. 16g4! Tournament practice has demonstrated this to be White's strongest continuation. 16... f5! (Diagram 75)

Diagram 75 {W)

Diagram 76 {W)

Fighting back

An important resource for Black

After witnessing Black's suffering in the previous game, the appeal of the text is obvious. Whether or not White has a theoretical advantage is a subject for debate (though he probably can claim some sort of edge), but here Black is at least fighting back rather than waiting passively. 17 exf5 It is very important to consider the alternative 17 gxf5!? gxf518 Rhgl. The idea of using the g-file is very natural, yet for a long time has been considered less challenging than the text. However, some recent games have questioned this assessment, and I would recommend

220

Mar6czy Bind: Gurgenidze System both sides to pay close attention to this possibility. Following the natural18 ... Kh8 White has two main options: a) After 19 Nc3 fxe4 it is interesting to try 20 fxe4!?, intending to plant the knight on d5 (20 Nxe4 is playable, but not too dangerous). A.Dubois-P.Reyes Molina, Correspondence 2003, continued 20 ... Be5?! 21 Nd5 Bxd5 22 exd5 with a large advantage to White. Instead, Black should have preferred 20 ... Bf7! intending to meet Nd5 with ...e7-e6, e.g. 21 Rcfl Rf8 22 Nd5 e6 23 Nb6 Nxb6 24 Bxb6 Bg6 with approximate equality. b) 19 Bd3 is a tricky move, after which it is very important for Black to continue with the accurate 19.. .f4! (Diagram 76). [In R.Vaganian-R.Ponomariov, Calvia Olympiad 2004, the young super-GM fell into serious trouble after 19... Ne5?! 20 Rxg7! Kxg7 21 Bd4 with excellent play for the exchange.] Following 19... f4!, 20 Bxf4 only leads to equality after 20 ...Rf8 21 Bg5 (or 21 Ke3 Rxf4! 22 Kxf4 Bd4! with dangerous threats) 21...Rxf3 22 Bxe7 Bh6+ 23 Bg5 Bxg5+ 24 Rxg5 Rf2+ as pointed out by Mikhalevski. 20 Bf2 keeps a bit more life in the position, but Black should still be ok thanks to his undisputed control over the e5-square. L.Dominguez-R.Felgaer, Dos Hermanas 2005, continued 20 ... Ne5 21 Ke2 Rc6 22 Nb6 Rf8 23 Nd5 Rf7 24 c5 Bh3 25 cxd6 exd6 26 Bd4 Ng6 27 Bf2 Ne5 28 Rc2 Rxc2+ 29 Bxc2 Nc6 30 Kd3 Ne5+ 31 Ke2 Nc6 with a draw by repetition. 11 ... gxfs 18 h31 (Diagram 77)

Diagram 77 (B) Maintaining the tension

Diagram 78 (B) A crushing blow

This is White's most challenging approach, reinforcing the g4-pawn and maintaining the tension on the kingside. He intends to increase the pressure with moves like f3-f4 and Bf3, combined with Nc3-d5 and perhaps a rook transfer to el. Black's principal plan involves the preparation of ... d6-d5, by which he hopes to demonstrate that the white king is not so ideally placed after all. The immediate advance 18 g5?! releases the tension prematurely and presents Black with the opportunity to strike with 18... d5! immediately, when he has chances to take over the initiative. S.Garcia Martinez-S.Tiviakov, Ubeda 1999, was a good example of how quickly things can turn pear-shaped for White after inaccurate play: 19 f4 dxc4 20 bxc4 Rd8

221

Starting Out: The Accelerated Dragon (20 ... Rc6 and ... Rac8 also looks promising) 21 Ke1 Rab8 22 Nb6? Nxb6 23 Bxb6 Rd7 24 a4 Bd4 25 Rb1 Rc8, when White had serious problems defending his weaknesses and the rook on h1 is a mere spectator; Black went on to win. White does better with 19 cxdS BxdS 20 Rhd1 (20 Nc3 can be met by 20 ... Bf7 followed by ...e7-e5 with some initiative- Tiviakov), which should be enough for equality; A.Rustemov-P.Tregubov, Moscow 2004, concluded 20 ... b5 21 Nb6 Nxb6 22 Bxb6 h6 23 h4 hxgS 24 hxgS BeS 25 Ke3 f4+ 26 Kf2 e6 1J2-1J2. 18 ... Rf8 Definitely the most logical move; the rook has little future on the c-file and clearly belongs back on f8. Now one of Black's main positional ideas involves the exchange of darksquared bishops with .. .f5xg4 followed by ... Be5-f4 and ... NeS. Alternatively, in certain situations he might opt for the advance .. .f5-f4, gaining space and securing the use of the eS-square for his minor pieces.

TIP: This type of pawn advance should not be played on a whim, as it robs the black position of much of its dynamism. For example, it would be pointless to play it just for the sake of attacking the bishop on e3. The bishop will move, and then what have you gained? Such a committal move should only be played when there is a genuinely good reason. Before moving on I will briefly mention that a few alternatives have been tried, such as 18... Rab8 and 18 ... Rc7!?, both of which should be met by the centralizing 19 Nc3. 19f4! This is the strongest way for White to prevent Black from carrying out the objectives outlined in the previous note. It is also worth investigating the consequences of 19 Nc3, which is almost always useful in these positions, though in this instance it is less precise than the text. One important point of 19 Nc3 is that the plan described above is no longer effective, as after 19.. .fxg4?! 20 hxg4 BeS 21 NdS! BxdS 22 cxdS the threats of Rc7 and Bd3 make it impossible for Black to exchange the dark-squared bishops. Fortunately, he has a much stronger option in the tactical break 19... b5!, e.g. 20 cxbS (20 NdS BxdS 21 cxdS Nf6 wins the dS-pawn) 20 ... axb5 21 BxbS f4! (on this occasion Black has a genuine reason for advancing this pawn, namely to shut the white bishop out of the game before making the following exchange) 22 Bf2 Bxc3+ 23 Kxc3 Rxa2 when Black has at least equal chances (analysis by Khalifman). Therefore White had better follow the game continuation if he hopes for an advantage. 19 ... Rad8 This is the most common move; Black is powering up for a ... d6-d5 break. The immediate 19... d5? is premature in view of 20 gxfS! (this is one reason why White was better off postponing the advance of the g-pawn) 20 ... Bxf5 (if 20 ... Rxf5?! 21 Bg4) 21 cxdS, when Black will have to waste some time regaining the d5-pawn. M.Makropoulou-T.Papadopoulou, Greek Ch., Athens 1999, continued 2l...Nf6 (if 2l...Be4?! 22 Rhg1 Bxd5 23 Bd4 Nf6 24 Nb6 Rad8 25 NxdS RxdS 26 Bc4 wins) 22 Nb6 Rad8 23 Bf3 NxdS 24 NxdS e6 25 Bb6 Rd7 26 Rc7 Rf£7 27 Rc8+ Rf8 28 Rc7 Rf£7 29 ReS Bf8 30 Rg1 + Bg6 31 Rc7 Rd6 32 Rx£7 Kx£7, and now White could have won with the beautiful move 33 f5! (Diagram 78) as pointed out by Khalifman. Black's other main option is 19 ... Nf6, attacking g4 while supporting the advance of the dpawn (though the response Na4-c5 must always be considered). After the sensible 20 Rhg1 Rad8 (20 ... d5? 21 NcS! was very good for White in N.Bojkovic-G.Heinatz, Dresden 1994) 21

222

Mar6czy Bind: Gurgenidze System Bb6! Rc8, the most accurate continuation seems to be 22 Bf3! when V.Kotronias-S.Tiviakov, Gibraltar 2003, continued 22 ...fxg4 23 hxg4 Nd7 24 f5! Bf7 25 g5! (Diagram 79) with a clear advantage to White, who went on to win. Evidently Tiviakov was unable to find a fully satisfactory improvement, hence his switch to 19... Rad8 in the present game.

Diagram 79 (B)

Diagram 80 (B)

White's kingside pawns dominate

An important choice for Black

20g5

Preventing the black knight from coming to f6. White has a reasonable alternative in 20 Nc3 d5 21 cxd5 Nf6 22 Rhgl!? (22 Bf3 fxg4 23 hxg4 Bxg4 is equal- Khalifman) 22 ... Nxd5 23 Nxd5 Bxd5 24 Kel with just a slight edge in V.Chuchelov-K.Van der Weide, Utrecht 2000. 20... Ncs

Now that the £6-square is unavailable, it makes sense for Black to activate his knight by a different channel. After the immediate 20 ...d5 the simple response 21 cxd5 Bxd5 22 Rhdl was pleasant for White in L.Nisipeanu-A.Khalifman, FIDE World Ch., Las Vegas 1999, as his pieces are very active and his king position is not a worry. 21 Nc3! (Diagram So)

This is the only way to fight for the advantage. The tame 21 Nxc5 dxc5+ 22 Kel Bd4 23 Kf2 was agreed drawn in A.Motylev-S.Tiviakov, Bastia (rapid) 2003. We now reach a critical position where Black is at a crossroads. 21... Ne4+?

This could have led to serious problems, and Black should definitely look to improve at this point. There are at least two alternatives worth considering: a) 21...Bxc3+!? 22 Kxc3 Ne4+ is quite a radical approach, ceding the bishop pair in order to secure the position of the knight on e4. While there is no denying the power of that piece, I do not find the idea convincing as White's two bishops have tremendous potential. Following 23 Kb4! it should not be too difficult for White to play around the knight and create threats on both sides of the board. It is also worth noting that the opposite bishop position arising from 23 ... Ng3 24 Rhel Nxe2 25 Rxe2 is quite depressing for Black after 25 ... Bd7 26 Bd4 Kf7 27 Reel Rfe8 28 h4- Khalifman.

223

Starting Out: The Accelerated Dragon b) Black's best chance may lie in 2l...Bd7!? {Diagram 81).

Diagram 81 (W}

Diagram 82 (B)

This may be Black's best bet

Black has serious problems

The bishop was not doing much on e6, so it is quite logical for Black to reroute it to c6. This move is an untested suggestion of Tsesarsky, who goes on to analyse 22 Bf3 Bc6 23 NdS Rd7 24 b4 Ne4+ 2S Bxe4 fxe4, or 24 Rhe1 e6 (24 ... Ne4+ also looks entirely reasonable) 2S BxcS exdS 26 Bb4 dxc4 27 Bxc6 bxc6 28 Rxc4 cS with equality in both cases. I would add that 22 Rhe1 Bc6 23 b4 Ne4+ 24 Nxe4 Bxe4 2S Bb6 Rd7 26 Bd3 dS! offers Black sufficient counterplay. It should also be noted that if White tries 22 b4 Ne4+ 23 Nxe4 fxe4 24 Bb6 (24 h4 can be met by 24 ... bS!?, freeing the way for the d-pawn) 24 ... Rc8 2S Ke3, hoping for a position similar to the following note, then he will be disappointed after 2S ...eS! 26 fxeS BxeS when his loose king position is a cause for concern. 22 Nxe4 fxe4 23 Bg4?!

Missing an opportunity to gain a serious advantage with 23 Bb6! {Diagram 82). White intends simply to play Ke3 when the e4-pawn will be seriously weak. This position was tested in A.Goloshchapov-A.Wirig, Izmir 2004, which continued 23 ... Rc8 24 Ke3 Rc6 2S BaS BfS 26 Bc3 (26 Bb4 eS 27 fxeS BxeS 28 Rc£1 is also clearly better for White- Tsesarsky) 26 ... Bxc3 27 Rxc3 bS 28 h4 b4 29 Rc2 ReS 30 Rd1 Rfc8 31 Rcd2 RaS 32 RdS RccS 33 RxcS dxcS 34 Rd2 e6 3S hS h6 36 Bfl and White converted his advantage. After the move played White is still better, but Black has more chances to equalize. 23 ... Bfs 24 Bb6 Rde8 25 Rhf1 es 26 Bxfs Rxfs 27 fxes Rfxes 28 h4 e3+! Surprisingly, all of the preceding moves had been seen before. The text is an improvement on D.Gross-O.Brendel, German League 2000, which concluded 28 ... h6 29 gxh6 Bxh6+ 30 Be3 RhS 31 Bxh6 Rxh6 32 Rf4 e3+ 33 Kd3 Rg6 34 Re1 Rg3 3S Re4 Rxe4 36 Kxe4 Rg4+ 37 KdS1-0. 29 Ke2 Re4!

Tiviakov is just about managing to squeeze enough activity out of his position before White can surround and win the e-pawn. 30 hs Now Black can equalize relatively comfortably, as White will have two weak kingside

224

Mar6czy Bind: Gurgenidze System pawns to worry about instead of just one. The superior 30 Rhl! would have caused Black a few more difficulties. Tsesarsky provides a detailed analysis from here, the main line runs 30 ... Bd4 31 Bxd4 Rxd4 32 Rcdl Rg4! 33 Rh3! Rg2+ 34 Kd3 e2 35 Rel bS 36 cxbS axbS 37 Re3 Rxe3+ 38 Kxe3, when Black's split queenside pawns mean that he is not completely safe yet. 30... Bd4 31 Bxd4 Rxd4 32 Rcd1 Rh4 33 Rh1 Rg4 34 Rhg1 Rh4 35 Rh1 Rg4 36 Rh2 RxgS 37 Rxd6 Rg1 38 Rd1 Rxd1 39 Kxd1 Kg7 40 h6+ Kg6 41 Ke2 aS 42 Rh3 Re6 43 cs ReS 44 Rxe3 Rxcs 45 Re6+ Kf7 46 Rb6 Rc2+ Yz-Yz

Summary and Conclusions It is easy to see why the Gurgenidze System remains a popular choice amongst Accelerated

Dragon players of all levels. Black develops quickly and actively and can, in most cases, arrange his pieces according to the standard recipe outlined in the opening pages of the chapter. Despite the predictability of his play, in many cases Black will succeed in achieving his desired ... b7-b5 break, which is usually enough to ensure equality and in some cases to fight for the initiative. White has a variety of ways of combating Black's plan. 7 f3!? is not so popular nowadays, and instead of following the long theoretical lines I have focused on the more modern plan involving 10... a5!? as in Game 24. After the more common 7 Be2 Nxd4 8 Qxd4 Bg7 White must make two major decisions. The first concerns the development of the dark-squared bishop which could be deployed on e3, g5, or perhaps postponed in favour of 9 0-0 followed by 10 Qe3!? or 10 Qd3!?. Neither of the last two should be terribly dangerous for Black, although the Gurgenidze player should certainly be familiar with them. In most games White does develop his queen's bishop on the ninth move, thus allowing him to meet 9 ...0-0 with the harmonious retreat of the queen to d2. Then after 10... Be6 White must decide whether he wishes to castle immediately, or leave his king in the centre. Castling is a natural and popular choice, after which White can choose between kingside aggression with f2-f4 and solidity involving a more patient build-up. We have seen that both options can lead to rich middlegames with chances for both sides. The final part of the chapter dealt with the endgame variation, which is considered by many experts to constitute the most stringent test of the Gurgenidze System. White delays castling and hopes that after a queen exchange his king will be ideally placed on d2. The game Karpov-Kavalek remains a highly instructive example of a positional squeeze from White. In general, my recommendation for most Gurgenidze players would be to follow the exchange of queens with 1S ... Nd7 followed by a subsequent ... f7-f5 as in Nijboer-Tiviakov. This will at least enable you to avoid the kind of passivity from which Kavalek suffered in Game 26. The resulting positions are unlike anything you are likely to encounter in most other openings, and both sides will need to demonstrate a good level of theoretical knowledge and positional understanding. If the reader has thoroughly acquainted himself with the annotations to Game 27, then he should have a good chance of outplaying the majority of opponents from either side of this strategically unbalanced position. From an ultimate theoretical viewpoint White's chances appear to be slightly preferable, though Tsesarsky's suggested improvement of 21... Bd7!? could challenge that verdict. It remains to be seen whether practical testing will reveal any effective methods for White to fight for the advantage here.

225

Chapter Eight

Mar6czy Bind: 7 ••• Ng4

R R R R R R R R

Basic Theory and Early Deviations The Provocative g ••• es!? The Ambitious Approach The Solid Approach The Main Line with 9 ••• Ne6 Conventional Play with ••• 0-0 Larsen's Creative ••• g6-g5 Plan Summary and Conclusions

Mar6czy Bind: 7 ... Ng4 This chapter will deal with the positions occurring after the opening moves: 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 Ng4!? (Diagram 1)

Diagram 1 (W}

Diagram 2 (W}

A novel way to trade knights

White is already worse

As is the case with most of his major systems in the Maroczy, Black is looking to exchange a pair of knights. This particular method is rather more visually striking than usual, as the knight on g4 is currently en prise. But as the reader will no doubt have noticed, the capture 8 Qxg4 (which is almost always played) allows 8 ... Nxd4 regaining the piece. After the standard 9 Qd1 Black can choose between 9...e5!?, attempting to maintain the knight's powerful position, and the more reliable retreat 9... Ne6, when the knight's unusual position in the centre can give rise to some interesting and unique possibilities.

Basic Theory and Early Deviations We will begin our coverage by looking at some unusual options for both sides. Beginning at move 8, there is only one decent option for White. 8Qxg4

Nothing else will do. 8 Nxc6? is weak, though it is worth taking a few moments to familiarize yourself with the following analysis, as this could easily be the reaction of an inexperienced player encountering 7... Ng4 for the first time. After 8... Nxe3 9 NxdS (9 fxe3?! allows Black to choose between 9 ... dxc6, 9 ...bxc6 or 9... Bxc3+!? 10 bxc3 dxc6, with a clear advantage in all cases) g... Nxd1 (Diagram 2) White faces an uninspiring choice: a) 10 Nxf7?? Nxc3 11 Nxh8 Bxh8 12 Bd3 Na4 is winning for Black. b) 10 Ne6?! has the idea of compromising Black's pawn structure, but after the straightforward 10... dxe6 11 Rxd1 Bxc3+ 12 bxc3 Bd7 Black is clearly better. c) 10 Nxd1 Kxd8 gives Black a favourable ending thanks to the powerful and unopposed 'Dragon Bishop'. d) 10 Rxd1 Kxd8 (10 ... Bxc3+!? 11 bxc3 Kxd8 is also possible) 11 c5 b6 12 c6 (12 cxb6 is advantageous for Black after 12... axb6 or 12 ... Bxc3+ 13 bxc3 axb6) 12... Bxc3+! 13 bxc3 d6!

227

Starting Out: The Accelerated Dragon (13 ... Kc714 cxd7 Bxd7 is also strong, but the text should win material) 14 Bc4 f6, and as long as Black is careful he should have little trouble in rounding up the c6-pawn after ... Kc7 and ... Rb8. 8... Nxd4

8 ... Bxd4?! (Diagram 3) is a risky line with a dubious reputation,

Diagram 3 (W)

Diagram 4 (B)

A risky line for Black

Black faces an important choice

though it is far from easy to refute and could perhaps be considered as an (occasional!) surprise weapon. Recapturing with the bishop is actually quite logical as an exchange of dark-squared bishops can often prove beneficial for Black in the Maroczy Bind (compare the Classical main lines in Chapter Six, as well as 5 ... Bh6!? in Chapter Ten). However, in this specific position the idea is under a dark cloud because of the strong riposte 9 Bxd4 Nxd4 10 0-0-0! (10 Qd1 e5 is not dangerous for Black) 10... e5 (Black should try to maintain his powerful knight) 11 Qg3 d612 f4 f613 f5! (White's long-term chances are hampered by his bad bishop, so he must play energetically to exploit his temporary initiative) 13... Kf7 (13 ...gxf5?! 14 Qg7 Rf8 15 Nd5 is dangerous for Black), at which point White can choose between 14 Ne2 and 14 NbS!?. Black is by no means losing by force, but he is certainly walking a tightrope and I would advise the reader to opt for something a little more solid. 9 Qd1 (Diagram 4) Once again this is practically the only move played. The (very rare) alternatives include: a) 9 0-0-0?! Nc6 will leave the white king vulnerable to an attack. b) 9 Bd3 d510 Qd1 dxe411 Bxe4 0-0 12 0-0 has been tried a few times, when Black's most straightforward response looks like 12... Bf5!, after which 13 Bxf5 (13 Bxb7? Rb8 regains the pawn and gains the advantage) 13 ... Nxf5leaves Black very comfortably placed. After 9 Qd1 Black can choose between the strategically risky but unrefuted 9 ... es!? and the sounder main line of 9... Ne6, both of which will be discussed separately. (The alternative retreat 9 ... Nc6?! would not be very logical, as the knight would have fewer active prospects on this square.)

228

Mar6czy Bind: 7 ... Ng4

0

NOTE: The positioning of the black knight on e6 is not often seen in the Maroczy Bind, but in the 7... Ng4 variation it can provide some unique opportunities for Black's counterplay.

Theoretical? Play can become quite sharp after 8 ... Bxd4, so a degree of theoretical knowledge will be useful for both players, especially Black who needs to take care to avoid being blown off the board! White also needs to show some early precision in order to generate a meaningful initiative with moves like 10 0-0-0! and 13 f5!.

Statistics Black scores 45% after 8...Bxd4, but this drops to a pitiful 17% if White gets as far as 13 f5!. You have been warned!

The Provocative 9... es!? We will now move onto another tricky continuation. Unlike the 8... Bxd4?! variation, this one has the enormous advantage of not leaving Black vulnerable to an early mating attack, though it does carry an element of strategic risk. 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 Ng4!? 8 Qxg4 Nxd4 9 Qd1 eSI? (Diagram 5)

Diagram 5 (W)

Diagram 6 (B)

A patzer's move? Not quite!

An ambitious, if impractical choice

On first impressions this looks almost like a patzer's move! Black saddles himself with a

backward d-pawn, blocks his prized 'Dragon bishop' and presents White with a permanent outpost on d5. However, 9 ... e5 does have one very important redeeming feature: it allows the knight to remain on the magnificent d4-square, at least for the time being. White can and should attempt to exchange this piece, but it turns out that Black has a surprising num-

229

Starting Out: The Accelerated Dragon number of tactical resources which have tripped up many an unsuspecting opponent. White has two main ways of handling the position. He can play solidly, completing development and aim for a modest edge; or he can play ambitiously and aggressively, striving for a larger advantage but with a much smaller margin for error. Having researched both options extensively, I strongly believe that the solid approach represents a more pragmatic choice for most players, for reasons that will be explained. Before considering White's two main approaches we should briefly note there is little value in doubling Black's pawns with 10 Bxd4?! exd411 Nd5 0-0 12 Bd3 d613 0-0 Be6. White is not worse here, but he has no real chance of an advantage either. TIP: Try not to place too much emphasis on pawn structure. Yes, there are times when it matters, but here the weakening of White's dark squares, combined with the improved scope of the g7-bishop, makes it a bad trade for White.

The Ambitious Approach 10 NbS!? (Diagram 6}

This is regarded as the most ambitious approach, with which White immediately challenges Black's pride and joy on d4. 10...0-0! Black must rely on tactical resources. The feeble 10... Nxb5? would be a gross positional error, as the theoretical weakness of the doubled b-pawns is completely insignificant here. After 11 cxb5 White will follow up with Bc4, obtaining a crushing positional bind. WARNING: In the g...es variation Black should only exchange knights on bs if he is assured of gaining something in return, e.g. achieving the freeing ... d7-d5 break. If White has time to follow c4xb5 with the clamping Bc4-d5 then he will almost always gain a massive positional advantage.

A Dangerous Pitfall If the reader is new to this variation, the first thought to cross his mind after 10...0-0 is likely to be: "what happens if White captures on d4?". This course of action is in fact fraught with danger and should be avoided at all costs, for reasons that we will soon see ...

11 Nxd4? exd4 12 Bxd4 QaS+! Forcing the king to move. 13 Ke2 ReS 14 f3 (Diagram 7) So far this may not seem so bad for White. True, he has had to move his king, but his posi-

tion appears solid and it should only take a couple of moves to finish developing and castle artificially. Well, if that were the case then the entire 9...e5 variation would be refuted; unfortunately for White there are some unpleasant surprises in store. 14... ds! 15 Bxg7 15 cxd5? would be met by 15 ... Rxe4+! 16 fxe4 Bg4+ winning the queen.

230

Mar6czy Bind: 7 ... Ng4 1S ... Rxe4+!! This is the really clever point behind Black's play. Now the white king is drawn into the open, as once again the rook is taboo. 16 Kd3 This is not the only move, but White is struggling in all variations. One example was K.Robatsch-N.Stanec, Austrian Ch. 1995, which continued 16 K£2 Qc5+ 17 Kg3 Qe3! 18 h3 Qf4+ 19 Kf2 Kxg7 20 Qcl Qf6 21 cxd5 Rd4 22 Bc4 Qb6 23 Qc3 Kg8 24 K£1 Bd7 with ongoing pressure. 16... Rxc4! (Diagram 8)

Diagram 7 (B)

Diagram 8 {W)

How to continue the attack?

Black has a very strong attack

The model game L.Brunner-R.Ekstroem, Swiss Grand Prix 1990, continued 17 Ke3 QcS+ 18 Kd2 Kxg719 Bxc4 dxc4 20 Qe1 QgS+ 21 Kc3? (21 Ke2! was the only chance, though Black would have a pleasant choice between 2l...Qe5+, 21...Qxg2+ and 2l...Bh3!?) 21...Qf6+1 22 Kc2 BfS+ 23 Kc1 c3! 24 bxc3 ReS 25 Kb2 Qb6+ 26 Kc1 QaS 27 Kb2 QbS+ 28 Kc1 Qb4 0-1 Obviously White can do much better than 11 Nxd4?, but this game is quite indicative of the dangers facing White in the 10 NbS variation. Theoretically he may be slightly better, but the margin for error is considerably tighter than after 10 Bd3. We will now look at what can happen if White is well prepared and plays in the most ambitious fashion possible. One would think that 9 ... e5 ought to be semi-refutable with perfect play, but it turns out that Black's resources are considerable.

Best Play 11Qd2! This is considered the most challenging move, though 11 Be2!? is a reasonable alternative. At this point Black has two decent continuations: a) 11... Qh4 (Diagram 9)

231


Diagram 9 (W)

Diagram 10 (W)

Activity is paramount for Black

Looks wild, but it's all theory

This is the main line, though 11...Qe7!? also appears playable as we will see later. 12 Bd3

12 Nd6 Qe713 Nxc8 doesn't look particularly convincing after either 13 ... Rfxc8 14 Bd3 (R.Vaganian-L.Espig, German League 1990) when Black should try 14 ... Ne6!? (Nielsen and Hansen) intending ... Bf6-g5; or 13 ... Raxc814 Bd3 f515 exf5 gxf516 Bxd4 exd4+ 17 Kfl (P.Freisler-M.Konopka, Lazne Bohdanec 1999) and now 17... d5!? looks best, when 18 cxd5 Qc5 regains the pawn with approximately equality. 12 ... d5! (Diagram 10)

This clever move helps Black to activate his pieces as quickly as possible. 13 cxds

The most popular capture. It is not my intention to analyse all possibilities in detail, so I will only mention briefly that 13 exd5 Bh3! leads to great complications, with the general consensus being that Black is ok. 13 ... Nxbs

Here 13 ...Bh3? is refuted by 14 Bg5! Qg4 (if 14...Qh515 Nxd4 Bxg2 16 Rg1 wins a piece for no compensation, e.g. 16... Qxh2 17 Ne2 etc) 15 Nxd4 Qxg2 (15 ... Bxg2 16 f3 wins) 16 0-0-0 exd4 17 Rhg1 Qf3 (or 17... Qxh2 18 Bf4) 18 Rg3 Qh5 19 Be2 and Black can resign. 14 Bxbs Qxe4 15 o-o (Diagram 11)

After a brief tactical skirmish the position has stabilized. The outcome of the game is likely to hinge on whether the passed d-pawn turns out to be a strength or a weakness. There can be no doubt that White's chances are to be preferred in Diagram 11. At the same time he is certainly not winning by force, as there is no clear way to promote the pawn or achieve a dominant position. Black must aim to establish a secure blockade by establishing a piece outpost on either d6 or d7. If he can do this without making any major concession then he should be ok, and may even have chances to take over the initiative if he can turn the dpawn into a weakness.

232


Diagram 11 (B)

Diagram 12 (B)

Is the d-pawn strong or weak?

A nightmare for Black

15 ... Rd8! This looks like the best chance. The game G.Pinter-C.Philippe, Budapest 2000, provided a model example of White's chances after 15 ...Qf5? 16 Bc5 Rd8 17 Be7 Bd718 Bxd8 Bxb5 19 Bc7 Bxf120 Rxfl Qd7 21 d6 ReS 22 Qd5 Bf8 23 Rd1 (Diagram 12), when the mighty d-pawn paralysed Black's forces and soon enabled White to win the ending. 16 Rfd1 QfS 17 RaC1 Bd718 Be2 e4! (Diagram 13) Gaining some space and, more importantly, liberating the g7-bishop.

Diagram 13 {W)

Diagram 14 {W}

Black must play energetically

Black is hanging on

19 Rc7 19 g4!? wins a pawn, but in A.Biaux-J.Lecoq, Correspondence 2002, Black gained enough activity after 19...Qe5 20 Bd4 Qxd4 21 Qxd4 Bxd4 22 Rxd4 Rac8 23 Rxc8 Rxc8 24 Rxe4 K£8

233

Starting Out: The Accelerated Dragon (intending to invade with the rook) 25 B£3 when a draw was agreed. 19 ... Bc6! 20 dxc6!

The critical continuation. 20 Bc4 and 20 d6 are both adequately met by 20... Be5. 20... Rxd2 21 Rxd2 bxc6 22 Bc4

White has very dangerous compensation for the queen, but Black's defensive resources just seem to hold up. 22 ... Be5 23 Rxf7

23 Rb7? turned out badly for White in J.Quist-L.Espig, Berlin 1993: 23 ... Rf8 24 Rxa7 Bf4 25 Ra3 Bxe3 26 Rxe3 Kg7 27 Rde2 Rd8 28 Bb3 Rd4 29 Bc2 Qc5 30 Bxe4 Rd 1+ 31 Re 1 Qxe3 32 fxe3 Rxe1+ 33 Kf2 Ra134 a3 c5 35 Bd5 Rd136 e4 Rd2+ 0-1. 23 ... Qxf7 24 Bxf7+ Kxf7 (Diagram 14)

The ending may appear to be winning or at least clearly better for White, but his ambitions are considerably hampered by the reduced material and Black's impending queenside counterplay. The game E.Gufeld-L.Espig, Sukhumi 1972, finished 25 Rd7+ Ke6 26 Rxh7 a5! 27 b3 a4 28 bxa4 Rxa4 29 Ra7 Rb4 30 Kfl Rb1+ 31 Ke2 Bc3 32 f3 Rb2+ 33 Kfl exf3 34 gxf3 Rxh2 35 a4 Ra2 Vz- 1/z. I spent some time looking at various alternatives over the past ten moves, but was unable to find anything truly convincing for White. Black needs to be accurate, but ultimately he seems to be ok. It is hard to know how to assess such a variation, in which Black is somewhat worse but is able to hold the draw. From a strictly theoretical viewpoint one must acknowledge that 'if it's a draw, then it's a draw!' and that Black should be happy. Then again, it's always good to have some winning chances as well, especially when facing a lower rated opponent. Here is a second possibility for Black which also deserves serious consideration. b) 11... Qe7!?

This is played less frequently than ll ... Qh4, but still seems to be holding up. 12 Be2!

This looks best. 12 0-0-0?! has been the most popular choice, but has scored very badly for White after 12... Nxb513 cxb5 d5! 14 exd5 (14 Qxd5? Be6leaves White in deep trouble, especially in view of 15 Qd6? Rac8+ 16 Kb1 Rfd8! winning) 14... Rd8. Now ... Bf5 is a big threat, as the reply Bd3 can be met by ... Rxd5. Several games have continued 15 d6 Qe6 16 Kb1 Bf8 (Diagram 15), when White is under pressure; his statistical score of 32% from this position says it all. 12 ... Nxb5 13 cxb5 b6!

This looks quite an effective plan, and is certainly more promising than 13... d5? 14 exd5 Rd815 0-0 which gave Black very little for the pawn in J.Rowson-Y.Afek, Wijk aan Zee 2000. 14 0-0 Bb7 15 f3 It is important that Black was able to gain time attacking the e-pawn, otherwise White

would have played Bc4 with a positionally won game. 15 ... d5! (Diagram 16) It is essential to get this in before White clamps down with Bc4-d5. The pawn sacrifice is

only temporary as Black will regain it after doubling rooks on the d-file. V.Zakharstov-

234

Mar6czy Bind: 7 ... Ng4 V.Tseshkovsky, Krasnodar 1998, continued 16 exdS Rad817 Bc4 Rd718 Rad1 Rfd8 19 Qd3, and now Black should have played 19...BxdS 20 BxdS RxdS 21 QxdS RxdS 22 RxdS h6 when he shouldn't encounter too many difficulties in the endgame. White's slight material advantage of two rooks vs. queen is offset by his inferior pawn structure, and Black can aim for ...eS-e4 to activate his bishop and open up the white kingside. I believe the position to be equal, though it would be nice to see a practical test.

Diagram 15 (W)

Diagram 16 (W)

White is playing with fire

Equalizing?

Theoretical? The positions occurring after 10 NbS are quite sharp and theoretically demanding. From White's point of view this might be worth the effort if he could be sure of obtaining a big advantage, but that does not seem to be the case - White is struggling to find anything truly convincing against both 11...Qh4 and 1l...Qe7!?. Black will certainly need to know his stuff, too, as a single mistake can easily land him in a bad position. It is also important to know the refutation of moves like 11 Nxd4?- there is no point in offering a poisoned pawn if you don't know how to administer the poison!

Statistics 10 NbS scores a total of 63% for White. After the usual moves 10...0-0 11 Qd2, Black scores 36% with 11 ... Qh4 and 42% with 11 ... Qe7!?.

The Solid Approach 10Bd3! I believe this to be a better practical choice than 10 NbS, as White should be able to obtain an advantage by relatively simple means, without having to know too much theory. In this line White doesn't try to do anything fancy in the early stages, and will definitely not have to worry about poisoned pawns and mating attacks. Instead, he will complete development

235

Starting Out: The Accelerated Dragon and later attempt to force the exchange of the knight on d4, usually for his own knight. If he can do this successfully, he should have good chances of exerting unpleasant pressure along the d-file and on the queenside generally. 10...0-0 110-0 d6 (Diagram 17)

The alternative bishop development with 11...b6?! looks inferior, as White can easily carry out his plan of challenging the d4-knight after 12 Qd2 Bb7 13 f3 a6 14 Ne2. The Israeli IM Yochanan Afek has had this position on a couple of occasions. In S.Kindermann-Y.Afek, Fuerth 1999, the continuation was 14... b5 15 cxb5 axb5 16 Nxd4 exd417 Bh6 Qa518 Qxa5 Rxa5 19 Bd2 with a clearly advantage to White. A year later in W.Hendriks-Y.Afek, Wijk aan Zee 2000, he tried to improve with 14... Nxe2+ 15 Bxe2 Bc6, but after 16 a4 Rb8 17 b4 Re8 18 b5 axb5 19 axb5 Bb7 20 Rfd1 once again found himself in a bad position.

Diagram 17 (W} White wants to exchange knights

Diagram 18 (W}

Black is ok

Returning to 11...d6: Both sides have developed naturally and it is time for White to decide on a plan. This will usually mean trying to trade off Black's pride and joy on d4, and there are various ways in which he can prepare to do so. Before examining the main line it is worth familiarizing ourselves with a few different approaches: a) The direct 12 NbS allows Black to free himself with 12... Nxb5 13 cxb5 d5!, after which 14 exd5 Qxd5 15 f3 Be6 was equal in Z.Almasi-B.Chatalbashev, Krynica 1998. b) 12 Ne2 avoids compromising White's pawn structure, but is slightly less active, so Black may be tempted to try 12... f5!? (12 ... Be6 is a more solid alternative), when 13 Bxd4 exd4 was prematurely agreed drawn in A.Yermolinsky-L.Christiansen, Woburn 1999. c) White can postpone knight-challenging operations in favour of continuing development, though this does not seem too threatening. After 12 Qd2 Be613 Racl (13 Rad1 a614 b3 Qc7 15 Bb1 b516 cxb5?? Qxc3! 0-1 was a disaster for White in R.Pert-C.Ward, British Ch., Douglas 2005) 13 ... a614 b3 Rb815 Ne2 b5 16 Nxd4 exd417 Bh6 Bxh6 18 Qxh6 Qa5 (Diagram 18) gave Black enough counterplay and a draw was agreed here in O.Korneev-V.Georgiev, Elgoibar 1999. In my opinion White's most promising move is ...

236

Mar6czy Bind: 7 ... Ng4 12 a4! (Diagram 19)

Diagram 19 (B)

Diagram 20 (B)

The critical variation

Challenging the black knight

The main idea is to prepare Nc3-bS, after which White will always have the useful option of recapturing with the a-pawn, maintaining the central clamp while opening another file for the queenside attack. A secondary plan involves a4-aS to gain some territory. I believe this to be the critical test of 9...eS, though Black's position remains playable. 12 ... a6!? is one logical response, preventing NbS once and for all. This will be examined in Game 28. Here is a summary of the main alternatives: a) 12 ... as!? was suggested by Nielsen and Hansen but remains untested as far as I can tell. The idea is to prevent the further advance of the a-pawn while answering 13 NbS with 13 ... Bd7, intending to meet 14 Nxd4 exd4 1S Bd2 with 1S ... Qb6 16 b3 Bc6 followed by doubling rooks on thee-file, after which it is hard for White to do anything active. Instead, 14 Be2!? (Diagram 20) looks more promising, hoping to force Black to exchange his strong knight voluntarily, after which White can pile up on the d-file. If Black is determined to maintain the knight on d4, for example with 14 ...Qb6 1S Qd2 BxbS, then White gains a big advantage with 16 cxbS! intending to activate the light-squared bishop on c4. b) 12 ... Nc6!? has been played only rarely, but is quite a logical move. Black again prevents the further advance of the a-pawn while pre-empting White's intended NbS by moving the knight out of harm's way. In volume three of his Opening for White according to Kramnik series, Alexander Khalifman cites this move as his main reason for not advocating 10 Bd3, so the idea must be taken seriously. Following the logical13 NdS Be6, R.Vera-C.Matamoros Franco, Cienfuegos 1997, continued 14 Qd2 Nd4 (the knight has done its duty on c6 and, as White's knight has strayed from bS, returns to its best square) 1S Ra3 and now 1S ... aS! (Diagram 21) was a good attempt to stifle White's queenside play. Vera opted for 16 b4!? here, opening the b-file. The decision looks justified, as White's concentration of forces on the queenside should be enough to offset the weakening of his structure there. Still, it is not the only good plan and I wonder if a case could be made for 16 Bxd4!? exd417 Rb3, now that some holes have been created on the black queenside.

237


Diagram 21 (W) Blocking the queenside

Diagram 22 (B) A tough ending for Black

After 16 b4!? the game continued 16 ... BxdS 17 exdS axb418 Qxb4 Qd719 Rb1 Ra7 20 aS Rfa8 21 Qb6 B£6 22 cS Bd8 23 Qxd6 Qxd6 24 cxd6 RxaS 2S RxaS RxaS 26 Bc4 ReS 27 Rxb7! Rxc4 28 Rb8 Kg7 29 Rxd8 ReS 30 £4 RxdS 31 fxeS (Diagram 22), reaching an endgame in which White had won a pawn but Black retained some drawing chances. Khalifman opines that Black should be able to hold this ending with accurate defence and thus concludes that 10 Bd3 is not quite strong enough to recommend to his readers. However, with the greatest respect for the judgement of a player far stronger than myself, I am not at all sure if Black can draw this position, and even if a hypothetical 13-piece tablebase were to prove it, I doubt that many players would relish the task of having to do so over the board. In any case, for our purposes it is largely irrelevant to focus on this particular endgame. The most important thing is that for the entire game it has been White who has possessed the initiative and put pressure on his opponent. His position has clearly been the more pleasant from start to finish, and I see no reason to dampen my enthusiasm for White's chances after 10 Bd3. c) Black can also simply develop with 12... Be6, hoping that after the logical sequence 13 NbS a6 14 Nxd4 exd4 1S Bd2 Rc8 16 b3 fS his slightly more active pieces will compensate for his structural deficiency. Several games have continued 17 exfS BxfS18 Qf3 dS! (Diagram 23).

Black has played in a logical and principled fashion, and it appears that he has succeeded in freeing his position. Unfortunately, tournament practice has indicated that White should be able to retain an edge as long as he is careful to maintain a solid blockade of the d4pawn. In V.Smyslov-G.Fabiano, Rome 1990, the former World Champion demonstrated his technical prowess after 19 Racl Qd7 20 cxdS Bxd3 21 Qxd3 QxdS 22 Rxc8 Rxc8 23 Re1 QfS 24 Re4! (24 QxfS?! would damage Black's structure but surrender the blockade) 24 ... Rf8 2S f3 (Diagram 24), when White had achieved a stable endgame advantage. The d4-pawn may be passed, but it is securely blockaded and liable to become weak after further major piece exchanges. Smyslov went on to win with some model endgame technique: 2S ... QcS 26 Re1 Rc8 27 Rcl QfS 28 Rxc8+ Qxc8 29 Kf2 Qd7 30 Qc4+ Kh8 31 Ke2 Bf6 32 Bh6 gS 33 QcS Qe8+

238

Mar6czy Bind: 7 ... Ng4 34 Kd1 d3 35 Qf8+ Qxf8 36 Bxf8 Be5 37 g3 Bc3 38 Bh6 b5 39 axb5 axb5 40 Bxg5 Kg7 41 g4 K£7 42 h4 Ke6 43 h5 Kd5 44 Bh6 Kc5 45 Bf8+ Kd5 1-0.

Diagram 23 (W) Trying to get active

Diagram 24 (B) Model technique from White

Theoretical? Not really. The main advantage of 10 Bd3 over 10 Nb5 is that it offers at least as many (and arguably even greater) chances of an opening advantage while also reducing the theoretical burden.

Statistics 10 Bd3 scores a total of 63%- exactly the same as 10 Nb5. After the usual moves 10... 0-0 11 0-0 d6, 12 a4! scores a whopping 78%. The conclusion is obvious: White is in excellent shape, and Black will need to find some sort of improvement if he wishes to play 9 ... e5.

Illustrative Game The 9 ... e5 variation has seldom featured in recent Grandmaster practice, but the following game from 1998 remains a highly instructive model for White. Game28

D A.Shabalov •

A.Baburin

Mermaid Beach 1998

1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 Ng4!? 8 Qxg4 Nxd4 9 Qd1 es!? 10 Bd3 o-o 11 o-o d6 12 a4! a6

We have already seen Black struggling to equalize with other moves. The text is logical in that it prevents White's principal positional threat of Nb5, though it does allow him toreveal the secondary purpose of his twelfth, namely the acquisition of queenside territory with ...

239

Starting Out: The Accelerated Dragon 13 as! (Diagram 25) Fixing b6 as a potential outpost for the future.

Diagram 25 (B)

Diagram 26 (B)

White will play on the queenside

A pleasant advantage for White

13 ... Be6 Black should focus on completing development. 14 Qa4 ReB 15 Nds Bxds This is not forced, but it is understandable that Black did not wish to concern himself with the prospect of a knight invasion on b6. 16 cxds (Diagram 26) White has every reason to feel satisfied with the outcome of the opening. The black knight may be superbly placed on d4, but his other minor piece is passive and in this sort of position White can play around the knight, perhaps aiming to create a target on the queenside while retaining the option of Bxd4 at a moment of his choosing. 16 ... Rcs If Black tries for counterplay with 16.. .£5 then White can make a favourable exchange 17 Bxd4 exd418 exf5 gxf5, when Black's fractured kingside pawns render his position slightly but permanently unfavourable, notwithstanding the presence of opposite-coloured bishops. 17 Rfc1 Qe7!? Allowing a change in the pawn structure; one can certainly sympathize with Baburin's wish to alter the course of this somewhat unpleasant position. Once again 17.. .£5 does not solve Black's problems after 18 Bxd4 Rxcl+ (if 18 ... exd4?! 19 Rxc5 dxc5 20 exf5 gxf5 21 Qb3! is troublesome) 19 Rxcl exd4 20 exf5 gxf5 21 Qb4 when White has a lasting advantage. 18 Rxcs dxcs 19 Rc1 Rd8 20 Qc4 Bf8 21 f4! (Diagram 27) This is an excellent move, undermining the d4-knight while opening things up for the white bishops. 21... Qf6

240

Mar6czy Bind: 7 ... Ng4 21...exf4 22 Bxf4 Bg7 sets an elementary trap in 23 QxcS? QxcS 24 RxcS Ne2+! 2S Bxe2 Bd4+, but White can maintain a clear advantage with 23 Khl, as pointed out by Tsesarsky.

Diagram 27 (B)

Diagram 28 (B)

Opening the position

A dream pawn centre

22 b41

22 Rfl!? also looks strong. The text continues the undermining policy as well as creating a strong aesthetic impression. 22 ... exf4 23 Rf1 bs?

This does not help; but even after the superior 23 ...Qe7 White can retain a near-decisive advantage with 24 Bxf4 NbS 2S eS cxb4 26 d6 Qd7 27 Qxb4. 24 axb6 Qxb6 25 Bxf4?

White is still much better after this, but given that his earlier play revolved around undermining the enemy centre it is a pity that he missed the opportunity to wrap up the win in wonderfully thematic fashion with 2S Rxf4! Bh6 26 bxcS! Qb8 27 g3! Bxf4 28 gxf4 NbS 29 eS (Diagram 28), when Black will soon be overwhelmed by a tsunami of white pawns. 2S ... Qxb4 26 Qxa6 Nb3?

Now the win becomes relatively straightforward. Tsesarsky' s 26 ... Bg7! would have been more resilient, though 27 Bc7 or the patient prophylactic move 27 Khl! would maintain White's large advantage. 27 Bes!

Threatening Qf6. 27 ... Be7

27...Bg7? is impossible due to 28 Bxg7 Kxg7 29 Qf6+. 28 Qa7! (Diagram 29)

The attack along the seventh rank is deadly: both e7 and £7 are vulnerable, and the impending advance of the d-pawn will break the defence. 28... C4

28 ... Kf8 29 d6 Rxd6 30 Bxd6 Qd4+ 31 Khl Qxd6 32 Bc4 would force resignation.

241

Starting Out: The Accelerated Dragon 29 d6 ReS 30 Qd7 Qb6+ 31 Kh1 Qd8 32 Bxc4 1-0

Diagram 29 (B)

Diagram 30 (W)

Black is doomed

The main line

The Main Line with 9 ... Ne6 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 Ng4!? 8 Qxg4 Nxd4 9 Qd1 Ne6 {Diagram 30)

This is the soundest way for Black to handle the 7... Ng4 system. The knight retreats to an excellent central square, from which it may influence events on either side of the board, while clearing the long diagonal for the g7-bishop. ~

._,

It is

WARNING: White must remain vigilant here as he is already faced with the positional threat of 10... Bxc3, wrecking his queenside pawns.

~that

this exchange would involve the loss of Black's most prized asset on g7, but this is not so terrible when Black has not yet castled. We will examine the consequence of the ... Bxc3 exchange shortly. The majority of games that reach Diagram 30 continue with 10 Rc1, which is definitely the most accurate move in my opinion. 10 Qd2 is also frequently seen, but is slightly less flexible as the queen does not always have to move here, whereas the rook will usually end up on c1 regardless. Assuming that White prevents ... Bxc3 by opting for one of these two moves, then Black will need to make a fundamental choice between two distinct strategies: 1) The conventional method involves short castling combined with the standard development of the queen's bishop on either d7 or b7. This is a reasonable enough strategy, though White ought to be able to maintain a slight edge. 2) There is, however, a much more interesting way to play the position, which I have used with success on several occasions. The idea is to utilize the position of the knight on e6 to support an audacious kingside pawn advance with ...g6-g5!?. We will investigate both of these strategies later in the chapter, after first conducting a brief excursion to investigate the positions in which Black is allowed to execute his positional threat of ... Bxc3.

242


Doubling the c-pawns After 9... Ne6 a surprising number of games have continued with the stereotypical developing move 10 Be2?!. The 'Dragon bishop' has justifiably earned itself a reputation as a formidable piece, and I suspect that in many cases it simply does not occur to White that the opponent may be planning to 'sacrifice' it for a mere knight, regardless of the structural repercussions. If you reach this position with Black then I would strongly advise you to take the opportunity to cripple the white queenside with 10... Bxc3+! 11 bxc3 (Diagram 31)

l:)

y_

Diagram 31 (B)

Diagram 32 (B)

Pawn structure over bishops

Excessive greed from Black

TIP: The idea of ... Bxc3, shattering the white queenside pawns, is especially promising when Black has not yet committed himself to short castling.

Black s correct strategy from this point involves a harmonious development of the queenside pieces while delaying castling in order to avoid presenting White with a fixed target. In my opinion the most accurate continuation is 11 ...Qc7!.

WARNING: Black must not get too greedy! After 11... Qa5 12 o-o! Qxc3? 13 cs! (Diagram 32) 13 ...0-o 14 Rb1 Qes 15 Qds White had more than enough for the pawn in D.Tunks-A.Greet, British League 2003. Naturally Black would have been better off refraining from 12... Qxc3?, but in any case I think that the queen is better off on c7, as it could become exposed on aS, for instance after 12 ... b6 13 QdS!. Following 11... Qc7!, the game T.Imanaliev-Z.Lanka, Moscow 1979, continued 12 0-0 b613 Bd4 (Nielsen and Hansen suggest 13 f4 as an improvement, but add that they still prefer Black after 13 ... Bb7) 13.. .f6! (still delaying castling, while usefully increasing Black's influence over the dark squares) 14 Re1 Bb7 1S Bfl d6 16 Qg4 NcS, when Black had successfully developed his queenside while maintaining a rock solid position. Later he can castle at a moment of his choosing. Similar positions can occur after 10 Qd2 Qas, in the event that White neglects to play the best move 11 Rcl!. True, Black has committed his queen to aS, but White has also lost a little flexibility in declaring the position of his own queen. After something like 11 Be2?!

243

Starting Out: The Accelerated Dragon Bxc3! 12 bxc3 d6 (12 ... b6 is also possible) the game M.Rocius-R.Bilinskas, Lithuanian Ch. 1995, continued 13 0-0 Bd7 (13 ... Nc5 may be slightly less accurate in view of 14 e5!?) 14 f4 Bc6 15 Bf3 f6 (Diagram 33) [15 ... Nc5 also looks entirely reasonable]

Diagram 33 (W}

Diagram 34 (B)

Another promising position for Black

Correctly avoiding doubled pawns

when Black had again developed his queenside pieces successfully while keeping the position closed. White is struggling for an active plan, and Black went on to win.

Theoretical? There is very little to learn. Black just needs to avoid being too greedy (as in Tunks-Greet), and develop his queenside pieces in such a way as to prevent White from forcing a favourable opening of the position.

Statistics After 10 Be2?! Bxc3+! 11 bxc3 Black has scored 60%, while the similar 10 Qd2 Qa5 11 Be2?! Bxc3! 12 bxc3 d6 has yielded an even better 65%. The conclusions are obvious. If you play this variation with White, then make sure you remember 10 Rcl!. If you play Black, then you should be very happy if your opponent gives you the opportunity to wreck their queenside.

White's Best Continuation 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 Ng4!? 8 Qxg4 Nxd4 9 Qd1 Ne6 10 Rc1! (Diagram 34)

With this accurate move White eliminates any risk of doubled c-pawns, without prematurely committing his queen. From Diagram 34 Black has two main ideas. The standard plan involves short castling combined with the development of the light-squared bishop on either d7 or b7. This set-up

244

Mar6czy Bind: 7 ... Ng4 is reasonably solid, but is also quite passive and Black sometimes has trouble finding counterplay. The more creative plan involves 9...Qa5, delaying castling in favour of an early ...g6-g5 to control the dark squares. This is much more interesting and will provide the main focus for the remainder of the chapter.

Conventional Play with ...0-0 Black has a choice of two options for his queen's bishop. a) Quite often he opts for a double fianchetto with ... 10... b6 Here White should give serious attention to the space-gaining ... 11 b4!7 (Diagram 35)

t:;)

w

Diagram 35 (B)

Diagram 36 (B)

Playing against the knight

Black is struggling for activity

TIP: This can be a very useful resource for White, as it severely restricts Black's knight by taking away the c5-square.

This is not the only good plan. White also obtained a slight edge after the quieter 11 Bd3 Bb712 0-0 0-0 13 Bb1 ReB 14 b3 d6 15 Qd2 (15 f4!? is a more aggressive alternative) 15 ... a5 16 Nd5 in R.Hiibner-R.Hernandez, Las Palmas 1976. 11... Bb7 12 Bd3 0-0 13 0-0 ReS Black can hardly hope to activate the rook by opening the a-file, as this would leave his bpawn terribly weak. 14 Qd2 (Diagram 36) This position has often been reached. White has a comfortable advantage and has scored about 67%. Black's position is solid enough but it is difficult for him to generate counterplay, while White can aim to improve his position with moves like Rfd1 and a timely Nd5. Occasionally Black can aim for ... f7-f5, but in most cases this only brings him weaknesses. Personally this is the type of position that I prefer to avoid when facing the Mar6czy Bind.

245

Starting Out: The Accelerated Dragon b) Black's alternative mode of development involves putting the bishop on d7 after ... 10...d6 (Diagram 37)

0

Diagram 37 (W)

Diagram 38 (B)

A more solid set-up

The a-file provides counterplay

NOTE: For some reason most players have tended to play 10... b6 or 10... d6 before castling, although there is nothing wrong with the immediate 10... 0-0 followed by a subsequent bishop development.

After 10... 0-0 I would once again recommend 11 b4!? for White, reaching the same positions after a subsequent ... b7-b6 or ... d7-d6. Now several strong players have employed the familiar move ... 11 b4!? ... again with the intention of restricting the black knight. P.Orev-Y.Kraidman, Saint Vincent 2001, featured an alternative approach: 11 Bd3 0-0 12 0-0 Nc5 13 Bb1 a5 14 Qd2 Bd7 15 f4!?, with which White initiated active operations on the kingside. 11 ... 0-0

WARNING: In view of the effectiveness of White's b2-b4 advance, the reader may be wondering why Black doesn't nip this idea in the bud with 10... as. Unfortunately, this turns out badly after 11 cs! when Black will have trouble developing his queenside. 12 Be2

0

NOTE: Now that Black has hinted that his bishop is going to d7 rather than b7, it is less logical for White to develop his king's bishop on d3, as the e4pawn is unlikely to require additional defence. Instead he should put it on e2 to keep the d-file clear.

12 ... as!? This is something of a double-edged move, opening the a-file at the cost of weakening the b5- and b6-squares. Black can also revert back to the double fianchetto with 12... b6 13 0-0 Bb7, but White maintains the usual stable advantage after 14 Qd2 or 14 Nd5. 246

Mar6czy Bind: 7 ... Ng4 13 a3 axb4 14 axb4 Bd7 15 o-o Bc6 16 Qd2 Ra3

Black should not be tempted to win a pawn with 16... Bxc3?! 17 Rxc3 Bxe4?, as after 18 Bh6 Re819 Re3! Bc6 20 Qb2 (Khalifman) he can resign. 17 Nd5 (Diagram 38)

White's space advantage and central control should enable him to claim an edge here, although the open a-file is a major boon for Black. His best continuation is probably 17... Re8!?, when 18 Rfd1 Nf819 h3 Nd7 saw Black improve his slightly passive knight in S.MarjanovicD.Velimirovic, Sarajevo 1984. White tried 20 c5!?, but after 20...Ba4 21 cxd6 exd6 22 Re1 Bc6 23 Bg5 Nf6 24 Nxf6+ Bxf6 25 Bxf6 Qxf6 26 g3 Rea8 27 Bc4 a draw was agreed. A better example for White was shown in M.Wiese Jozwiak-D.Klusek, Polish Women's Ch. 1989, which continued with the more direct 18 h4!? Qb8 (Black could consider 18 ... h5!?, though one must always consider the long-term ramifications of such a move) 19 h5 b5 20 h6 Bh8 21 Rfd1 bxc4 22 Bxc4 Ba4 23 Re1 Bb5 24 Qe2 Bxc4 25 Qxc4 (Diagram 39)

Diagram 39 (B) Black has big problems

Diagram 40 (B) A typical set-up

when White was clearly better thanks to her powerful passed pawn and its troublesome team-mate on h6. The remaining moves were 25 ... Kf8 26 f4 Rxe3!? (the best chance, but it is not enough to save the game) 27 Nxe3 Nxf4 28 Kh1 Qa7 29 Rc2 Nh5 30 Rf2 e6 31 g4 Ng3+ 32 Kg2 Bd4 33 Ra2 Qb6 34 Ra6 Qb7 35 Qxd4 Nxe4 36 Ra7 e5 37 Rxb7 exd4 1-0.

Theoretical? Neither side requires more than a minimal amount of theoretical knowledge to play this variation. Positional understanding is far more important.

Statistics Black scores 33% with 10... d6 and a slightly better 37% with 10 ... b6, though this drops to just 26% if White plays an early b2-b4.

247

Starting Out: The Accelerated Dragon The lines with an early ... 0-0 are playable for Black, but there are times when he finds it difficult to obtain counterplay, especially after the double fianchetto. If the reader is considering playing the black side, then I would strongly recommend the development of the bishop on d7, intending to meet an early b2-b4 with ... a7-a5, opening the a-file. Objectively White is probably still a bit better, but Black's position is quite resilient and the open a-file will at least provide a measure of active counterplay.

Larsen's Creative ...g6-gS Plan This is definitely the most interesting way for Black to handle the 9... Ne6 variation. Following the standard opening moves ... 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 Ng4!? 8 Qxg4 Nxd4 9 Qd1 Ne610 Rc1! (or 10 Qd2)

... Larsen's plan begins with the active development of the queen: 1o...Qast?

Black intends to develop his light-squared bishop on b7, combined with the audaciouslooking ... g6-g5!? advance. Larsen introduced this ingenious plan in 1987 against Karpov and went on to use it with success in many subsequent games. Diagram 40 shows a typical application of Larsen's plan. Before saying any more I should clarify that this plan is not usually played with the intention of developing a checkmating attack on the white king. There may be occasions on which Black ends up attacking, but the primary objective of ... g6-g5 is to increase Black's control over the central dark squares. So in spite of its cavalier appearance, the ...g6-g5 plan should primarily be viewed as being positionally motivated. The e5-square is of especially great importance for Black. Often he will use it as a base for the queen or dark-squared bishop, both of which will coordinate very nicely with the knight on e6. A subsequent pawn advance with ... h5-h4 will increase his influence over the dark squares even further.

0

NOTE: At the risk of stating the obvious, the pawn advance ... g6-g5 is an indispensable part of this dark square strategy, because otherwise White could easily arrange to play f2-f4.

Before moving on to some illustrative games, I would like to offer two important pieces of advice for these positions: 1) A queen exchange will generally favour Black. His centralized king will feel slightly safer with queens off the board, and meanwhile he will maintain his dark square clamp over the position. 2) Black should always be on the lookout for an opportunity to exchange the dark-squared bishops, especially in an endgame. This will often highlight White's lack of influence over

the dark squares, as well as the passivity of his remaining bishop. One final point: It pains me to have to say this, but I have to be honest and inform the reader that, strictly speaking, the ...g6-g5 plan may not be entirely sound. It all depends on the specific move order employed by White. In the majority of cases the ... g6-g5 plan is quite viable, but there is one particularly effective 'anti-... g6-g5' move order, against which I have been unable to find a truly satisfactory way of making Larsen's plan work. Game 31

248

Mar6czy Bind: 7 ... Ng4 will give details of this critical line, together with some thoughts on possible solutions for Black.

Theoretical? If Black is to employ the ... g6-g5 plan with success then it is essential that he brings to the

table a decent level of theoretical preparation. One must, for example, be able to distinguish between different white move orders and react accordingly. The good news is that there are not too many forcing lines to remember, and once you reach move 15 or so you should be able to rely mostly on positional understanding. But just to repeat, it is absolutely crucial to know what you are doing on moves 11 to 15!

Statistics From the position after 14... g5! in Game 29 Black scores an impressive 55%. This is encouraging, as White's moves in that game are very natural and the majority of players tend to react in the same way when confronted with Larsen's plan. White scores much better when he omits the non-essential Qd2. The 14 Rf2 plan found in Game 30 scores 71% for White, though most of these games were played before Black figured out the best response of 16... Rd8!, instead of the dubious 16... d6?. If White finds the most accurate sequence of 10 Rcl Qa5 11 Bd3, then plans involving ...g6-g5 score a pitiful 18% for Black, a sure sign that some repair work needs to be done here.


D A.Karpov •

B.Larsen

Brussels 1987

This was not the first time that Larsen had employed the ...g6-g5 plan (he had used it against Gligoric as far back as 1957), but the present encounter was something of a landmark, as the great Dane pitted his creation against the legendary Anatoly Karpov, one of the greatest players in the history of the game. 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 Ng4 8 Qxg4 Nxd4 9 Qd1 Ne6 10Qd2

I have already mentioned that 10 Rcl is more accurate, but this had not yet been established in 1987. Indeed, it was only in view of the subsequent success of Larsen's ... g6-g5 plan that White began looking for such refinements in the move order. 10...Qas 11 Rc1 b6 (Diagram 41)

White's system with an early Qd2 is not too threatening, so Black can happily continue developing before rushing into ... g6-g5. We will see later that there are certain instances in which he has to be much more careful about the timing of the pawn thrust. 12 Be2

12 Qd5!? looks like a tricky move, especially considering that 12... Qxd5 13 cxd5 Nc5 14 f3 leaves Black clearly worse due to his lack of space and weakened queenside. However, it is much better to play 12 ... Rb8! when 13 Qxa5 bxa514 b3 Bd4! should be fine for Black. He

249

Starting Out: The Accelerated Dragon controls plenty of dark squares and the doubled a-pawns are not weak at all. 12... Bb713 f3

Black was threatening 13 ... Bxc3, winning the e4-pawn. 13 ... h5! (Diagram 42) 14 0-0

Diagram 41 (W)

Diagram 42 (W}

Black should develop first

Commencing the kingside advance

White has also tried 14 b3, hoping that after a subsequent exchange of queens his king will be well placed on d2 (compare the final section of Chapter Seven). Black should continue with the consistent 14 ... h4! (White is not committed to the queen exchange, so 14 ... g5?! could be met by 15 h4!), when V.Kramnik-L.Ljubojevic, Monte Carlo (rapid) 1994, continued 15 Nd5 (or 15 0-0 g5) 15 ... Qxd2+ 16 Kxd2, and now 16... Rc8!? looks best, after which Black can follow up with ... Bd4 to exchange bishops, or perhaps ... g6-g5 combined with ... Be5. 14... gs! 15 Rfd1 d6 16 Nds

I have encountered this move in my own games; it seems to be the natural reaction of a player reaching this position for the first time. But as I have already mentioned, Black will usually feel quite comfortable after an exchange of queens in the ... g6-g5 variation. White can keep the queens on with 16 a3!? h4 (16 ... g4!?) 17 b4, when 17... Qe5! transposes to N.Short-B.Larsen, Brussels 1987 (the last round of the same tournament}, which continued 18 Nd5 Kf819 Bfl Bc6 20 Qd3 Nf4 (Diagram 43) 21 Nxf4 (21 Qd2 Ne6 repeats) 2l...gxf4 22 Bd4 Qxd4+ 23 Qxd4 Bxd4+ 24 Rxd4 Rh5 (24 .. .£6! -Larsen) 25 c5 dxc5 26 bxc5 b5 27 e5 h3! undermining the kingside pawns, when Black had obtained sufficient counterplay and eventually managed to win. TIP: Both sides should make a mental note of the undermining move ... h4h3, especially in the endgame. We will see further examples of its effectiveness in some of the later notes to the present game. 16...Qxd2 17 Rxd2 (Diagram 44) 17... Bes

17...h4 will usually transpose to the main game. Though it is unlikely to make much differ-

250

Mar6czy Bind: 7 ... Ng4 ence, my personal preference has always been for the preliminary advance of the h-pawn, to ensure that the e5-bishop is not driven away by g2-g3 and f3-f4. After 17... h4 one independent possibility is 18 Bd1!?, hoping to find an active role for the bad bishop. G.Kenworthy-A.Greet, British League 2002, continued 18... Be519 Ba4+ Kf8 20 Bd7 Nf4 (20 ... Rd8 21 Bxe6 fxe6 also looks fine for Black) 21 Nxf4 (21 Nb4!?) 21...gxf4 22 Bf2 Rd8 23 Bg4, and now 23 ... Bc8 would have been safest with roughly equal chances.

Diagram 43 (W)

Diagram 44 (B)

A complex but balanced position

A complex semi-endgame

18 b4 18 g3 has hardly ever been tried. A plausible continuation might be 18 ...h419 f4!? Bxd5 20 cxd5 (20 fxe5 Bxe4 21 gxh4 dxe5 22 hxg5 f5! looks better for Black) 20 ... gxf4 21 Bb5+ Kf8 22 gxf4 Nxf4 23 Rc7 with unclear play. 18 ... Rc819 a4 White's only active plan is to prepare a queenside advance. If he is feeling less ambitious, 19 Bd1!? h4 20 Bb3looks like a sensible relocation of his worst placed piece. This should not be too terrifying for Black, though it does mean that he will have put his king somewhere other than the usual f7. 19 ... h4 20 Bf1 This prophylactic move seems to be necessary. In V.Farfan Ortiz-H.Frey, Chilean Ch. 2004, White tried an immediate 20 Ra2 preparing a4-a5, but after the strong reply 20 ...Bd4! 21 Kf2 Bxd5 22 Bxd4 (22 exd5?! Bxe3+ 23 Kxe3 Nf4! creates a double attack on g2 and d5) 22 ... Nxd4 23 exd5 e5! Black was left with the superior minor piece and went on to win. 20.. .f6! This system is all about the dark squares! With 20 .. .f6! Black applies the finishing touch to a highly impressive chain of pawns while preparing to connect his rooks with ... Kf7. 21 Ra2 Planning a4-a5. 21... Bd4! (Diagram 45) 22 Kf2 The game P.Wolff-A.Miles, Philadelphia 1987, provides an instructive example of Black's chances in this line, following 22 Bxd4 Nxd4 23 Rd1 Nc6 24 aS Ne5 and then:

251


Diagram 45 (W)

Diagram 46 (W)

Black loves to swap these bishops!

The thematic kingside undermining

a) The most obvious move is 25 axb6, when there might follow 25 ... Bxd5 26 cxd5 axb6 27 h3!? (taking a tempo to secure the kingside, at the cost of more dark square weaknesses; but 27 Ra6 h3! 28 Rxb6 Kf7 gives Black compensation) 27 ... Kf7 28 Ra6 Rc2 29 Rxb6 Rhc8 30 b5 R8c3 (intending ... Rxf3!) with a difficult position to assess. White has an extra pawn, but Black can aim for counterplay against the king. b) Wolff tried 25 Ne3, which was met by the typical undermining 25 ... h3! (Diagram 46). After 26 axb6 axb6 27 Ra7 Bc6 28 b5 hxg2 29 Bxg2 Bd7 30 Nd5 Rb8 31 Nxb6 Be6 White had won a pawn, but Black clearly has some compensation. Now 32 Nd5! would have kept things unclear after 32 ... Bxd5 33 cxd5 Kf7 34 Rbl Rb6. Instead, White went wrong with 32 Rd4? Rh4 33 Na8?! Bxc4 34 b6 Kf7 35 Ra4 Bb5 36 Ra3 Bc6 37 Nc7 Rxb6 and Black won easily. 22 ... Kf7 23 a5 Bxd5 24 exd5 Bxe3+ 25 Kxe3 Nf4 26 Kd2

Karpov opts for further prophylaxis. Nielsen and Hansen give the line 26 Real Ng6 27 axb6 axb6 28 Ra6 Ne5 29 Kd4? h3! when White has problems. 29 Rcllooks better, but even here Black can obtain excellent chances with 29 ... Ra8! 30 Rxb6 h3!.

0

NOTE: We have seen the undermining ... h4-h3 cropping up time and again in the notes to this game. The reader might be forgiven for wondering why White doesn't just play h2-h3 himself to prevent it once and for all.

The problem is that, by doing so, he almost permanently immobilizes his kingside pawns and abandons hope of ever contesting the dark squares. 26... Rc7 27 axb6 axb6 28 Ra6 Rhc8 (Diagram 47) 29 Rxb6?!

This is harmless. Instead: a) Nielsen and Hansen state that 29 Real is well met by 29 ... b5, though they may have overlooked the strong riposte 30 g3! hxg3 31 hxg3 followed by Bh3 when White is on top. However, Black looks fine after 29 ... Ng6! intending 30 Rxb6 Ne5 regaining the c-pawn. b) Perhaps White's last chance of an advantage lay with 29 g3!. His bishop has contributed

252

Mar6czy Bind: 7 ... Ng4 very little to the game so far, so it is very logical to activate it on h3. After the likely continuation 29...hxg3 30 hxg3 Nxd5 31 Bh3 Nxb4 32 Bxc8 Rxc8 33 Rxb6 Nc6 White is the exchange for a pawn up, though Black should be able to hold the draw without too many problems, e.g. after 34 c5 dxc5 35 Rxc5 Rd8+ 36 Ke3 Ne5. 29 ... Nxd5 30 Rb5 Nf4 31 Ra5 Ng6! (Diagram 48)

Diagram 47 (W}

Diagram 48 (W}

Counterattacking the white pawns

The knight heads for es

Transferring the knight to the wonderful e5-square. Black is now slightly better and Karpov has to take care to hold the draw. NOTE: Black's beautiful pawn chain d6-e7-f6-g5-h4 has been rock solid throughout the endgame. 32 c5 Ne5 33 Rc3!

33 ...h3 was threatened. 33 ...dxc5 34 bxc5 Rb8

34 ... Nd7 does not win a pawn after 35 Ba6 Rxc5?? 36 Bxc8 followed by 37 Bxd7, as noted by Nielsen and Hansen. 35 Bb5 Rd8+ 36 Ke2 Nc6

There was still no way to win the c-pawn, so Larsen acquiesces to the exchange of minor pieces. 37 Bxc6 Rxc6

The draw is now more or less inevitable. 38 g3 Rd4 39 Rb5 hxg3 40 hxg3 Rd5 41 g4 Rc7 42 Ke3 e6 43 Rc2 Ke7 44 Rc3 Kf7 45 Rc2 f5 46 gxf5 exf5 47 Kf2 Kg7 Yz-Yz

This game showed that Black can count on a full share of the chances after an exchange of queens on d2. The notes also demonstrate that in many variations it is White who has to be careful, especially when it comes to the undermining ... h4-h3.

253

Starting Out: The Accelerated Dragon We will now examine a slightly more venomous approach for White, in which he omits the imprecise Qd2 to reach a fascinating middlegame. Game 30

0

G.Drzymala •

T.Caba

Polish Junior Championship 2004 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 Ng4 8 Qxg4 Nxd4 9 Qd1 Ne6 10 Rc1! Qa5 11 Be2

White's play represents a more critical test than in the previous game, but the most accurate move of all is 11 Bd3!, as we will see in Game 31. 11... b6 12 0-0 Bb7 13 f3 g5 14 Rf2! (Diagram 49)

Diagram 49 (B) The start of a clever regrouping

Diagram 50 {W) A unique type of position

White intends to regroup with Bfl and Rd2, activating his rook along the d-file while avoiding a queen exchange. This plan was first introduced by Nigel Short as an improvement over his Brussels encounter with Larsen.

0

NOTE: White can also opt for this plan using the 11 Bd3 move order.

14... h5 15 Bf1 Qe5 16 Rd2 Rd8!

In N.Short-B.Larsen, Hastings 1987/88, Black tried 16... d6?, but after 17 Nd5 Kf8 18 b4 Bh6 19 Qb3 g4 20 Bxh6+ Rxh6 21 Qe3 Qg7 22 f4 he was clearly worse and only managed to scrape a draw when the Englishman blundered in a winning position. And in my opinion 19 Rf2! would have been even stronger. The rook has done its duty on the d-file and makes way for Bd2-c3, while preparing to meet 19... g4 with 20 f4! Qxe4? 21 Bd3. Caba's choice is better as it allows the queen to retreat along the h2-b8 diagonal. 17 Nd5 Qb8 18 b4 Be5 (Diagram 50)

One of the most appealing aspects of the ... g6-g5 plan is that it leads to positions quite

254

Mar6czy Bind: 7 ... Ng4 unlike anything else you might encounter in the Accelerated Dragon, or any other opening for that matter. Black has been shuffling his heavy pieces around the back rank, but is about to succeed in provoking a weakening of White's castled position. 19 h3

The alternative is 19 g3 h4 20 g4 Nf4 (Black can also consider 20 .. .f6!? intending ... K£7) 21 c5 bxc5 22 bxc5 e6 23 Nxf4 gxf4 24 Bf2 Bc6 25 Rb1 Qc7 26 Bb5 h3 27 Bxc6 Qxc6 with approximate equality in J.Soyer-R.Moltchanov, Avoine 2000. 19 .. .f6 20 c5!?

White hurries to open a queenside file. The untested 20 a4 was a valid alternative, intending a timely a4-a5. 20... bxc5 21 bxc5 Bc6 22 Rb1 Qc8 23 Qc1

23 Qc2!? looks interesting, so that in the event of ... Bxd5, e4xd5 Black will have to worry about a queen invasion on g6. The text is not bad though. 23 ... Kf7?!

This allows Black to ferry his heavy artillery over to the kingside if needs be, but it seems a little on the slow side. Instead, I think that Black's number one priority must be to execute the attacking thrust ... g5-g4 to generate some kingside play. As we will soon see, this can be very risky if White can respond with f3-f4. Therefore 23 ... Bg3!? looks quite logical, so that after a subsequent ...g5-g4, the reply f3-f4 will not hit the bishop. To be perfectly honest I am not sure if I completely trust the black position even here, but with ... g5-g4 on the way the game will certainly remain complicated, irregular and difficult for both sides. 24 Qa3! g4? (Diagram 51)

Diagram 51 (W)

Diagram 52 (W)

Over-ambitious play by Black

A complex endgame

This is the move Black wants to play, but it contains a massive flaw. 25 Qxa7?

Returning the favour. White could have obtained a crushing position with 25 f4!; for example, 25 ... Bxd5 26 exd5 Nxf4 (this may have been what discouraged White, but the loss of the pawn is immaterial) 27 Ba6 Qa8 (or 27 ... Qc7 28 Rb7 Qc8 29 c6) 28 Qa4! (a typically powerful

255

Starting Out: The Accelerated Dragon computer move!) 28 ... Ng6 29 Bb7 Qb8 30 d6 Kg7 31 Be4 Qc8 32 Bxg6 Kxg6 33 dxe7 Rde8 34 Rxd7 when the exposed king and mighty passed pawn on e7 mean that Black is unlikely to survive for long. 2S ... gxf3 26 Nb6

26 gxf3 Rdg8+ followed by ... Rg3 gives Black good attacking prospects, and if 27 Rg2?! there is 27 ... Rxg2+ 28 Bxg2 Qg8!. The text leads to a tricky endgame with mutual chances. 26...Qc7 27 Qxc7 Bxc7 28 gxf3 Rhg8+ 29 Rg2 Nf4 30 Rxg8 Rxg8+ 31 Kf2 h4! (Diagram 52)

Fixing White's weak h-pawn while securing an outpost on g3. White enjoys the potentially decisive advantage of an extra passed pawn, but in the mean time he has to worry about his fragile kingside.

0

NOTE: This is a good example of a middlegame initiative persisting deep into the endgame, without being dampened by a queen exchange.

32 a4f5!?

Blacks light-squared bishop was not doing much on c6, so Caba takes steps to erode the restrictive white pawn wedge. 33 Bxf4 Bxf4 34 Bbs

White hurries to exchange the soon to be liberated bishop. 34...fxe4 (Diagram 53)

Diagram 53 {W)

Diagram 54 {W)

Still tricky for both sides

White's pawns are weak

35 Bxc6

35 fxe4 Bxe4 36 Rb4 looks tempting, but the game remains unclear after 36 ...Bg3+ (not 36 ... Rg2+?? 37 Kfl) 37 Ke3 Bg2 38 Bxd7 e6, when Black will obtain a powerful passed pawn of his own. 3S ... e3+ 36 Ke2 Rg2+ 37 Kd3 dxc6 38 Nc4 Ra2 39 as es 40 Re1

40 Kc3! was better, intending Kb3 to chase the black rook from the a-file. 40... Ke6 41 Nxe3?

256

Mar6czy Bind: 7 ... Ng4 Both sides have played the endgame quite consistently up to now, but the text is a definite mistake. Instead 41 Kc3 e2 42 Kd3 Kd5 43 Rxe2 Rallooks about level. 41... Rxa5 (Diagram 54) 42 Ke4??

White was clearly worse, but this is just disastrous. 42 ... Ra4+ 43 Kd3 Ra5??

Black could have won easily with 43 ... Ra3+ 44 Ke4 Rxe3+ 45 Rxe3 Bxe3 46 Kxe3 Kd5. 44 Ng2 Bg3 45 Rcl Ra2 46 Rc2 Rxc2 47 Kxc2 Kd5 48 Kd3 KxcS

White should now be able to hold by using his king and knight to blockade the light squares. Black is not helped by the fact that he has the wrong rook's pawn for his bishop. 49 Ne3 Bf2 50 Nfs?!

50 Ng4 would force the bishop back to g3, as 50 ... Bd4 51 Ke4 Kd6 52 f4 draws easily. 50 Nc4 also seems fine. 50 ... Kds 51 Ne7+ Kd6 52 Ng6? Bg3 (Diagram 55)

Diagram 55 (W)

Diagram 56 (B)

White has trapped his own knight!

Preventing ... g6-g5

53 Ke4Ke6

53 ... c5! looks best, as the knight cannot escape. 54 f4!?

This should have been a reasonable drawing attempt. 54... exf4 55 Nxf4+??

A feeble end to a mostly enjoyable game. This leads to an immediate defeat, whereas 55 Ne5 would have kept some hopes alive. 5S ... Bxf4 56 Kxf4 KdS 57 Ke3 Kc4 58 Kf4 Kd3 59 Kg4 cs 60 Kxh4 C4 0-1 It was a pity that the final part was marred by numerous blunders (the players were rated

1931 and 2121 respectively), but the middlegame and early endgame demonstrated a variety of important themes for the ...g6-g5 variation.

257

Starting Out: The Accelerated Dragon Finally we move on to what must be currently be viewed as the last word in the ...g6-g5 system. If you play the Mar6czy Bind as White you should definitely pay close attention to White's example, while if you play Black then an improvement will be needed. Game 31

D J.llczuk •

L.Kristol

Correspondence 2003

1 e4 C5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 Ng4 8 Qxg4 Nxd4 9 Qd1 Ne6 10 Rc1! Qa5 11 Bd3! Here we will see the best way for White to combat the ... g6-g5 plan. The first point is that the bishop should come to d3 rather than e2, when the added protection of e4 makes f2-f4 a strong possibility. u ... g5!? This may appear premature, but if Black does not play this now or on the following move he risks missing the opportunity for ...g6-g5 altogether. After the standard 11...b6 White can react strongly with 12 0-0 Bb7 (12 ... g5 reaches the main game) 13 f4! (Diagram 56), when Black's hopes of ... g6-g5 and ...Qe5 or ... Be5 have been well and truly dashed. TIP: Here we see the real value of the bishop's development to d3 instead of e2. In Diagram 56 the additional protection of the e4-pawn means that Black is no longer threatening ... Bxc3 followed by .•. Bxe4.

After 13 f4! Black has tried a few different ideas, none of which are particularly inspiring: a) 13 ... Nc5 14 Bb1 Bxc3?! 15 Rxc3, when taking on e4 loses due to the forking Qd4. b) 13 ...Bd4 14 Bxd4 Nxd415 Nd5 Bxd5 16 cxd5 b517 Bb1 Qb618 Kh1left the d4-knight very awkwardly placed in J.Horvath-S.Conquest, Budapest 1987. c) 13 ... 0-0 may be best, but after 14 Bb1 d615 Rf2 Rac8 16 Nd5 Bxd5 17 exd5 Nc5 18 a3 Na4 19 b4 Qa6 20 Qb3 White was clearly better in A.Rodriguez Cespedes-R.Hernandez, Cuban Ch. 1988. 12 0-0 It may seem tempting to try and exploit the 'premature' kingside advance, but after 12 h4 h6 Black's position is solid enough. E.Prie-F.Berend, Val Thorens 1989, was certainly nothing special for White after 13 hxg5 hxg514 Rxh8+ Bxh8 15 Qh5 Qe516 g3 b617 Qh7 Qg718 Qxg7 Bxg719 f4 Bb7. 12... b6 Black needs to catch up with queenside development. 13 Bd2! (Diagram 57) 13 f3 Bb714 Rf2 h515 Bfl would reach Game 30, but the text is even better. 13 ... Qe5 14 b4! Planning Nd5. White has benefited enormously by avoiding the superfluous and timewasting f2-f3 and Qd2. Instead, all of his moves have been direct and purposeful, and the rewards will speak for themselves. 14... Bb7 15 Nd5 Qd6

258

Mar6czy Bind: 7 ... Ng4 15 ...1<£816 Bc3 Nd4 was N.Lakos-S.Diaz Castro, Dos Hermanas 2003, when 17 Bb1 would have left the d4-knight in a very awkward pin. 16 c5!

Diagram 57 (B)

Diagram 58 (B)

Refuting the ...g6-gS plan?

A crushing attack

16 Qh5!? was a good alternative, but it looks even better to open a queenside file. 16 ... bxc5 17 bxc5 Qc6 17... Nxc5 is refuted by 18 Rxc5! Qxc519 Bb4 Qc8 (or 19... Qc6 20 Bb5!) 20 Nxe7 Qc7 21 Bb5! (Diagram 58), e.g. 2l...Be5 22 Nd5 Qd8 23 Qg4 intending QfS with a massive attack. 18 Qb3 a6 Preventing Bb5. 19 Be3! (Diagram 59)

Diagram 59 (B)

Diagram 60 (B)

Major difficulties for Black

Total domination

259

White continues to improve his position. Now Rbl is threatened, as the reply ... Nxc5 is no longer available. To put it mildly, this is not the sort of position Black should be aiming for with ...g6-g5. 19 ... Bc8 19... Nd4 20 Qb4 does not help. 20 Rb1 Bes 21 Qa3! An excellent move, clearing the b-file for the rook while setting up veiled threats against the sensitive e7-square. 21 ... Ra7 22 Rb6 Qa8 23 Bc4 aS Black can hardly move a piece. 23 ...Bb7? is no good on account of 24 c6 threatening mate on e7. 24 Rfb1 Nd8 25 Bxgs 1-0 (Diagram 60) Black was obviously unable to stand the misery any longer. In Diagram 60 there is no immediate mate, but Black's position is so dreadful that the end would surely not be long in coming.

Possible Solutions? If, like myself, you like the look of ...g6-g5 but are put off by the brutal demolition of the previous game, then I can suggest two possible solutions.

1) Play it anyway! A search on my database reveals that, in well over a thousand games featuring 10... Qa5 (after both 10 Qd2 and 10 Rcl), only four of them ended up reaching 13 Bd2! in Game 31. I have employed the 8 ... Ne6 variation perhaps a dozen times and have yet to face the best move order (most of my opponents have followed in Karpov's footsteps from Game 29; others have allowed an early ... Bxc3). So, while a purist might justifiably claim that the ... g6g5 plan is not technically sound, it seems fair to assume that you are unlikely to encounter many opponents who know the correct response. 2) Play it against 9 Qd2, and have a backup system against 9 Rc1. There are some players who cannot stand to employ a plan which deep down they know to be flawed. If that description applies to you then I would suggest using the ... g6-g5 plan selectively. Game 29 and its annotations indicate that ... g6-g5 is in pretty good shape against an early Qd2, so if you encounter this on move 9 you have nothing to fear. If your opponent plays 9 Rcl, you can revert to the more solid plan involving short castling. In that case I would recommend developing the light-squared bishop on d7 rather than b7, so that in the event of b2-b4 you can play ... a7-a5 to open the a-file without leaving a weak pawn onb6.

Summary and Conclusions 7... Ng4 should be considered by Accelerated Dragon players as a serious alternative to the Classical and Gurgenidze Systems. Black achieves an early exchange of minor pieces while

Mar6czy Bind: 7 ... Ng4 steering the game towards positions of quite a unique character for the Maroczy system. After the practically forced 8 Qxg4, we began the chapter by examining the rather suspicious 8... Bxd4?!. This has rightly acquired a dubious reputation, but could perhaps be used as an occasional surprise weapon. 8... Nxd4 is much more reliable, when 9 Qd1 reaches a major crossroads. In the second part of the chapter we investigated the committal and provocative move 9... es!?. Optically this looks like a gross positional error, but so far it has defied all attempts at a direct refutation. My advice for White players would be to opt for the solid 10 Bd3! leading to a slight but relatively risk-free advantage, rather than the sharper 10 NbS!?. Finally there is the main line of 9... Ne6. This avoids the chronic structural damage associated with the advance of thee-pawn, while still maintaining some unique prospects for counterplay. At this point it is imperative for White to safeguard his queenside structure against ... Bxc3, and to this end 10 Rc1! is the most accurate move for reasons that have been discussed. Black's most interesting method of obtaining counterplay involves Larsen's ingenious plan of ... QaS combined with a subsequent ... g6-g5 to take control of the dark squares. My personal experience of playing the black side of this line has been very positive, and I have no hesitation in recommending it for general use. Unfortunately Game 31 demonstrates a very effective way to combat Larsen's plan, and I would urge White players to familiarize themselves with that game. From Black's point of view it is fortunate that very few opponents seem to be aware of the correct path, and in the majority of games the chances of facing it are minimal. I have offered a couple of possible approaches for Black to work around this problem.

261

Chapter Nine

Mar6czy Bind: Avoiding Exchanges with Nc2

-

Basic Theory and Early Deviations

-

Playing Dynamically with ••• f7-f5

-

The Main Line with 8••• Nd7 9 Bd2

-

Delaying Nc2

-


Mar6czy Bind: Avoiding Exchanges with Nc2 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Nc2!? {Diagram 1)

Diagram 1 (B)

Diagram 2 {W)

Avoiding minor piece exchanges

Unusual, but perhaps underrated

This variation has become very popular since receiving the endorsement of GM Nielsen in Experts vs. the Sicilian. Diagram 1 shows the Nc2 variation in its purest form, though it can also arise from the move order 5 ... Nf6 6 Nc3 d6 (perhaps hoping to employ the Gurgenidze with ... Nxd4) 7 Nc2 Bg7 8 Be2, transposing to the main line below. The reasoning behind White's system is crystal clear: he controls more space, so it is logical to assume that the avoidance of exchanges will benefit him.

0

NOTE: If the d4-knight is going to move somewhere then c2 is clearly the best available destination. It conveniently covers the important b4 and d4squares, and may later be deployed on e3 or, in the event of a subsequent ... a7-a5 from Black, to a3 and b5.

Although moves like 6 Nb3 and 6 Nb5 are occasionally seen, they do not provide the same quality of future prospects for the knight.

Basic Theory and Early Deviations 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Nc2 d6

The most flexible move, though 6 ... Nf6 is perfectly fine and will almost always transpose to the main line after 7 Nc3 d6 8 Be2 etc. 7 Be2

7 Nc3?! would needlessly present Black with the option of 7... Bxc3+! 8 bxc3 Nf6. (This should be compared with the note to White's 9th in the main line; in the present position Black can accelerate his development by doing away with ... Nd7-c5.) 7... Nf6

This is usual, but players who enjoy dynamic positions should give serious consideration to alternatives based on an early ... f7-f5. This plan can be found in the following section. 8 Nc3 Nd7

The main line. Otherwise Black can develop conventionally with 8 ... 0-0 9 0-0 Be6!? {Dia-

263

Starting Out: The Accelerated Dragon gram 2). [This is more active than 9 ... Bd7, while 9... Nd7 reaches positions similar to the main line, except that the possibility of ... Bxc3 is slightly less attractive after Black has committed his king. 9 ... a6!? may also be considered.] It is strange that most sources have paid little attention to this logical development. White can probably maintain a slight edge, but he must be very careful as one small slip could lead to serious difficulties on the queenside. WARNING: White must always be on the lookout for ideas such as ... ReS and ... Nas, attacking c4, when the defensive move b2-b3 will leave White open to tactics involving the long diagonal.

Possible continuations include: a) 10 Ne3 does not work out too well for White after 10... Rc8 11 Ned5 Na5! 12 Qa4 (not 12 b3? Nxe4) 12... Nxd5 (12 ... b6!?) 13 Nxd5 Bd7, when 14 Nxe7+ Qxe7 15 Qxa5 Qxe4 or 14 Qb4 e6 15 Nc3 Qc7looks very nice for Black. b) 10 Be3 is better, continuing development. Then 10 ... Rc8 11 Rcl (11 f3 Na5! 12 b3 Ng4 13 Bd4 Bxd4+ 14 Qxd4 Nc6 15 Qd2 Qb6+ 16 Kh1 Nf2+ 17 Rxf2 Qxf2left Black an exchange to the good in T.Schmitz-G.Kreuzer, German Junior Ch. 1992) 11 ... a6 (11 ... Nd7 and 1l...Qa5 can also be considered) 12 Qd2 Ng4 13 Bxg4 Bxg4 14 Bh6 was slightly better for White in J.Lechtynsky-P.David, Czech League 1997. I suspect that White should be able to maintain a slight edge, but this is true of several main lines as well, so 9 ...Be6 could be worth considering, if only for practical reasons; i.e. that anyone who plays the Nc2 system will probably have spent most of their study time concentrating on the main lines with ... Nf6-d7. Returning to the most popular move 8... Nd7 (Diagram 3):

Diagram 3 (W)

Diagram 4 (W}

A key manoeuvre

A double-edged position

This is a highly thematic manoeuvre for Black in the Nc2 variation. The knight heads for the excellent c5-square and opens the h8-a1 diagonal. Furthermore, the capture ... Bxc3 becomes a serious possibility, giving up the powerful bishop in order to wreck White's queenside structure. 264

Mar6czy Bind: Avoiding Exchanges with Nc2 TIP: The obvious drawback of ... Bxc3 is that Black may suffer from dark square problems on the kingside. Therefore the idea becomes all the more potent when the black king is not yet committed to short castling.

Black quite often prefers the move order 9...0-0 10 0-0 Nd7, but it seems to me that the text is slightly more accurate for the reason highlighted above. 9 0-0!?

This is unusual and somewhat provocative. I do not regard 9 0-0 as particularly promising, but that's not to say that Black will necessarily have an easy time. 9 Bd2 is the main move and will be considered in a later section. 9... Bxc31

9... 0-0 is possible, but the text is clearly the most principled and, in my opinion, the strongest continuation. 10 bxc3 Ncs (Diagram 4)

We are set for an interesting and unbalanced middlegame. White's queenside structure is in ruins, but if he succeeds in opening things up for his bishops then Black may not survive as far as his desired endgame. Possible continuations include: a) In E.Bacrot-S.Pujos, French Team Ch. 2003, White offered thee-pawn with 11 Bh6!?, presumably intending something like ll...Nxe4 12 Bf3 when his two bishops on an open board would provide strong compensation for the pawn. Instead, Black played more solidly with 11...£6! 12 Nd4 Ne5 13 Qc2 Bd7, which led to a complex position and an eventual draw. b) White preferred the more natural11 f3, in M.Stangl-V.Korchnoi, Garmisch Partenkirchen (rapid) 1994, and was successful after ll ...Qa5 12 Qel!? Be613 Bh6 Na4? 14 Bg7 Rg8 15 Bd4 0-0-0 16 Nb4 Kb8 17 f4 Nxd4 18 cxd4 when it was obvious that something had gone seriously wrong for Black. Nielsen and Hansen believe that his 12th was the mistake, and offer 12... Na4 13 Bd2 Be6 as an improvement. However, it seems to me that it was on the following move when things started to turn pear-shaped, and if the great veteran had played 13 ... Ne5! 14 Ne3 (14 Bg7 Rg815 Bxe5 dxe5 can't be good for White) 14.. .£6 then he would have obtained a full share of the chances.

Theoretical? None of these early deviations require much theoretical knowledge, and are mostly played with the express intention of avoiding the more heavily analysed variations.

Statistics I was surprised to see 9 0-0 Bxc3 10 bxc3 Nc5 scoring as little as 30% for Black, though this could just be a statistical anomaly caused by a very small sample size (just 5 games on my database). After 8...0-0 9 0-0, 9... Be6 has yielded a slightly disappointing 31%.

Playing Dynamically with ...f7-f5 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Nc2

Now 6... d6 7 Be2 Nf6 is the main line, but it is possible to try and 'exploit' White's early knight retreat with the active ...

265

Starting Out: The Accelerated Dragon 6... Nh6!? (Diagram 5)

.. .intending to strike at the centre with ... f7-f5!?.

Diagram 5 (W} Aiming for .. .f7-f5

Diagram 6 (B) Fascinating and highly irregular

For those interested in playing this line as Black, it is good to know that 6... Nh6!? has been largely neglected by most other texts, so it could very well carry a degree of surprise value. 7 Be2

This is the most natural move, so I will consider it as the main line. Instead, 7 Bd2!? has occasionally been seen, planning a possible Bc3 while taking advantage of the fact that ... Bxb2 would hang the knight on h6. After the natural 7... 0-0 (7 .. .f5 8 exf5 Nxf5 9 Bc3 Qb6 10 Qd2 was S.Voitsekhovsky-L.Aronian, Minsk 1998, and now 10... 0-0 would have been best with an unclear position; all the same, castling immediately looks more logical), White can choose between: a) 8 Bc3?! d6 9 Bxg7 Kxg7leads to something resembling the Classical Mar6czy from Chapter Six, except that Black has saved time while also achieving the positionally desirable exchange of dark-squared bishops. b) 8 Qcl Ng4 9 h3 Nf610 Nc3 d6 11 Be2led to a somewhat irregular Mar6czy position in R.Fuchs-H.Platz, East German Ch. 1953, but I think Black should have taken the opportunity for 10... e6! intending a quick ... d7-d5. c) 8 h4!? (Diagram 6} led to a bizarre-looking position in J.Barlow-V.Aleshnia, Correspondence 2000, which continued 8 .. .f5 9 Nc3 d6 10 h5 N£711 hxg6 hxg6 12 exf5 Bxf5 13 Ne3 Qd714 Nxf5 Qxf5 15 Qb1 Qe6+ 16 Qe4 Qxe4+ 17 Nxe4 Bxb2 18 Rb1 Bg719 Rxb7 Rab8 20 Rb5 a6 21 Rb3 Rxb3 22 axb3 Rb8 23 Rh3 Nd4 24 c5 Nxb3 25 cxd6 Nxd2 26 Nxd2 Nxd6 1/2-1/2. 7 ...0-0

7... d6 is also possible, although White may then consider 8 g4!? as recommended by Donaldson and Silman. I am not convinced that this is as bad for Black as they suggest, but at the same time it seems logical to avoid the whole issue with 7...0-0. 8 0-0

8 g4!? could be tried here, too, though in this position Black is better placed to hit back with

266

Mar6czy Bind: Avoiding Exchanges with Nc2 8.. .f5!. V.Ciocaltea-D.Drimer, Rumanian Ch. 1961, continued 9 exf5 gxf510 g5 Nf7 with an unclear but roughly equal position; 10... Ng4!? also deserved consideration. Perhaps a better try for White was 9 gxf5 gxf5 10 Rg1 Kh8, with a highly unclear position in which neither king is likely to feel completely safe. 8... d6

The immediate 8... f5!? could also be investigated. 9 Nc3 fs (Diagram 7)

Diagram 7 (W)

Diagram 8 (B)

A successful opening for Black

Offering a poisoned pawn

Now the pawn advance ... f5-f4 constitutes something of a positional threat, securing an outpost on e5 while cramping the white kingside, and perhaps planning a future attack with ... g5-g4. Therefore a number of games have continued 10 exf5 Nxf5 11 Ne3 Nfd4, with unclear but approximately level chances. There are many ways in which the game may develop, but if you are the type of player who enjoys a creative and dynamic game, without having to learn much theory, then this could be the ideal weapon against 6 Nc2. NOTE: Just a final reminder that this option is only available via the move order s... Bg7 6 Nc2. Fans of the Gurgenidze who employ the move order s... Nf6 6 Nc3 d6 will need to look elsewhere against 7 Nc2.

Theoretical? One of the attractions of this variation is that very little theory exists.

Statistics 6... d6 7 Be2 f5 scores a respectable 48% for Black, while 6... Nh6!? achieves an even more encouraging 52%, a figure which rises to 57% after the common sequence 7 Be2 0-0 8 0-0 d6 9 Nc3 f5.

The Main Line with 8 ... Nd7 9 Bd2 267

Starting Out: The Accelerated Dragon 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Nc2 d6 7 Be2 Nf6 8 Nc3 Nd7 9 Bd2

Guarding against ... Bxc3. 9...0-o 10 o-o Ncs

In view of White's next, some players have preferred 10 ... a5!?, increasing Black's control of b4 at the cost of weakening bS. This approach can be found in Game 32. 11 b4! (Diagram 8)

This active move has superseded the older 11 f3, since it became clear that accepting the pawn sacrifice brings great risk to Black. Besides, after 11 f3 Black seems to have a good answer in ll ... Qb6!?, when R.KerndlC.Lund, Correspondence 1998, continued 12 Kh1 Qxb213 Rb1 Bxc314 Rxb2 Bxb2 and Black enjoyed full material equality with a rock-solid position in return for the queen. Another possibility is 12 NdS, but here Nielsen and Hansen give 12...Qxb2 13 Rb1 Qxa2 14 Ncb4 Nxb4 15 Nxe7+ Kh8 16 Bxb4 Be6 with an edge to Black. Otherwise, Black can settle for the solid 1l...a5, leading to very similar positions to those found in Game 32. 11... Ne6

This is the usual move, though a very interesting alternative, ll ... Nd7!?, can be found in Game 34. I would strongly advise even the most materialistic reader to resist the temptation to bag a pawn with 1l...Bxc3 12 Bxc3 Nxe4.

0

NOTE: ... Bxc3 is often a worthwhile idea to cripple White's queenside pawns, but seems far less attractive when White can recapture with a piece, thus maintaining the fluidity of his queenside, even if it means sacrificing a valuable central pawn.

The above point is vividly demonstrated after the vigorous follow-up 13 Bb2 Be614 bS! (Diagram 9).

Diagram 9 (B) Massive compensation for a pawn


White has more than enough for a pawn, with two strong bishops and Qd4 on the way. In

268

Mar6czy Bind: Avoiding Exchanges with Nc2 P.Svidler-S.Tiviakov, European Cup 2002, things went from bad to worse for Black after 14 ... Ne5? 15 Qd4 Nf6 16 f4 Ned7 17 g4! when White was able to win a piece thanks to the dual threats of f4-f5 and g4-g5, and was soon victorious. Nielsen mentions a couple of possible improvements on moves 13 and 14, but the whole variation still looks decidedly dodgy for Black. After 11...Ne6, White nearly always takes the sensible precaution of removing his queen's rook from the long diagonal. But which square should it go to- b1 or c1? 12 Rb1 has the advantage of providing additional support to the b4-pawn, but in the last few years ... 12 Rc1 ...has been all the rage, thanks in no small part to its advocation by Nielsen in Experts vs. the Sicilian. The main idea is to provide additional support to the knight on c3. 12 ...asl? Now that the white rook has deserted the a-file it is logical for Black to turn it into an avenue for counterplay. Nevertheless, this approach carries certain risks in that the b5 and b6squares are irreparably weakened. The more solid 12 ... Ned4 can be found in Game 33.

13 a3 13 b5?! would be a mistake. True, it may enable White to keep the a-file closed, but the immobilization of the queenside pawns and the ceding of the c5-square are a high price to pay. Following 13 ... Ncd4 14 Nxd4 Nxd4 15 Be3 e5 Black was fine in I.Balinov-M.Ivanov, Schwarzach 2002. 13 ... axb4 14 axb4 Ned4 1S Nxd4 Nxd4 16 Be3 (Diagram 10) White should be a little better here due to his space advantage and active pieces. Now a number of games have seen 16... Nxe2+ 17 Qxe2, but this type of position should be quite enjoyable for White, as he can play for a win without taking too many risks. Sometimes Black will force an opposite bishop position with a timely ... Bxc3, but this will still leave him with passive pieces and a slightly vulnerable kingside. More interesting is: 16 ...es!? This secures the strong knight on d4, albeit at the price of blocking the g7-bishop.

0

NOTE: It hardly needs stating that White should definitely not consider 17 Bxd4 here. This may damage the black pawn structure, but the loss of his valuable bishop combined with the improvement of Black's own bishop make it a bad idea.

The critical continuation must be ... 17 NbS!? White offers to concede doubled pawns of his own, but it's worth it to remove the powerful enemy knight without making the concessions noted above.

0

NOTE: This position bears an obvious resemblance to the line s ... Bg7 6 Be3 Ng4 7 Qxg4 Nxd4 8 Qd1 es!? 9 NbS from Chapter Eight (page 230).

11 ... Nxbs 17... Ra2!? was seen in E.Geller-D.Velimirovic, European Team Ch., Skara 1980, and here Nielsen suggests the logical18 Re1 in order to renew the threat of Nxd4.

269

Starting Out: The Accelerated Dragon 18 cxbs Be6 (Diagram 11)

Diagram 11 (W)

Diagram 12 (W)

A critical position

A novelty

19 b6!? An improvement over 19 Bc4 Qd7 20 Qd3 Rfc8 which was agreed drawn in E.GellerE.Pigusov, Sochi 1989. 19 .. Js! This seems to be best. A third Geller game (E.Geller-E.Pigusov, Cappelle la Grande 1992) continued 20 f3 Ra3 21 Qd2 Ra2 22 Rc2 Rxc2 23 Qxc2 Qd7 24 Rcl f4 25 Bf2 ReB 26 Qxc8+ Qxc8 27 Rxc8+ Bxc8 28 Bh4 Kf7 1/2-1/2. Instead, Nielsen offers 20 Bc4!? Bxc4 21 Rxc4 f4 22 Qd5+ Kh8 23 Bd2 Qxb6 24 Rfcl with good compensation. This seems like a fair assessment, although practical testing would be helpful and White must always be on the lookout for a timely .. .f4-f3.

Before moving on to some illustrative games, I wish to share an original and untested suggestion which may be worth considering: 14... Nf4!? (Diagram 12) The credit for discovering 14... Nf4!? goes to Jovanka Houska, who suggested it in 2006 while we were analysing this variation together. If White does not capture on f4, then Black will capture on e2, resulting in a position resembling 14... Ned415 Nxd4 Nxd4 16 Be3 Nxe2+ 17 Qxe2, except that an extra pair of knights remain on the board. This difference favours Black, as one of the white knights may turn out to be superfluous. Possibilities include 15 Nd5?! Nxe2+ 16 Qxe2 e6 and 15 Nb5 Nxe2+ (15 ...Be5!?) 16 Qxe2 Ra2, with good chances for Black in both cases. Therefore White should prefer ... 15 Bxf4 Bxc3 16 bs 16 Rb1 is met by 16... Be6. 16... Nb4 16... Nb8?! 17 Ne3 Bg718 c5! and 16 ... Na517 Ne3 Bg718 Nd5 are both good for White.

270

Mar6czy Bind: Avoiding Exchanges with Nc2 17 Ne3!

17 Nxb4 Bxb4leaves White's queenside pawns immobile and potentially weak in an endgame, while 17 Bh6 Re8 18 Nd4 Na2!? looks rather unclear. 17... Bg7 (Diagram 13)

Diagram 13 {W)

Diagram 14 {W)

An unusual position

Preventing b2-b4, but weakening bS

It is hard to give a definite evaluation of this position. White will probably play Nd5 in the next few moves, when the response ... Nxd5, c4xd5 will open the c-file. If White can arrange b5-b6 followed by Rc7 then he will stand better, so Black needs to mobilize his forces in time to obstruct that plan. A couple of possibilities: a) 18 Qb3 Qa5 19 Nd5 Nxd5 20 cxd5 Bd7 21 b6 Rfc8 seems ok for Black. b) 18 NdS Nxd5 19 cxd5 Ra3!? (19 ...Bd7 is also possible) 20 Qd2 (20 Be3 Rxe3! 21 fxe3 Qb6 22 Qd2 Bd7looks like a promising exchange sacrifice) 20 ...Bd7 (20 ...Qb6 21 Be3 Rxe3? is insufficient after 22 Qxe3 Qxe3 23 fxe3- the exchange of queens is a big help to White) 21 Be3 Qa8 22 Rc7 Ra2 with some counterplay. It is hard to say whether 14... Nf4!? is enough for full equality. White should be able to maintain a slight edge somehow, but the positions certainly appear playable for Black and one should never underestimate the value of surprise.

Illustrative Games Game 32

D L.Van Wely •

J.Van der Wiel

Dutch Championship, Leeuwarden 2004

1 Nf3 cS 2 C4 Nf6 3 Nc3 g6 4 d4 cxd4 5 Nxd4 Nc6 6 Nc2 Bg7 7 e4 d6 8 Be2 Nd7 9 Bd2 0-0 10 o-o aS!? (Diagram 14)

The text is a solid move which prevents b2-b4, thereby creating a more secure home for a knight on c5. The disadvantage is that the b5-square is permanently weakened. Nielsen

271

Starting Out: The Accelerated Dragon also makes a good point when he states that once ... a7-aS has been played White is in a better position to move his dark-squared bishop away from d2, as a subsequent ...Bxc3 will become less effective, for two reasons: Firstly, the black queen will be deprived of the aSsquare, and secondly Black may become weak along the b-file. 11 Re1

A logical move, intending to meet ... NcS with Bfl conveniently protecting the e4-pawn. According to my database this has been the most popular move here. Bizarrely it is not even mentioned by ADP in their repertoire book, although the authors do partially redeem themselves with some worthwhile suggestions in some of the other lines: a) 11 Rb1 is a solid but slightly slow approach, so Black may be justified in fighting for the initiative with 1l...NcS12 f3 fS!?; e.g. 13 Kh1 (13 exfS?! BxfS14 NdS Nd4! 1S Nxd4 Bxd4+ saw Black bag an exchange in S.Bromberger-F.Grafl, Augsburg 2002) 13 .. .fxe414 Nxe4 (M.Magomedov-S.Vokarev, Moscow 2002, and now ADP suggest 14...BfS! 1S Ng3 Bxc216 Qxc2 Nd4 when Black is very active. b) 11 Na3!? is recommended by Nielsen, so must be taken seriously. Following ll ... NcS 12 NabS Black should occupy an important central square with 12... Nd4!. Now after the natural13 Nxd4 Bxd4 14 Bh6, the passive 14 ... Bg7?! 1S Bxg7 Kxg7 16 Bg4! Bxg4 17 Qxg4 gave White a pleasant advantage in V.Bologan-A.Motylev, Togliatti 2003. Therefore Black should prefer the principled continuation 14... Bxc3! 1S bxc3 (1S Bxf8 Bxb2 16 Rb1 Bf6 17 Bh6 Nxe4 is better for Black- ADP) 1S ... Re8 (Diagram 15) with a very interesting and strategically unbalanced position. ADP evaluate it as favourable to Black, while Nielsen prefers White after 16 f3, adding that White should aim for a slow build-up with moves like Qd2 and Rabl. I believe that the truth lies somewhere in the middle. Both sides have chances and it is probably just a matter of personal taste as to whether one prefers White or Black. (I should add that the only practical test that I could find- N.Zadruzny-R.Edouard, European Junior Ch. 200S- ended in a draw after 37 moves.)

Diagram 15 (W)

Diagram 16 (W)

A tough position to evaluate

Fighting for the initiative

11 ... Nc5 12 Bf1 b6!?

Slightly unusual, though it looks quite sensible to me. Lots of other moves have been tried,

272

Mar6czy Bind: Avoiding Exchanges with Nc2 but in most cases the general character of the position will stay roughly the same. One notable exception is the active 12 .. .£5!? (Diagram 16). Black's knights are very harmoniously placed, so this aggressive action seems quite justified. In S.Dvoirys-K.Van der Weide, Dieren 2000, White soon got into trouble and lost an exchange after 13 exf5 Bxf5 14 Be3?! Nb4! 15 Nd4 Bxd4 16 Qxd4 Nc2, although he later managed to draw. Food for thought! 13 Na3! White often plays this thematic move at one time or another, with the intention of planting the knight on b5. Black continues to develop solidly, but falls slightly short of equality. 13 ... Bb7 14 Rc1 Rc81S BgS! (Diagram 17)

Diagram 17 {B)

Diagram 18 {W)

White has developed harmoniously

Ingenious, but insufficient

I like this move; White improves his bishop and hopes to provoke a weakening pawn move on the kingside. 1S ... Nd416 Qd2 ReS 17 f3 Nce6 18 Be3 fS 19 NdS! A good decision; there is no need to get excited about the advance of the £-pawn here. Instead White turns his attention to the principal weakness in Black's position: the b6-pawn. Note that Van Wely has also refrained from placing the a3-knight on b5, as he prefers to keep the b-file clear for a rook. 19 .. .fxe4 20 fxe4 Rf8 21 Rc3! Continuing the plan - the threat is Rd3 to chase the knight away, followed by Rb3. 21 ... Rf3!? (Diagram 18} An amazing move! Vander Wiel sees what is coming and plans to counter with an exchange sacrifice on e3. The idea receives full marks for creativity, even though his compensation is probably insufficient. Objectively, he probably should have preferred Mikhalevski's suggestion of 2l...Nc6! 22 Rb3 Nc5 23 Rxb6 Rb8! with some compensation as White's pieces are not too well coordinated and the rook on b6 is unstable. 22 Kh1 Rxe3!

273

Starting Out: The Accelerated Dragon Continuing the plan. It was too late to back out with 22 ... Rf7?, as 23 Rd3 would win a pawn for no compensation whatsoever. 23 Qxe3 Nes 24 Reel e6 25 Ne3 Qh4

Mikhalevski suggests 25 ... Be5!? and if 26 Rcdl Q£6, which does indeed look quite attractive for Black. But I believe that White can do better with 26 NabS!, removing the powerful knight. After 26 ... Nxb5 27 cxb5 White's pawn structure has been compromised slightly, but this is of no great consequence and he can improve his position with g2-g3 and Bg2 to defend e4, followed by Rfl to take the £-file. 26 Redl Rf8 27 g3 Qf6 28 Bg2 Qf2 29 Qel! (Diagram 19)

Diagram 19 (B)

Diagram 20 (B)

Skilful consolidation

White can win the ending

29 Qxf2?! Rxf2 would leave the black rook extremely well placed. 29 Qcl may seem passive, but Van Wely correctly judges that Black's apparent activity is an illusion, and in reality his pieces will soon be driven back. The immediate threat is Rfl embarrassing the queen. 29 ... Qf6 30 Qe3 Qf2 31 Qel Qf6 32 Rfl!

After repeating once, presumably to gain time on the clock, White improves his position. 32 ... Qes 33 Rxf8+ Kxf8

33 ... Bxf8? just misplaces the bishop and White continues 34 Q£4 anyway. 34 Qf4+ Qxf4 35 gxf4 (Diagram 20)

White's technical task should not be too difficult from here, although Black can resist for a while. 3S ... Ke7 36 Kgl Ne6?

This does not help; indeed, as Mikhalevski points out, the knight on d4 was one of Black's best pieces and it certainly doesn't help him to remove it voluntarily. 36 ... e5 would have been more stubborn. 37 Nabs! Ba6?!

Losing almost immediately, though there was not much hope anyway by this stage. 38 es!

274

Mar6czy Bind: Avoiding Exchanges with Nc2 Winning further material. The last few moves can pass without comment. 38 ... Nb4 39 Rxd6 Bh6 40 Ne2 Nbd3 41 Rxb6 Nxf4 42 Nxf4 Bxf4 43 Rc6 Bxb5 44 cxb5 Bxe5 45 Kf1 Nd7 46 Bh3 Bxh2 47 Rxe6+ Kd8 48 Ra6 1-0

Despite this reversal the plan of stabilizing the cS-knight with lO ... aS appears to be quite playable. In particular, the active plan with 12.. .f5!? may be worth investigating. Game 33

D E.Bacrot •

S.Tiviakov

Wijk aan Zee 2006 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Nf6 6 Nc3 d6 7 Nc2 Bg7 8 Be2 Nd7 9 Bd2 Nc5 10

o-o o-o 11 b4 Ne6 12 Rc1 Ned4 Earlier in the chapter we looked at 12... a5!? which is more active (just ask the a8-rook!), but on the negative side bS- and b6-squares were permanently weakened. In the present game Tiviakov opts for solidity. 13 Nxd4 Nxd4 14 Be3 Nxe2

Compared with the analogous position after 12... a5, it is worth noting that 14 ... e5?! looks highly dubious here in view of 15 NbS! NxbS 16 cxbS, when the white b-pawns are less weak and the black rook on a8 is passively placed. M.Matthiesen-T.Jaksland, Danish Ch. 2007, continued 16... Be6 17 Bc4 Qd718 a4 Rfd8 19 BdS h6 20 Qd2 Kh7 21 Rc3 with a huge positional advantage for White. 15 Qxe2 {Diagram 21)

Diagram 21 (B)

Diagram 22 (B)

Black is solid but passive

b6 is a long-term weakness

This is an important position for the evaluation of the Nc2 variation. There can be little doubt that White is slightly preferable; the question is whether it is enough to put Black under any significant pressure. In the present game Bacrot makes quite a convincing case for White, and I would certainly advise fans of the Nc2 system to pay close attention to the

275

Starting Out: The Accelerated Dragon way in which he increases his advantage. 1S ... Be6 This is certainly a logical move, developing a piece and pressurizing the pawn on c4. a) 15 ... Bxc3? is too passive and after 16 Rxc3 a517 b5 f618 c5! dxc5 19 Rd1 Qe8 20 Bh6 R£7 21 Rcd3 Be6 22 Qd2 Black found himself in a dreadful mess in J.Nogueiras-I.Zugic, Havana 2004. b) 15 ... b6 16 Rfd1 Bb717 Bd4! gave White a slight but stable advantage in P.H.NielsenK.Lie, Drammen 2005. It is quite helpful for White to exchange the dark-squared bishops here, as this will weaken the black kingside while depriving him of the bishop pair. Nielsen later won by centralizing his pieces and playing a well-timed Nd5, leading to a favourable major piece ending which was not dissimilar to the Larsen game in the introduction to Chapter Six. 16 Nds

Preventing any ... Bxc3 ideas which can lead to drawish positions, albeit slightly worse for Black. A top GM with excellent defensive technique might be content to defend such a position to achieve a draw as Black against a formidable rival, though I would not recommend the reader to play in this fashion. 16 ... Qd7

16... Bxd517 cxd5 Qd718 Rc2 Rfc819 Rfcl a6 20 f3looks very pleasant for White. 17 Rfd1 b6?! 17...Bxd5 was a better try. Then 18 cxd5 a6 would leave Black practically a tempo up on the previous note, even if White can still claim an edge. 18 exd5 is also possible, though here 18 ...Qa4!? may offer some chances for counterplay. 1S bs! (Diagram 22) A very important move, immediately latching onto the weakness on b6. Now White can aim for a4-a5, and a knight transfer to c6 could also become a tempting proposition; whereas Black will not be able to make use of the weakened c5-square any time in the foreseeable future. 1s.. .fs! A good practical decision! Passive defence would enable White to arrange a4-a5 at his leisure. 19 exfs Rxfs

19 ... Bxf5 could be met by 20 Bd4 or 20 c5!? bxc5 21 Bxc5 (Roiz). 20Nb4

Aiming for c6. 20 ... Bf7

The exchange sacrifice 20 ... Rc8 21 Nc6 Rxc6 22 bxc6 Qxc6 is refuted by 23 Bxb6! (Roiz). 21 Nc6 ReS 22 Qd3 ReS

White must remain vigilant here, as the exchange sacrifice has become a possibility again. If Black could carry this out in safety then his extra pawn, pair of bishops and superior structure would provide decent compensation, and would certainly represent an improvement over his current situation.

276

Mar6czy Bind: Avoiding Exchanges with Nc2 23Qe41

Bacrot is a world class player and is careful not to present Tiviakov with such an easy opportunity to solve his problems. The text reinforces the knight while attacking e7. 23 ... Rc7 24 a4 (Diagram 23)

Diagram 23 (B)

Diagram 24 (B)

Maintaining the pressure

A pretty finish

Preparing a queenside assault. 24... hs

24 ...Qe6!? is suggested by Roiz, though 2S Qxe6 Bxe6 26 Re1 Bc8 27 aS would still be unpleasant for Black. 25 h3

Also tempting is Roiz's suggestion of 2S cS!? bxcS 26 BxcS Bf6 27 Be3 with a clear advantage. 2s ...Qe6

The white queen was very well placed in the centre, so it makes sense for Black to offer this exchange. 26 Qxe6 Bxe6 27 Rel Bc8

Obviously not 27...1<£7?? in view of 28 Nd8+. Or if 27... Bf7 28 aS bxaS 29 Bxa7, when Roiz gives 29 ... Kf8 30 Bb6 Rb7 31 Bd8 Bf6 32 NxaS or 29 ... Rxa7 30 Nxa7 Bd4 31 Rxe7 Rxf2 32 Kh1 a4 33 Nc6 BcS 34 Nd8, with a White victory in both cases. 28 aS bxaS 29 Bxa7 Kf7 30 Bb6 Rxc61

The best chance. 31 bxc6a4

Black's position is critical, although the strong a-pawn offers him some hope. 32 Redl a3 33 Bd4 Bxd4 34 Rxd4 Be6?

The final error. He had to try 34 ... RcS with the possible continuation 3S Ra1 Rxc6 36 Rxa3 eS, when the weakness of the c4-pawn would keep some drawing chances alive. 35 RdS! (Diagram 24)

277

Starting Out: The Accelerated Dragon A neat way to wrap up the game. The rook is immune from capture, as the c-pawn would become unstoppable (as the reader can quickly verify). 35 ... Ke8

Now White is able to execute a favourable exchange of rooks, leading to an easily winning endgame. 36 Rxf5 gxf5 37 Ra1 Kd8 38 Rxa3 Kc7 39 Re3 Bxc4 40 Rxe7+ Kxc6 41 f3 1-0

A fine technical display from Bacrot, and a convincing demonstration of White's chances in the Nc2 variation. The following example sees Black adopt a radically different approach. Game 34

D Art.Minasian •

V.Malakhov

European Club Cup, Fuegen 2006

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Nf6 6 Nc3 d6 7 Nc2 Bg7 8 Be2 0-0 9 0·0 Nd7 10 Bd2 Nc5 11 b4 Nd7!? (Diagram 25)

Diagram 25 (W)

Diagram 26 (W)

An amazing idea

White has six(!) 'extra tempi'

What is this? Is it a misprint? Did Black really put his knight back on d7 instead of e6? Yes, it's true! One of the world's leading experts on the Accelerated Dragon thinks that White's thematic b2-b4 advance represents a weakness, the provocation of which is worth the sacrifice of two tempi! 12 Rb1 b6 13 f4

A logical move, gaining space and increasing White's influence in the centre and kingside. The only other game I could find with 11...Nd7!? was N.Sedlak-V.Zvjaginsev, Serbian Team Ch. 2005, which continued 13 a4 Bb7 14 Rb3 ReS 15 Bg5 h6 16 Bh4 aS! with good play for Black. It seems to me that an early a2-a4 from White is just asking for ... a7-a5. 13 ... Bb7 14 Be1

Preventing intrusions on d4 and perhaps thinking of relocating the bishop to h4 in the fu278

Mar6czy Bind: Avoiding Exchanges with Nc2 ture, from where it may combine with a knight on d5. 14... Rc8 15 Rf3

This also seems very logical. The rook prepares to take up an aggressive post on h3 while guarding his team-mate on c3 against an x-ray attack from his opposite number on c8. 1S... Ncb8!?

This may seem strange, but after seeing Black's 11th you should not be surprised by anything! Black is intending to position his knights on f6 and d7 in accordance with the wellknown 'Hedgehog' formation. 16 Rh3 Nf6 17 Bd3 Nbd7 (Diagram 26}

The Hedgehog can arise from different openings, most notably the English and certain variations of the Sicilian. The difference is that normally Black's knights will use two tempi in taking up their posts on f6 and d7. In the present position Malakhov has achieved the identical formation after expending eight(!) tempi on knight moves. Yet the surprising fact is that Black's chances still seem to be fully adequate. I should also add that Minasian is an experienced Grandmaster, rated 2596 at the time of the present game. 18 Ne3 a6 19 g4?

Perhaps White was starting to get impatient, but this move clearly does not have the desired effect. 19... h6!

Highlighting the drawback of having a rook on h3- there is no h2-h4 to support g4-g5. 20 Bd2 e6 21 Qe1 gS! (Diagram 27)

Diagram 27 {W)

Diagram 28 {W)

Seizing the initiative

Black is winning

A thematic move, but a beautiful one nevertheless. The idea is to wrest control of the crucial e5-square. TIP: It is worth making a mental note of the positional theme of playing ... g6-g5 (or ...g7-g5) in order to wrest control of es. It can arise in many different openings, most notably a number of Sicilian variations.

279

Starting Out: The Accelerated Dragon 22 fxgs hxgs 23 Nc2?!

Already in a worse position, White pins his hopes on an unsound attack. Better was 23 Qe2 Ne5 24 Rfl when White is clearly worse but the game goes on. 23 ... Nxg4 24 es?! Bxesl

Freeing the g7-square for the king 25 Rhs Bxh2+! 26 Rxh2 Nxh2 27 Kxh2 NeS (Diagram 28)

Black's advantage has already reached decisive proportions. It is not only the extra material (rook and three pawns for two minor pieces), but also the great activity of his pieces, combined with the vulnerability of the white king. 28 Be2 Rxc4!

Make that a rook and four pawns! 29 Kg3

Obviously 29 Bxc4?? Nf3+ is no good. 29 ...Qf6 30 Qgl

This time 30 Bxc4 would have allowed 30 ... Qf3+ 31 Kh2 Qg2 mate. 30... Rxc3+! 0-1 (Diagram 29)

Diagram 29 {W) A neat finish

Diagram 30 {B) The delayed version

White resigned in view of 31 Bxc3 Qf4+ 32 Kh3 Qh4 mate. A beautiful finish to an excellent game. At the time of writing 11...Nd7!? remains relatively unexplored, but it definitely warrants further investigation and could well become more popular in the future.

Theoretica I? Both sides should know a certain amount of theory, but nothing too heavy. Most of the time positional understanding will be more important.

280

Mar6czy Bind: Avoiding Exchanges with Nc2

Statistics 10... a5 scores 33% for Black, while 10... Nc5 yields a slightly better (though still not great) 37%. After 11 b4, the enterprising 11 ... Nd7!? has scored two wins from two, but with so few encounters it is hard to draw any firm conclusions. Overall I think that the main lines of the Nc2 system contain enough interesting resources to make them eminently playable for both sides. Before closing out the chapter we should take a look at a variation which usually arises from a Classical Maroczy move order, but seems more appropriate to be included here.

Delaying Nc2 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd7 10 Nc2!? (Diagram 30)

Delaying Nc2 is not without logic, because in Diagram 30 Black no longer has the standard ... Nd7-c5 manoeuvre at his disposal. If White wishes to play this way, he must still be ready to meet lines such as 7... Ng4, but an argument could certainly be made for avoiding the 'pure' 6 Nc2 systems, which allow Black such interesting options as 6... Nh6!? as well as the main lines involving ... Nd7-c5. Instead White plays traditionally with 6 Be3 but switches to 10 Nc2 in the present position to implement the idea in an arguably more favourable setting. Before moving on, I will briefly mention the related but less potent possibility of 10 Nb3!?. As usual, I don't believe the knight to be as well placed here as on c2. Black's most obvious answer is 10... as!, immediately threatening to harass the knight, and after the usual retreat 11 Nd2 (11 a4?! would permanently immobilize White's queenside pawns) 11... Bc8!? (Diagram 31) looks quite attractive.

Diagram 31 (W)

Diagram 32 (B)

A cheeky but effective retreat

Black has problems

By playing this slightly cheeky move Black is effectively claiming that the white knight is

281

Starting Out: The Accelerated Dragon no better on d2 than d4, and plans to follow with ... Nd7-c5. One high level encounter, V.Korchnoi-V.Anand, Wijk aan Zee 1990, continued 12 Rcl Nd713 Nb3 b6 14 Nd4 Nxd4 15 Bxd4 Bh6!? (an exchange of bishops should also be fine for Black- compare Chapter Six) 16 f4 Bb717 Be3 Nc5 when Black had no problems whatsoever. After 10 Nc2 Black must decide on a plan of action, which will usually involve aiming for ... b7-b5. This is not an unrealistic goal when you consider that White's last move has reduced his control over that square. White's strategy can vary; on some occasions he will aim for restraint on the queenside, preventing counterplay before turning his attention to a kingside assault. On others he may try to expand on the queenside and even take over the initiative there. 10... a6!? This is arguably Black's most flexible move. I have already mentioned that Black's usual plan is to aim for ... b7-b5, so the text is likely to be necessary sooner or later. Alternatives include: a) 10... Rc8 is possible and will usually transpose to another line after a subsequent ...Qa5 and/or ... a7-a6. b) On 10... Qa5 the ambitious space-gaining 11 f4!? deserves serious consideration. Now N.Short-U.Andersson, Wijk aan Zee 1990, continued 1l...Rac812 Rb1 a613 b4!? Qd814 Qd3 Bg4 15 Kh1 Bxe2 16 Qxe2 with an edge to White, while 14 h3!? may have been even stronger, preventing the bishop exchange. The benefits of the Nc2 retreat are all too obvious here, as the black position is noticeably congested. Instead, 1l...Rfc8 is in some ways more logical, when the queen can slot back to d8 and the two rooks make themselves useful on the queenside. Unfortunately, S.Sahu-L.Karlsson, Wrexham 1996, showed a strong approach for White in 12 c5! (Diagram 32). The main idea is that 12 ... dxc5 loses a piece to 13 e5. In the game Black could find nothing better than 12... Rd8, but after 13 cxd6 exd614 Na3 Be615 f5 gxf516 exf5 Qe5 17 Qd3 Bd518 Nxd5 Qxd519 Rad1 Qxd3 20 Bxd3 White had a large advantage. Black may have small improvements along the way, but I have been unable to find anything fully satisfactory. 11 f3 ReS ll ... Rb8 12 Qd2 Be613 Racl Nd714 Rfd1 also leaves White with the usual slight advantage. In I.Smirin-K.Arakhamia Grant, Port Erin 2004, Black attempted to break out with 14.. .£5?!, but after 15 exf5 gxf5 16 Kh1 Kh8 17 Bh6 Bxh6 18 Qxh6 Rf6 19 Qd2 Qe8 20 Nd5 Bxd5 21 cxd5 Nce5 22 f4 Ng4 23 Bf3 White had a large advantage and went on to win convincingly thanks to Black's numerous weaknesses. 12Qd2 12 Rcl!? may be slightly more accurate, and will be considered in Game 35. 12 ... Ne5 13 Na3!? (Diagram 33) This looks like the most secure method of defending the c4-pawn. The main point of Black's play is revealed after 13 b3 b5! (Diagram 34) 14 cxb5 Qa5! (forcing the following loosening of White's queenside) 15 b4 Qc716 Bd4 and now 16... Bxb5 17 Ne3 was about equal in C.Lamoureux-Y.Afek, Oakham 1993, while 16... Nc4!? 17 Bxc4 Qxc4 offers excellent compensation (the immediate threat is ... e7-e5 winning a piece).

282

Mar6czy Bind: Avoiding Exchanges with Nc2 TIP: The knight may look poor on a3 but it has some positive points as well. In the event of Nc3-d5 followed by an exchange on d5, the reply c4xd5 will enable the knight to come to c4 and perhaps b6, where it may help White to conquer the c-file.

If, following c4xd5, Black prevents this manoeuvre by playing ... b7-b5, the knight may look to occupy the newly created outpost on c6, which can be reached via c2 and b4 (or d4).

Diagram 33 (B)

Diagram 34 (W)

Can Black avoid passivity?

An important tactic

13 ... Be6

Another playable idea is 13...Bc6!? 14 Racl Re815 b4 b6 16 Rfd1 Ned7 setting up a Hedgehog formation, as in M.Tempone-H.Spangenberg, Trelew 1995. 14Nd5 It now looks as if White has succeeded in stifling any counterplay while retaining a useful space advantage. However, Black may be able to free his position with the surprising ... 14... Bxd5!?

A playable alternative is 14... Nfd7!?, which can sometimes be followed by ... f7-f5. 15 cxd5

Perhaps White could consider 15 exd5!? e616 dxe6 fxe617 Racl, as in D.Prochazka-N.Guth, Correspondence 1999. 15 ... e6! 16 dxe6 fxe6 (Diagram 35)

Black's position may appear a little loose, but he ought to have enough dynamic resources to hold the balance. He is planning ... d6-d5, after which the likely sequence e4xd5, ... Nxd5 will leave him with two well-placed knights in the centre. In V.Savon-E.Gufeld, Sukhumi 1972, the continuation was 17 Racl Rxc118 Rxcl d5 19 Rdl, at which point the players agreed an early peace treaty.

283


Diagram 35 {W)

Diagram 36 (B)

...d6-d5 is coming

The most accurate move order?


0 N.Short • R.Felgaer Najdorf Memorial, Buenos Aires 2001

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd7 10 Nc2 a6 11 f3 ReS 12 Rc1!? (Diagram 36} 12 ... Re8!? This can often be a useful waiting move in both the traditional Dragon and the Accelerated version. One of the points of White's 12 Rcl move order (rather than 12 Qd2) is seen after the plausible sequence 12... Ne5 13 Na3 Qa514 Qb3! when the threat of Bb6 will force an embarrassing retreat. Instead Black should prefer 13...Be6, when 14 b3 Qa515 Nab1 was unconvincing in Y.Salaun-S.Maze, St Lorrain 2001, so White may do better to investigate 14 Nd5!? Bxd515 exdS!? with a slight advantage. 13 Qd2 Qa5 Black returns to a normal plan now that Qb3 ideas are no longer an option. 14 Rfd1 Ne5 15 b4!? Short seems to enjoy the space-gaining approach (compare the Short-Anderson game cited earlier). However, White may have a stronger alternative in 15 c5!?, intending to meet 15 ... dxc5 with 16 f4 Neg4 17 e5, when the onus is on Black to find sufficient compensation for the soon-to-be missing piece. 15 ... Qd8 16 Na3! This knight move crops up quite often in the Nc2 lines. Here it provides a convenient way of overprotecting the c-pawn. 16... a5!? This seems the most challenging continuation. Instead 16 ... Bc6?! 17 b5 axb5 18 cxb5 Bd719 284

Mar6czy Bind: Avoiding Exchanges with Nc2 b6 (Ftacnik) looks uncomfortable for Black. White controls a lot of space, and will follow up with a knight invasion on c7 or a7. 17 bs (Diagram 37)

Diagram 37 (B)

Diagram 38 (B)

Complicated, but better for White

Should Black offer a queen trade?

The position is strategically very interesting. Normally it would not be desirable for White to fix his queenside pawns in this way, but here the policy is justified by his ideally placed pieces. 17... Be6

There isn't time for 17...b6?, as Ftacnik gives 18 Na4 Rb8 19 Bd4 Qc7 20 Qe3, winning the b6-pawn. Black can offer more resistance with 19... Be6, intending a knight retreat to d7, but 20 Nxb6! Rxb6 21 Bxe5 Nxe4 22 fxe4 Bxe5 23 c5! looks good enough, when the passed pawns should decide the game. 18 Na4 Nfd7

18 ... Ned719 b6! is also very good for White. 19 b6! If Black were able to play ... b7-b6 himself, then all of his problems would disappear. 19... Ncs!

The best chance. If instead 19... Nc6 then 20 NbS maintains strong pressure. 20 Nxcs dxcs 21 Qxas

21 Qxd8 Rexd8 22 Rxd8+ Rxd8 23 Bxc5 Rd2 24 Rc2 Rxc2 25 Nxc2 Bxc4 26 Bxc4 Nxc4 27 Kf2 is slightly better for White, but Black has good chances to hold. 21 Rb1!? Qxd2 22 Rxd2 may have been worth considering though. 21 ... Nd7!

21...Bd7? 22 Bxc5 is no good. 22 Nbs (Diagram 38) 22 ... Ra8

Felgaer opts to keep the queens on the board, but 22 ... Qxb6 was a legitimate alternative, when 23 Qxb6 Nxb6 reaches a strange kind of endgame in which both sets of pieces are

285

Starting Out: The Accelerated Dragon somewhat restricted by the mutual pawn weaknesses. White would probably need to organize some sort of kingside advance in order to make progress. 23 Qd2 Qxb6 24 f4

White keeps an edge thanks to his space advantage and strong b5-knight. At the same time he must take care not to allow his weak queenside pawns to become too significant a factor. 24... Nf6 25 Qc2 Bg4 26 eS BfS! 27 Qb3 (Diagram 39)

Diagram 39 (B) White remains on top

Diagram 40 (W) A missed opportunity

27 ... Ng4

Ftacnik calls this dubious, but it doesn't seem any worse than 27... Nd7, after which 28 Rd2 Rad8 29 Rcd1 maintains a pleasant advantage. 28 Bd2

Also possible was 28 Bxg4 Bxg4 29 Rd5 Rec8 30 Nc3! Qa7 (or 30 ... Qxb3 31 axb3 b6 32 Rd2) 31 Rd2 intending Nd5 and maintaining the advantage. 2s ... hs 29 Bf3?

Allowing an opportunity for counterplay. 29 h3 Nh6 30 Bf3 would have kept control. 29 ... Red8?

Returning the favour. If Black had found 29 ...g5! (Diagram 40), things could have become much more interesting; e.g. 30 h3 gxf4! 31 hxg4 hxg4 32 Bd5 Bxe5 is highly unclear. White may be able to claim an objective superiority, but the important thing is that Black has changed the unfavourable pattern of the game to reach an unclear position in which three results are possible. 30 Be1 f6

Once again 30 ... g5! could have been played. 31 a4! fxes 32 as Qf6 33 h3! (Diagram 41)

Now Black is forced to sacrifice material under much less favourable circumstances. 33 ... exf4

33 ... Nh6 fails after 34 Bxb7 Rxdl 35 Rxd1 Rf8 36 a6 etc.

286

Mar6czy Bind: Avoiding Exchanges with Nc2 34 hxg4 hxg4 35 Bxb7

The difference between this and the vaguely analogous position after 29 ... g5! is huge. Here Black is not just giving up a piece; he is also forced to relinquish the b7-pawn, after which White's a-pawn will be practically unstoppable. 35 ... Rxd1

35 ... Rab8 loses trivially after 36 a6. 36 Rxdl Rf8 37 g31

A final accurate move, after which the win should be fairly straightforward. 37 a6? g3! would allow Black to fight on. 37 ... f3 38 Qe3 (Diagram 42)

Diagram 41 (B) White is winning

Diagram 42 (B) The end is nigh

White has an overwhelming material advantage and the win should be simple. 38 ... Kh7 39 Rd2 Qat?!

A blunder in a lost position - Black forgets that his own king is also vulnerable! 40 Rh2+ Kg8 41 BdS+ 1-0

Theoretical? There is not much heavy theory, although a little bit of specific knowledge of different plans and move orders will be helpful for both sides.

Statistics 10 Nc2 scores an impressive 68% for White. If you don't fancy facing it as Black then there is always the 9 ... Nxd4 move order as shown on page 152.

Summary and Conclusions The Nc2 system has the advantage of being theoretically challenging, without requiring an encyclopaedic knowledge. It also carries the considerable practical benefit that, in contrast

287

Starting Out: The Accelerated Dragon to the previous three chapters, it is White who gets to dictate the character of the early struggle, whereas if he leaves the knight on d4, he has to be ready to face the Classical, Gurgenidze or 7... Ng4 systems (as well as certain sidelines discussed in the next chapter). The 'Nc2 versus everything' approach provides a certain amount of consistency and coherency to one's repertoire. We have seen that Black's options will vary according to the specific timing of the knight retreat. Thus after s... Bg7 6 Nc2, players with a love for the dynamic and irregular need look no further than 6... Nh6!?, which more or less guarantees the above without being tactically unsound. Another practical advantage of this approach is that it is not mentioned by Nielsen, whose repertoire coverage in Experts vs. the Sicilian is likely to be the primary reference point for many Nc2 practitioners. However, none of this will be of any relevance to Gurgenidze players, who are required to postpone ... Bg7 in favour of the move order s... Nf6 6 Nc3 d6, after which 7 Nc2 renders the early ... £7-f5 plan impossible. TIP: If the reader is a fan of the Ciurgenidze but also likes the look of the ...f7-f5 plan against Nc2, he may do well to consider varying his move order against different opponents.

If you know before the game that your opponent is a Nc2 player, then the chances are that you will be able to 'bluff' with 5 ... Bg7 when he will probably play 6 Nc2 anyway, rather than 6 Be3. (An alternative is to opt for the 'Accelerated Gurgenidze' to draw the opponent into unfamiliar territory - see page 181 ).

Fascinating as the ... f7-f5 plan is, the most solid and popular option for Black will always be the classical plan of development with ...Bg7 and ... Nf6. Then 8 ... 0-0 followed by 9 ... Be6!? may be underrated, but 8... Nd7 is currently regarded as the main line. In my opinion White should prevent the doubling of the c-pawns that results from ... Bxc3. A few strong players have experimented by allowing it, hoping to exploit the weak dark squares, but Black should be absolutely fine here, especially if he has been prudent enough to delay castling. In most games White will play Bd2 in response to ... Nd7, securing his pawn structure for the time being. Then Black faces a final decision concerning the d7-knight which is headed for c5. Game 32 featured the interesting 10... a5!?, safeguarding the future of the knight on its best square. Even though Black ended up losing that game, his lOth move appears to be fully playable, as evidenced by the annotations. After the more common 10... Nc5 White's most dangerous response is 11 b4!. Now it looks too risky for Black to accept the pawn sacrifice with ll...Bxc3?!, so he should retreat the knight. 11... Ne6 may not fully equalize, but with careful play Black should be able to reach an acceptable position; it would also be interesting to see Houska's 12 Rc1 Nf4!? tested in practice. Finally, Black can consider the cheeky 11... Nd7!? as in Game 34. The move is certainly sound and worked to perfection in that particular encounter, though it remains to be seen whether further testing will reveal a convincing way for White to utilize his space advantage. In the final part of the chapter we considered the idea of delaying Nc2 in order to reduce Black's options. This certainly deserves attention from White, but it can only be used in specific circumstances, i.e. the Classical Mar6czy with 9... Bd7. If you want to play this way with White, it's all well and good, but please do bear in mind that you will need to consider all of Black's alternatives (such as the Gurgenidze and 7... Ng4 systems) which the early Nc2 avoids. Finally I will remind the reader that if Black does not wish to face the delayed Nc2 line, he can always utilize the alternative move order of 9... Nxd4 in the Classical.

288

Chapter Ten

Maroczy Bind: Sidelines

-

The 'Mar6czy-Vugoslav Attack'

-

The Remarkable

-

Lines with ••• Nh6 and ••• f7-f5

-

The Double Fianchetto Variation

-


-

Final Word

s... Bh6!?

Starting Out: The Accelerated Dragon 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 (Diagram 1)

Diagram 1 (B)

Diagram 2 (B)

The Mar6czy Bind

White's aggression will be punished

We conclude our coverage of the Maroczy Bind by considering a variety of uncommon options available to both sides. It is mostly Black who has the opportunity to steer the game away from heavily analysed territory if he so chooses. For White it is not so easy to find purposeful ways of deviating from standard development. Having established the traditional Maroczy pawn formation (e4/c4), the queen's knight clearly has no better square available than c3, while the king's bishop will almost always head for e2. Essentially White will spend the next few moves developing his pieces on their usual squares, while reacting appropriately to Black's chosen set-up. True, White does get to make certain decisions, the main one being whether to retreat the d4-knight to c2 or reinforce it with Be3, but both of these options have been covered in the preceding chapters.

The 'Mar6czy-Vugoslav Attack' I have only been able to think of one meaningful way for White to steer the game into independent paths. I have decided to call this system the 'Maroczy-Yugoslav attack' for reasons that will become clear. Occasionally White players have 'forgotten' that they are playing a Maroczy Bind and switched to a kind of Yugoslav Attack involving a direct assault on the enemy king. While this plan is entirely normal in the Dragon, it is almost never played in the Maroczy Bind, as the pawn on c4 is hardly conducive to the successful realization of a mating attack. There is no reason for Black to fear this crude approach, but it has been known to catch players off guard, especially at club level, so we will take a quick look at how he should react. The precise details will vary depending on Black's chosen move order. First we will consider the sequence with 5... Bg7. 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 f3

8 Be2 would be the normal choice, but we will assume that White is not interested in short castling.

290

Mar6czy Bind: Sidelines 8... d6 9 Qd2 Bd7 10 g4?! Commencing an immediate attack, though the drawbacks of this approach will soon become apparent. Instead, 10 0-0-0 would commit the king and present a clear target for Black's counterplay after 10...Qa5 (10 ... a6 is also possible) 11 Kb1 Rfc8 intending some combination of the moves ... Ne5, ... a7-a6 and ... b7-b5. 10... Rc8 11 h4 NeS 12 b3 Qas 13 hS? (Diagram 2) In view of what follows White should try something else, but it is hardly worth our time analysing every possibility. All I am doing here is providing a general picture to show why White is in no position to execute a crude attack. 13 ... Nfxg4! 14 fxg4? The lesser evil was 14 Nd5 Qxd2+ 15 Bxd2 Nx£3+! 16 Nxf3 Bxa117 Nxe7+ Kg718 Nxc8 Rxc8, though even here Black has an extra pawn and a large advantage in the ending. 14... Nf3+! 15 Nxf3 Bxc3 and Black won the queen in G.Fant-E.Taina, Stockholm 1992. Occasionally White tries the same plan against the Gurgenidze System. The details are slightly different, but the overall verdict is the same: White's position is not secure enough to justify a full scale kingside assault. 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6 7 f3 If White intends to play aggressively then this is the most logical move order. As we saw in Chapter Seven, 7 f3 can be a fairly respectable line, so for the time being White can avoid revealing his intentions. 7 ... Nxd4 8 Qxd4 Bg7 9 Be3 0-010 Qd2 aS!?

If Black somehow knew in advance that his opponent planned to attack on the kingside then 10... Qa5 would be the most logical move, but so far White's play has been quite consistent with a normal positional strategy. In Chapter Seven I recommended 10 ...a5!? for Black, so we will keep to that here as well. Now 11 Be2 would be the sane choice; instead we will look at a couple of ways in which White tries to play aggressively.

a) First, he might opt for an exchange of bishops with 11 Bh6?!. While this is often a highly desirable plan in the Dragon, here it is time consuming and rather anti-positional, given that c2-c4 has left White's dark squares a little tender. Black should draw the queen away from the centre with 11...Bxh6! 12 Qxh6 Qb6 (Diagram 3) with good counterplay. The white queen may look threatening on h6, but bringing up reinforcements takes far too long. Now 13 0-0-0?! is the consistent move: consistently aggressive and consistently bad!. Black easily gains the advantage after 13... Be6 14 h4 Rac8! (so that the other rook can defend the kingside from f7 if necessary), when there might follow 15 h5 Bxc4 16 hxg6 fxg617 Kb1 R£7! and Black is a pawn up with a safe king (but not 17... Bxfl ?? 18 Nd5!). b) 11 g4?! is not a great move either, though it at least avoids dark square weaknesses. We should also remember from Chapter Seven that in certain endgames the moves g2-g4 and h2-h4 can be quite useful for White, so he may also look to offer a queen exchange. Black should probably continue with 11...a4 and meet 12 h4 with 12 ... h5!? (Diagram 4), ensuring that the h-file will not be opened. Now 13 gxh5 Nxh5 just leaves White with weak pawns, while 13 g5 Nd7leads to a closing of the kingside and an end to White's attacking aspirations. Black intends the standard developing moves ... Nc5, ... Be6 and so on, reaching normal-looking positions, except that White has wasted time on his wholly unnecessary and unhelpful pawn moves.

291


Diagram 3 (W) White is already worse

Diagram 4 (W) White's 'attack' achieved nothing

Theoretica I? Not at all. These lines are not even mentioned by most sources, as it is usually taken for granted that the Maroczy Bind is wholly unconducive to Yugoslav-style attacking plans. Given that this is a Starting Out book I felt it was worth taking the time to cover this possibility, as I can imagine a newcomer to the opening (especially one who is not versed in the Dragon) feeling rather intimidated by an unexpected kingside assault. I hope the above lines have demonstrated that Black has absolutely nothing to fear.

Statistics It is hard to give a total figure, as there are many different move orders by which White

may employ a Maroczy-Yugoslav set-up. But needless to say, the general pattern is that White scores pretty poorly and the Maroczy-Yugoslav is almost never used by anyone who knows anything about the opening. We will now look at a number of ways in which Black may steer the game away from the major systems, beginning with a deviation as early as the fifth move!

The Remarkable

s... Bh6!?

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bh6!? (Diagram 5)

According to my database 5 ... Bh6 was first played by Uogele in 1980, but it was only after Tiviakov employed it against Beshukov in 2000 that the chess world really paid attention. The move may appear bizarre, but it is actually quite logical when you keep in mind that, in the majority of variations of the Maroczy Bind, the exchange of dark-squared bishops is positionally quite desirable for Black. By aiming for an immediate bishop exchange, Black is essentially playing for an improved version of Chapter Six.

292

Mar6czy Bind: Sidelines

Diagram 5 {W)

Diagram 6 (B)

An unusual but logical sideline

Ambitious but as yet untested

6 Bxh6

Almost certainly best, as it forces the knight to the edge. The alternative 6 Nc3 Bxcl 7 Rxcl allows Black to develop his knight more favourably with 7... Nf6, after which 8 Be2 d6 (8 ...Qb6?! 9 Nb3 0-0 was the original game G.Piesina-A.Uogele, Kaunas 1980, and now 10 Nd5! looks tempting) 9 0-0 0-0 10 Nc2!? was just a little better for White in J.GallagherC.Beaumont, British League 2002. So far no-one has tried the ultra-ambitious 6 f4!? (Diagram 6), aiming to prevent the bishop exchange at all costs. A possible continuation is 6... Bg7 7 Nc2!?, leading to a strange version of the Nc2 system covered in Chapter Nine; it is unclear which side is more likely to benefit from the inclusion of the move f2-f4. White can also play 7 Be3, when 7...Qb6!? is tempting, though White gains decent compensation for the sacrificed pawn after 8 Nf5 Qxb2 9 Nxg7+ Qxg7 10 Nd2, due to the undeniable strength of his dark-squared bishop. 6 ... Nxh6 7 Nc3

o-o 8 Be2 d6

This has been the most popular move, although Tiviakov later refined his play with 8 .. .£6!? 9 0-0 N£7 10 Rcl b6 11 Qd2 Bb7 (Diagram 7) when a draw was agreed in J.SmeetsS.Tiviakov, Dieren 2003. White stands slightly better here, but the presence of his bad bishop means that he will always have to worry about potential endgame difficulties. 9 o-o (Diagram 8)

As I mentioned previously, the game which really drew attention to 5 ...Bh6!? was S.Beshukov-S.Tiviakov, European Ch., Saint Vincent 2000. That encounter continued 9 Qd2 Ng4 10 Bxg4! (10 0-0, as in E.Chevelevitch-S.Bravo Lutz, Hamburg 2001, allows Black to recentralize the knight with 10... Nf6) 10... Bxg4 (Diagram 9), when White had achieved a favourable exchange of his worst piece. The game continued 11 h3 (11 0-0 would have been a bit more accurate, as the bishop has little future on g4 and may even retreat to e6 voluntarily; note that 11...Qb6 would be pointless due to 12 Nc2, as taking on b2 would cost Black his queen) 11. .. Be612 b3 Qb613 Nc2 a6 14 Ne3 Qc5 15 Rcl Qe5 16 0-0 Qf4 when a draw was agreed. White stands slightly better,

293

Starting Out: The Accelerated Dragon but Black's position is very solid and tough to break down.

Diagram 7 (W)

Diagram 8 (B)

Black is only marginally worse

Black should improve the h6-knight

Personally, instead of 9... Ng4, I prefer 9 ... Nxd4!? 10 Qxd4 f6 (Diagram 10), when 11 0-0 Be6 reaches a position from line 'c' below. If it comes to a choice between Black exchanging a pair of knights or a knight for White's light-squared bishop, then I would definitely go for the knights. Objectively he is a bit worse in both cases, but as long as White's bad bishop remains on the board, Black can always aim for a favourable endgame, whereas after its exchange White is more or less 'playing for two results'.

Diagram 9 (W)

Diagram 10 (W)

White has exchanged his bad bishop

A solid set-up for Black

~ 294

TIP: While White's 'bad' light-squared bishop remains on the board, there is always a danger he might land up in a bad endgame. Therefore White should always consider the possibility of exchanging it, while Black will often try to deny him this option.

Mar6czy Bind: Sidelines After 9 0-0 Black has tried: a) 9•• .fs7! (G.Gaboleiro-J.Flores, Benedita 2001) looks too risky after 10 c5! dxc5 (10 ... Nf711 Bc4) 11 Nxc6 bxc6 (1l...Qxd1? 12 Nxe7+ K£713 Raxd1 Kxe714 Rfe1 is too dangerous for Black) 12 Qcl Kg7 13 Rdl. White will regain the c-pawn with a subsequent Na4, maintaining an initiative as well as a better pawn structure. b) 9...Qb610 Nb3 Be611 Qd2 Ng412 Nd5 Bxd513 exd5 Nce514 Bxg4 Nxg4 (Diagram 11) 15 Nd4 (15 Qe2!? Ne5 16 c5 also looks promising) 15 ... Ne5 16 b3 Rae8 17 Rae1 e618 dxe6 fxe6left White with a stable advantage due to Black's 'hanging' central pawns in M.Ginzburg-A.Hoffman, Villa Ballester 2001.

Diagram 11 (W)

Diagram 12 (W)

White has the advantage

A satisfactory position for Black

c) 9... Be6!7 would probably be my choice. Now A.Collins-M.Basel, Internet 2001, continued 10 Rcl Qb6?! (the queen does not have much of a future here and will soon be driven away by Nd5) 11 Nb3 f5?! 12 Nd5 Qd8 13 Qd2 N£714 Nd4 with advantage to White, but 10... Rc8 looks better. White can also consider 10 Qd2, which should be met by 10... Nxd4! 11 Qxd4 f6 (Diagram 12). Black is solid, if a bit passive. At the same time White's bad bishop has remained on the board, so the first player must always take care not to slip into an unfavourable endgame.

Theoretical? There is very little theory to learn for either side.

Statistics 5... Bh6 has scored 34%, though there haven't been a huge number of games, so it is hard to say whether this figure offers an accurate representation of the chances. If the reader is looking to add a secondary defence to his repertoire, leading to solid positions while requiring almost no theoretical knowledge, then 5 ...Bh6 is well worth investigating. 295


Lines with ... Nh6 and ...f7-f5 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nh6!? (Diagram 13)

Diagram 13 (W)

Diagram 14 (B)

Black intends .. .f7-f5

The standard position

Here we will consider sidelines in which Black strives to battle the bind with an aggressive ... f7-f5 in the early stages of the game. We have already encountered the idea of ... f7-f5 against an early Nc2 in Chapter Nine. In that case my conclusion was that the pawn advance was a slightly risky but quite principled reaction to the retreat of the white knight from the centre. If, on the other hand, the knight remains on d4, then Black's show of aggression obviously carries a considerably higher risk factor. That is not to say that it is unplayable, but I would not consider it to be in the same league as any of the major systems when it comes to soundness and reliability.

Move Order Subtleties The ... f7-f5lines are slightly hard to classify, as there are a few different move orders, each of which present different opportunities to one side or the other. The majority of games reach Diagram 14, though both sides have experimented along the way. Beginning on move six, Black must choose in what order to play ... d7-d6 and ... Nh6, while White must decide which of his developing moves, Nc3 or Be2, should come first. These decisions may appear trivial, but there are some subtle differences depending on what options both sides wish to afford the opponent.

A Strong Deviation for White Firstly I would like to draw the reader's attention to a very effective, though comparatively rarely played line for White. After the most popular sequence of... 1 e4 C5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 d6 7 Nc3 Nh6

I believe that White should take the opportunity to shut the black knight firmly out of the game with ...

296

Mar6czy Bind: Sidelines 8 f3! (Diagram 15) 8 Be2 reaches the main position shown in Diagram 14, but I believe the text to be superior.

Diagram 15 (B)

Diagram 16 (B)

A strong line for White

An important resource for White

This strong move comes with Khalifman's seal of approval. Black is more or less forced to continue with ...

s...fs This seems the only consistent move.

WARNING: The advantage of 8 f3 is that Black's knight is more severely restricted on h6. In D.Hamilton-S.Liljedahl, Correspondence 1961, Black blundered with 8...0-0?? and lost quickly after 9 Nxc6! bxc6 10 Qd2. In the analogous position with 8 Be2 the g4-square would have been available. We should also note that 8... Qb6?? would be disastrous in view of 9 Nf5, when White wins a piece. However, this idea may work rather well with a slightly different move order, as we will see. 9 Qd2 Nf7 In A.Stummer-A.De Araujo, Paranana 1993, Black preferred 9 .. .£4!? 10 Bf2 0-0. Perhaps he was dreaming of a positional bind based on his control over e5, but after the powerfulll c5! (Diagram 16) it became clear that his enterprising opening play had brought him nothing but weaknesses. The game continued 1l...Bxd412 Bxd4 dxc513 Bc4+ N£714 Bxc5 Qxd2+ 15 Kxd2 Kg716 Nd5 Nfe517 Bb5 Rd818 Ke2 with a large advantage to White, who went on to win convincingly. 10 exfs gxfs 11 cs! (Diagram 17) Once again this excellent move puts Black under considerable pressure. N.NikcevicL.Popov, Herceg Novi 2000, continued 1l...dxc5 12 Nxc6 Qxd2+ 13 Kxd2 bxc614 Bxc5 when Black faced an uphill defensive struggle due to his split queenside pawns, and White gradually converted his advantage. 8 f3! looks like a strong antidote to Black's opening strategy. If the reader wishes play the

297

Starting Out: The Accelerated Dragon ... f7-f5lines, I would advise him to take steps to avoid this variation. He can do this by delaying ... d7 -d6 as we will now see.

Diagram 17 (B)

Diagram 18 (B)

That move again!

A less effective implementation

Black's Most Accurate Move Order: 6 ... Nh6 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nh6!? At this point White's most popular move is 7 Be2 which will be discussed shortly. For the time being we will assume that White is hoping to reach the 8 f3! variation above, and consider the ways in which he may try to transpose. a) On the immediate 7 f3, White must reckon with 7... Qb6!?, exploiting the temporary looseness of his pieces. A.Wabschke-H.Kauder, Ruhrgebiet 2004, continued 8 Nf5 (there is nothing better) 8 ...Qxb2 9 Nxg7+ Qxg7, when White's strong dark-squared bishop offered some compensation for the missing pawn, but not more. b) 7 Nc3 0-0 8 f3!? (Diagram 18) [8 Be2 d6 reaches the main line]. This is still possible, but the position is quite different with ... 0-0 having been played instead of ... d7-d6. For example, the c4-c5 resource, which was quite powerful in the 6... d6 7 Nc3 Nh6 8 f3 variation, does not attack anything with the d-pawn still on d7, while in other lines Black may benefit from the fact that he has already castled. From Diagram 18 Black may try: a) 8.. .fs, when 9 Nxc6 is tame and 9 ... dxc6! was equal in I.Ivanvco-M.Ujtelky, Czechoslovakian Ch. 1959, and following the dubious 10 Qb3?! Black could have seized the initiative with 10.. .fxe4! 11 Nxe4 Nf5. However, 9 Qd2 looks more challenging, when play might continue 9 ... Nxd4 (or 9 .. .f4!?) 10 Bxd4 Bxd4 11 Qxd4 fxe4 12 Qxe4 with a slight edge to White. b) 8...Qb6!? would probably be my choice. This was unplayable with the 6... d6 move order but with ... 0-0 it makes very good sense. Now 9 NebS!? may be White's best (9 Na4 Qa5+ 10 Bd2 Qe5 looks a bit awkward), when there may follow 9 ... Qa5+ 10 Kf2!? (after 10 Bd2 Qd8 it is doubtful that White has anything better than 11 Be3 Qa5+ repeating, or 11 ... f5!? if Black is feeling ambitious) 10.. .£511 Nb3 Qd8 with a decent position for Black. He has achieved the

298

Mar6czy Bind: Sidelines desired push of the £-pawn, while White's knights and king are rather oddly placed. In conclusion, it seems that Black can take the sting out of the £2-£3 plan by playing 6... Nh6 and 7... 0-0 while delaying ... d7-d6.


0

R.Hubner • A.Afifi

Manila lnterzonal1990

1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nh6

The actual move order was 6... d6 7 Nc3 Nh6 8 Be2 0-0, but having argued in favour of 6... Nh6 I will 'pretend' it was played in this game to reinforce the point. Furthermore, presenting the game this way will allow us to investigate a couple of other subtleties. In particular, White players must decide in which order to play Be2 and Nc3, while Black may still postpone ... d7-d6 to accelerate his counterplay with ... £7-£5, although this approach may not be completely sound. 7 Nc3

7 Be2 has been the most popular, though it usually ends up reaching the same position.

0

NOTE: The advantage of 7 Be2 is that it controls g4, so White may be in a position to harass the h6-knight with Qd2. Therefore the safest response for Black is 7...d6, so that Qd2 can be met by ... Ng4.

Instead of 7... d6 Black has a couple of independent possibilities: a) 7.. .£5 8 ex£5 Nx£5 9 Nx£5 Qa5+ (9 ... Bxb2 10 Nd2 gx£5 11 Bh5+ K£8 12 Qc2 B£6 13 Bh6+ Kg8 14 Qx£5 was very good for White in W.Unzicker-M.Filip, European Team Ch., Vienna 1957) 10 Nc3 Qx£5 11 0-0 0-0 12 Qd2 d613 Racl Be614 Nd5 (Diagram 19) gave White a typical slight advantage in H.Wirthensohn-R.Wade, Birseck 1971.


Diagram 20 (W) The basic ... Nh6 position

299

Starting Out: The Accelerated Dragon b) 7... 0-0 8 Qd2 Nxd4 9 Bxd4 d610 Nc3 Bxd4 (10 ... f5?! looks risky with the dark-squared bishops already exchanged) 11 Qxd4 (M.Makarov-K.Kiik, St Petersburg 1999) 11 .. .£612 0-0 Nf7leads to a position resembling Diagram 7 from the 5 ... Bh6 variation. 1 ... 0-0 8 Be2 d6 (Diagram 20) We are now back in Hiibner-Afifi, with Diagram 20 showing the basic ... Nh6 position. Black has one more independent option in 8 .. .£5!?, postponing ... d7-d6 for a while longer. Following the natural sequence 9 exf5 Bxd4 10 Bxh6 Rxf5 11 0-0 Black can still transpose to Hiibner-Afifi with 11...d6, but he has occasionally experimented with 11...Qb6!?, creating immediate threats against both f2 and b2. This is a principled approach, but also a very risky one as Black's queenside pieces are still undeveloped. White can look to exploit this with 12 Nd5! (Diagram 21) and now:

Diagram 21 (B)

Diagram 22 (B)

Black is playing with fire

Black is in trouble

a) 12... Bxf2+ 13 Kh1 Qxb2 (or 13 ... Qd4 14 Bg4! Qxd1 15 Raxd1 R£7 16 Nxe7+! Nxe7 17 Be6 Nf518 Bx£7+ Kx£719 Rxf2 and White won easily in A.Grushevsky-G.Veresov, USSR Team Ch. 1959) 14 Bg4 R£715 Nxe7+!! Rxe7 (15 ... Nxe716 Be6!) 16 Qd5+, when 16 ... Re6 17 Bxe6+ dxe618 Qf3, 16... Rf7 17 Rae1! and 16... Kh8 17 Qf3! Re8 18 Rae1! are all equally hopeless for Black. b) 12... Qxb213 Rb1 Qa314 Bg4 R£715 Nc7 Rb8 16 NbS Qc517 Nxd4 Qxd418 Be3 Qxd119 Rfxd1 Ne5led to success for Black in P.Marusenko-M.Turner, London Lloyds Bank 1994, but White can do much better with 13 Nxe7+! Nxe714 Rb1 (Diagram 22) when the queen cannot maintain her defence of the d4-bishop. Play may continue 14 ... Bxf2+ (14 ... Qc3 15 Rb3) 15 Rx£2 Qe5 (or 15...Qf6) 16 Bf3 with a large advantage due to the pair of marauding bishops and Black's undeveloped queenside. 9 0-0 9 Qd2 Ng4 10 Bxg4 Bxg4 11 0-0 Qa5 12 Racl Rfc8 13 b3 a6 left Black well prepared for queenside counterplay in E.Gufeld-B.Gurgenidze, USSR Ch., Leningrad 1960. g.. .fs (Diagram 23) This position is reached in the majority of games featuring the ... Nh6/.. .f7-f5 plan. Black's

300

Mar6czy Bind: Sidelines intentions are laudable enough, but I cannot help but feel a tinge of scepticism over the objective merits of his aggressive approach.

10exfs Bxd4

Diagram 23 (W)

Diagram 24 (W)

Dynamic counterplay or weakening?

Dark-squared bishops favour who?

Black has a choice here. a) 10... Nxf5 is played less frequently, but may not be inferior. After 11 NxfS Bxf5 12 Qd2 QaS 13 Racl Be6 (Diagram 24) we reach a position similar to the main game, but with the dark-squared bishops still on the board. It is hard to decide who this is more likely to favour. The usual rule is that Black benefits from the bishop exchange, but after the additional pawn trade this is not so clear cut. The removal of Black's f-pawn means his king is more exposed than normal, while the 'Dragon Bishop' can prove a useful defender as well as playing an active role on the long diagonal. On the other hand the e3-bishop is also useful in controlling important dark squares such as b6, cS, d4 and f4. In K.Sakai-N.Zambor, Correspondence 2003, White decided to exchange these bishops anyway with 14 Bh6 and obtained a slightly favourable position after 14 ...Bxh615 Qxh6 Rf7. b) 10... gxf5 is also seen from time to time. By recapturing with the pawn Black increases his influence over the centre; the drawback is that his rook and knight won't enjoy so many active prospects. White must fight for the centre while preventing the further advance of the black f-pawn. This can be achieved by 11 f4! (Diagram 25), when there may follow: b1) l l ...Qb6 12 NebS Nxd4 13 Bxd4 Bxd4+ 14 Qxd4 Qxd4+ 15 Nxd4left Black facing a difficult endgame in M.Stean-R.Bellin, Norwich 1972, due to his numerous weaknesses. b2) 1l...Ng4 12 Bxg4 fxg4 13 Qd2 Bf514 NxfS Rxf515 Qe2 h516 Ne4 Qe817 Rad1 Rd8 18 Ng3 RaS was V.Neverov-B.Abramovic, Yugoslav Team Ch. 2000, and now White should play 19 RdS when he is clearly better. b3) l l ...Bd712 Kh1!? is a sensible idea, intending to meet ... Ng4 with Bgl. A.WojtkiewiczA.Hoffman, Valencia 1990, continued 12... Kh813 Rcl Rg814 Bf3 Qe815 Re1 ReS 16 b3 Ng4 17 Bg1 QhS?! 18 Nxc6 Bxc619 NdS (or 19 Rxe7 Qxh2+ 20 Bxh2 Nf2+ 21 Kg1 Nxd1 22 Bxc6 Nxc3 23 Bxb7 Rce8 24 Rxe8 Rxe8 when White's extra pawn gives good winning chances)

301

Starting Out: The Accelerated Dragon 19...e5 20 Ne7 e4? (20 ...Bh6 would have kept some chances alive) 21 Bxg4 fxg4 22 Nxc8 Rxc8 23 Bd4 and White won easily. Returning to 10... Bxd4 (Diagram 26}:

Diagram 25 (B)

Diagram 26 (W}

Fighting for the centre

Which piece to capture?

11 Bxh6

11 Bxd4 Nxf512 Be3 Nxe313 fxe3 Rxfl+ 14 Bxfl Qb6 was not altogether clear in P.SmithJ.Dulik, Correspondence 2002. For the time being Black is poorly mobilized, but his knight will find a splendid home on e5. 11 ... Rxf5 12 Bf3

12 Qd2 is a reasonable alternative, leading to roughly similar positions. 12 ... Bg7

Later in the same tournament Afifi attempted to improve with 12... Bxc3 13 bxc3 Ne5, but after 14 Be4 Rh5 15 Be3 Be6?! 16 Bxb7 Bxc4 17 Bxa8 Qxa8 18 Re1 Black did not have enough for the exchange (S.Rachels-A.Afifi, Manila Interzonal1990). Nunn and Gallagher suggest 15 ...Qc7, but even here White can remain on top after 16 c5! Be6 (or 16... dxc517 Bf4 and White's bishops are far more important than Black's extra pawn) 17 cxd6 exd618 Bd5 with an enduring advantage thanks to the superior minor piece and safer king. 13 Bxg7 Kxg7 14 Be4 Rf7 15 Qd2 (Diagram 27)

The remainder of the game should serve as a warning to any readers contemplating playing this variation with Black. The position may not be losing, but it is clear that the aggressive opening strategy has weakened his position without providing any counterplay. 15 ... Qb616 Bd5!

Provoking a further weakness. 16 ... e6

16... Rf8 is bad after 17 Rae1 Ne518 Kh1 followed by f2-f4. 17 Bxc6 Qxc618 Rad1

Fritz points out an even stronger 'computer move' in 18 NbS!? d519 Nxa7! winning a pawn

302

Mar6czy Bind: Sidelines thanks to the potential fork on d4. 18 ...Qxc4 19 Qxd6 (Diagram 28)

Diagram 27 (B)

Diagram 28 (B)

Black's opening has failed

Black's problems persist

Black is still in big trouble due to his weak dark squares and undeveloped queenside. 19... Qc7 20 Nbs!?

White's initiative persists even after a queen exchange. 20... Qxd6 21 Nxd6 Rf8 22 Rc1!

Now White penetrates to c7. 22 ... es 23 Rc7+ Kf6 24 Rxh7 Be6 25 a3

White has won a pawn and now faces the technical challenge of converting his advantage. 25 ... b6 26 f3 (Diagram 29)

Diagram 29 (B)

Diagram 30 (W)

White has good winning chances

The Double Fianchetto variation

303

Starting Out: The Accelerated Dragon 26... Bg8?

This hastens the end. 26 ... Rfd8 would have put up more resistance. 27 Ne4+ Kfs 28 Re7!

White is very close to constructing a mating net, while Black cannot offer an exchange of rooks due to a knight fork. 28 ... Rfd8 29 Re1 RdS 30 Kf2 Rad8 31 Rxa7 1-0

Black is two pawns down with no counterplay, and his king is close to being mated.

Theoretical? The ... Nh6/... f7-f5 plan is not too popular but can lead to obscure and complicated positions, so both sides will benefit from a certain amount of theoretical knowledge.

Statistics Black scores 32% from the main position in Diagram 23 (after 9 .. .£5). The line with 8 f3! scores just 53% for White, but that is mainly due to a failure to follow up with the energetic c4-c5 at the required time. The lesson is clear: White can indeed obtain a good position after 8 f3, but only if he continues correctly.

The Double Fianchetto Variation 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3

o-o 8 Be2 b6!? (Diagram 30)

The Double Fianchetto system is a solid and reliable choice. Black is not doing anything too unorthodox - he simply develops his bishop on a sensible square -but he can sometimes struggle to create meaningful counterplay, and for that reason it has never enjoyed the same level of popularity as the main lines. 9 0-0 Bb7 (Diagram 31)

Diagram 31 (W) White must be careful

304

Diagram 32 (W) White is already in deep trouble

Mar6czy Bind: Sidelines At this point White has tried a number of moves, with 10 f3 being by far the most popular. First we will consider a couple of natural yet inferior options which can swiftly lead White into trouble.

WARNING: Aside from being strategically sound, the Double Fianchetto variation is also a 'trappy' line and I would strongly urge players of both sides to make a mental note of the variations contained in the following section.

Two Traps that White must Avoid The first well-known trap occurs after the natural developing move: 10 Rc1? This leads to serious problems after ... 10... Nxd4 11 Bxd4 Bh61 (Diagram 32) In Diagram 32 White already faces serious problems. The clever point of Black's last move is that, in addition to attacking the white rook, he also sets up a threat to capture the white e-pawn, as White will no longer be able to win a piece with Bxg7 and Qd4+. 12 es Nothing really works here. 12 f4 Nxe4 wins a pawn, while 12 Bxf6 Bxcl also leaves Black clearly better. Quite a few players have tried the text, hoping to muddy the waters by sacrificing the exchange, but it is hardly sound. 12 ... Bxc113 exf6 No better is 13 Qxcl NeB. 13 ... Bg5 13... Bxb2!? may be even stronger. 14 fxe7 Qxe7 15 Bf3 Bc6 16 Bds Rae8 White had practically no compensation for the exchange in F.Holzke-E.Pigusov, German League 1994. A slightly more subtle mistake is: 10Qd2? This time the 'refutation' is harder to spot. Objectively the text is a less severe inaccuracy than 10 Rcl?, but Black can obtain an easy game with chances of an advantage after. .. 10... Nxd4! 11 Bxd4 eSI (Diagram 33) Once again the idea is to take on e4 under favourable circumstances. White has no choice but to capture, as 12 Be3? Nxe4 would just leave him a pawn down for nothing. 12 Bxes Nxe4 13 Nxe4 Bxes (Diagram 34) This position should be equal, but White has a smaller margin for error, as evidenced by Black's score of 67% from Diagram 34. The two bishops are very powerful, whereas the isolated d-pawn is less important and can easily be guarded by a bishop on c6. 14Nd6 Bc6 14 ... Bxd615 Qxd6 Qg516 f3 Rfe8 17 Rf2 Bc6 is equal; the text is slightly more ambitious.

305

Starting Out: The Accelerated Dragon 15 f4!?

Diagram 33 (W) Another opening trap

Diagram 34 (W) White must struggle for equality

This looks like the correct decision; White should play actively, otherwise he will have difficulty defending his weaknesses after ...Qf6 or ... Qe7. Now Black can choose between 15 ...Bxd616 Qxd6 Re817 Bf3 Re618 Qd4 (D.Rakhimov-S.Nadyrhanov, Uzbeki Ch. 1993) 18 ...Qf6 with equality, or the more ambitious 15...Bg716 Rad1 Qf6 17 b3 Qb2 18 Bf3 Qa3 which led to him to an eventual victory in T.Horvath-M.Turner, Davos 2004, though the position is unclear at this stage.

Theoretical? There is not a lot of theory to know here, but if you intend to play either side of the Double Fianchetto then you should definitely know to avoid these early inaccuracies with White and exploit them as Black.

Statistics Black has done very well in both of these variations. After 10 Qd2? Nxd4! 11 Bxd4 e5! he has achieved 66%, while after 10 Rcl? Nxd4 11 Bxd4 Bh6! the tally rises to 71%.

Correct Play from White Here we will consider what can happen when White avoids the two traps mentioned above. 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 b6 9 0-0 Bb7 10 f3 (Diagram 35)

In light of the previous section, in which Black was allowed to exploit the tactical vulnerability of the e4-pawn, the text should require no explanation. Indeed, now that Black has committed his bishop to b7 it is hard to think of anything more natural for White than fortifying the long diagonal with his pawns. Although 10 f3 has been the overwhelmingly most popular move here, it is not the only decent option and the alternatives will be examined in Game38.

306


Diagram 35 (B)

Diagram 36 (W)

White plays solidly

Black intends ... Nf4

At this point Black must make an important decision regarding a plan for the middlegame. Against the majority of 'sensible' moves- such as 10 ...d6, 10...Rc8, etc- White should improve his position with the standard Qd2, Rfd1, Rac1 and so on. Black's position is fairly solid, but he may have a hard time achieving any active play. Therefore, I will pay special attention to three specific ways in which Black may attempt to gain counterplay. One principled reaction is 10...e61?, intending to fight for the centre with ... d7-d5. This at least forces White to deal with some concrete problems, and will form the subject of Game 37. Another popular move has been 10... Nh5!? (Diagram 36). Black's idea is to bring the knight to f4, relying on the tactical vulnerability of the d4-knight to activate or perhaps just exchange a few of his minor pieces. He may occasionally follow with ... f7-f5, though this decision should never be taken lightly. Most games have seen one of two reactions: a) Perhaps the most straightforward method for White involves the completion of development with 11 Qd2, after which 1l...Nf4 can be met by 12 Bxf4 Nxd4 13 Bd3. White intends to continue with Rfd1 and Bfl, or perhaps Rad1 and Bb1, followed by chasing the knight away from d4 while retaining the usual slight advantage. b) Another logical and slightly more ambitious continuation is 11 Nxc6!?. The entire ... Nf4 plan was based on the weakness of the knight on d4, and so by exchanging this piece White hopes to expose Black's lOth as a waste of time. In J.Nunn-T.Ristoja, Malta Olympiad 1980, the continuation was ll ... Bxc612 Rcl with a slight advantage. At this point Black understandably opted for 12.. .f5!?, as otherwise it would be very hard to justify his lOth. Nevertheless, after 13 exf5 gxf5 14 f4 Nf615 Bf3 Rc816 b3 Qe817 Nd5 the advance of the f-pawn had done little to help Black's chances. Perhaps a more principled line for Black is 11 ... dxc61?, hoping to equalize in the position with symmetrical pawns. Now the most challenging move is 12 c5!? (Diagram 37). I believe that White should be able to maintain a slight edge here: he controls more space,

307

Starting Out: The Accelerated Dragon while Black's minor pieces on b7 and hS are not brilliantly placed. Play may continue 12 ... bs 13 f4 b4 (in B.Lalic-A.Martin Gonzalez, Benasque 1996, Black tried 13... Nf6?!, when 14 eS! would have been very good for White) 14 Na4 Nf615 Qb3!? and now: b1) 15 ... Nxe4?? drops a piece after 16 Qxb4.

Diagram 37 (B)

Diagram 38 (W}

An irregular Mar6czy position

White must formulate a plan

b2) lS ... QaS (W.Meijer-R.Ralls, Correspondence 2003) 16 a3! Nxe417 axb4 Qc718 Nc3 leaves White with a better structure after 18 ... Nxc319 bxc3 or 18 ... Nf619 Bd4. b3) 15... Rb816 Qc4 Bc817 Rad1 Qa518 b3 when White kept a slight edge in Y.RazuvaevA.Dunnington, Cappelle Ia Grande 1995. Finally, Black has a very interesting option in 10...Qb8!?. It may look bizarre to place the queen on this square, but Black is planning an ingenious regrouping with ... d7-d6, ... Rd8-d7, ...Qf8 and ... Rad8, to be followed by ...e7-e6 and finally ... d6-d5! This may seem awfully timeconsuming, but the position is relatively closed and it is not immediately obvious what White should be doing. In some ways a long-winded plan can be an advantage in semi-closed positions, as it ensures that one will be playing with a definite purpose, rather than merely shuffling the pieces around. Following the natural11 Qd2 Rd8 12 Rfd1 d6 (Diagram 38}, it is time for White to decide on a plan. Most games have continued with 13 Racl Rd7, when Black proceeds as described above. White is objectively a bit better here, but it is hard to know exactly what to do and I would instead advise players to investigate a more direct approach. a) In I.Herrera-J.Bellon Lopez, Santa Clara 1997, White activated his pieces in the centre with 13 NdS Rd714 Nxc6 Bxc615 BgS Qd8 16 Nxf6+ Bxf617 Bxf6 exf6 (Diagram 39). Here White has a permanent advantage due to Black's damaged structure, but it is very hard to win such positions. White's bishop is ineffective, while Black only has a single weakness on d6. b) K.Arakhamia Grant-T.Canela, Novi Sad Olympiad 1990, was a model demonstration of what may be White's most effective strategy. Given that Black is planning to transfer all his heavy pieces to the kingside and the centre, it is highly logical for White to focus on the

308

Mar6czy Bind: Sidelines queenside. She did this with 13 Nb3!? intending to attack with a4-aS. The game continued 13 ...Rd7 (13 ...e614 a4 NaS!? IS NxaS bxaS16 NbS dS is recommended by Nielsen and Hansen, but 17 QxaSlooks extremely good for White, threatening both Bxa7 and Nc7) 14 a4! Qd81S aS bxaS 16 NbS Ba617 NxaS BxbS18 cxbS NxaS 19 QxaS QxaS 20 RxaS (Diagram 40) when White's strategy had worked to perfection. The game lasted just a few more moves: 20... Rdd8 21 Bxa7 Nd7 22 b6 Rac8 23 b7 Rc2 24 BbSl-0. Black's defence may not have been perfect, but it is not easy to find a fully satisfactory antidote to the advance of the white a-pawn.

Diagram 39 (W)

Diagram 40 (B)

Black should hold

White wins easily

Theoretical? The Double Fianchetto variation is not overly theoretical and the majority of White players are unlikely to know much about it. If the reader intends to play the Mar6czy with White then he should definitely remember to avoid 10 Rcl? and 10 Qd2?. Black needs to know the refutations of those two moves, as well as at least one playable line against 10 f3 and White's other 'good' lOth moves seen in Game 38.

Statistics The Double Fianchetto has scored a reasonable total of 43% after 8 ... b6; not bad considering that the more mainstream 8 ... d6 has yielded one percent less. Following the usual 9 0-0 Bb7 10 f3, the main lines of 10... e6, lO ... NhS and 10 ... Qb8!? have scored 42%, 41% and 44% respectively.

Illustrative Games Game 37

D M.Marin •

V.Ciashimov

Spanish Team Championship 2007 1 c4 cs 2 Nf3 g6 3 d4 cxd4 4 Nxd4 Bg7 5 e4 Nc6 6 Be3 Nf6 7 Nc3 0-0 8 Be2 b6 9 0-0 Bb7 10 f3

309


e6!? (Diagram 41)

Diagram 41 (W)

Diagram 42 (W)

Black intends the ... d7-d5 break

Can Black regain the d-pawn?

This is a direct attempt to solve Black's problems by breaking in the centre with ... d7-d5. 11Qd2 The most common move. White makes no attempt to prevent ... d7-d5; instead he plans to allow it and respond with c4xd5 and e4-e5, hoping to gain a positional advantage based on his control of the dark squares in the centre. This plan has scored fairly well, though it is not necessarily be the best. Here are a couple of other ways in which White may fight for an advantage. a) After 11 Ndb5!? Black must, for better or worse, carry through with his intended 11...d5 (ll ... Ne8 is too passive) 12 cxd5 exd5 13 exd5 Nb4. The question is whether or not he will be able to regain his pawn without suffering any unpleasant consequences. A.BachmannH.Skarke, Correspondence 1985, continued 14 d6 Nfd515 Nxd5 Nxd516 Bd4 Qg5 (Diagram 42) [16 ... Bh6!? has also been tried, when 17 Re1!? may be White's best]17 g3!? Be5 (17... Ne3 can be met by 18 f4! Qd5 19 Bf3 Qxb5 20 Bxe3) 18 Bxe5 Qxe519 Re1 Qe3+ 20 Kg2 Qc5 21 Qd2 Ne3+ 22 Kh1 Nc4 23 Qf4 Qxb5 24 Bxc4 Bxf3+ 25 Qxf3 Qxc4 26 b3 Qc5 27 Rad1 Rad8 28 d7 when the strong passed pawn gave White some advantage despite the equal material. b) 11 Rcl!? aims for an improved version of the main line and may be the best try for an advantage. White hopes that after a subsequent ... d7-d5, c4xd5, the placement of his rook on c1 will prove more useful than the queen on d2.

0

NOTE: At the risk of stating the obvious, this move carries none of the drawbacks of the disastrous 10 Rc1?, as the e4-pawn is already securely defended.

Play may continue ll ... d5 (1l...Nxd4 12 Bxd4 Bh6 13 Rc2 gave White a slight but stable advantage in L.Polugaevsky-J.Bellon Lopez, Logrono 1991) 12 Nxc6 Bxc613 e5 Nd714 cxd5 exd5 15 f4. Compared with the analogous 11 Qd2 variation, White's chances are slightly better due to the vulnerability of the black bishop on the open c-file. 11 ... ds 12 cxds exds 13 Nxc6 Bxc6 14 es (Diagram 43)

310


Diagram 43 (B)

Diagram 44 (W)

Where to retreat the knight?

The IQP is a long-term weakness

14... Ne8! An important moment. 14... Nd7 may seem more natural, but the text has been almost a universal choice amongst Grandmasters with Black. One advantage of the e8-square is that the knight can come to c7 to provide a solid support of the d5-pawn. After the inferior 14 ... Nd715 f4 Nc516 Rad1! looks best, when 16... Ne417 Nxe4 dxe418 Qc2 Qe819 Rd4 (19 Ba6!? is also good) 19... Rc8 20 Rcl Bb7 21 Rc4left Black facing a difficult defensive task in A.Dgebuadze-K.Znamenacek, Pardubice 1993, as the e4-pawn is a long-term weakness. 15 Bd41? This was a novelty, but the move is very logical. After the hitherto universal choice of 15 £4 Black would always respond with 15.. .£6! (just as in Chapter Two, it is vital to destroy White's central pawn wedge to liberate the Dragon bishop), and now White must decide whether the inevitable pawn exchange will take place on £6 or e5: a) 16 Bd4 fxe5 17 Bxe5 (17 fxe5 Nc718 Rx£8+ Qx£8 19 Bg4 Qe7 20 Re1 Re8 was equal in O.Loskutov-M.Ekdyshman, Serpukhov 2003) 17... Bxe518 fxe5 Nc7 should be fine for Black. His pieces are quite well placed and the e5-pawn could become weak. b) 16 exf6looks better, when Black replies 16 ... Nxf6. In the main game we reach something very similar, except that the white £-pawn is back on f3 instead of £4. This difference should favour White, as the e4- and g4-squares are defended and his kingside is generally more secure. Nevertheless, despite the slightly loosening effect of the pawn on f4, White should be a little better thanks to Black's isolated d-pawn. V.Nevednichy-V.Gashimov, European Team Ch., Gothenburg 2005, continued 17 Rad1 Qe718 Bd4 Rad8 19 B£3 Ne4 20 Qe3 Bxd4 21 Rxd4 Qf6 22 Rfd1 Rde8 23 Nxd5 Bxd5 24 Rxd5 Qx£4 25 Qx£4 Rxf4 26 Re1 N£6 27 Rxe8+ Nxe8 28 Rd8 R£8 29 Bd5+ Kg7 30 Rd7+ Kh6 31 Rxa7, when White's extra pawn and superior pieces should have been a decisive endgame advantage. 1S ... Nc7 16 Rad1 f6 17 exf6 Bxf6 18 Rfe1 Qd6 19 Bxf6 Qxf6 (Diagram 44) White has a slight advantage due to the isolated d-pawn. He controls the blockading square on d4 and can slowly aim for a favourable endgame. In this game Gashimov defends very

311

Starting Out: The Accelerated Dragon well (he is, after all, a very strong Grandmaster), but I would advise the average player to stay away from this type of position and instead investigate one of the alternatives on move 10. 20 BbS!

A clever way of activating the bishop. Black cannot win a piece due to the potential fork on dS. 20... Bb7!

Correctly avoiding exchanges. TIP: Generally speaking, in positions with an isolated pawn, the side playing against the isolani should try to exchange minor pieces- knights in particular. 21 Ba4

Rerouting the bishop to b3 where is will play an active role in pressurizing dS. 21 ... Rad8 22 Qf2 Kg7 23 Rd2 Rf7 24 Red1 Re7 25 Bb3 Red7 26 Ne2

White would like to play 26 Qd4!?, but perhaps he was put off by the line 26 ... Qxd4+ 27 Rxd4 Ne6 28 R4d2 Nc5, intending to weaken White's pawns with ... Nxb3. 26... Ne6 27 Nd4 Ncs 28 Bc2 Re7

The game is in a manoeuvring phase with both sides trying to optimize their pieces. White's position is always a little more comfortable, but Black is never in any real danger and in the end it is White who is content to force a draw. 29 Qg3 Kh8 30 h4 Rf8 31 Re1 Qf4 32 Qf2 Rfe8 33 Rxe7 Rxe7 (Diagram 45)

Diagram 45 (W} Black has played well

Diagram 46 (B) Early aggression

34Re2

Black has become quite active, so Marin settles for a draw after the following forced sequence. 34...Qc1+

312

Mar6czy Bind: Sidelines 34 ...Rxe2? 35 Nxe2 Q£6 36 Qd4 Kg7 37 Kf2 would give White a pleasant endgame advantage. 35 Kh2 Rxe2 36 Qxe2 Qf4+ 37 g3 Qxd4 White can afford to drop the knight as there is an instant perpetual. 38 Qe8+ Yz-Yz Game38

D K.Sasikiran • L.Nisipeanu Sofia 2007 1 Nf3 c5 2 c4 g6 3 d4 cxd4 4 Nxd4 Nc6 5 e4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 b6 9 0-0

If White intends to advance his £-pawn aggressively then he can also consider the immediate 9 £4!?. The advantage of doing this a move sooner is that Black is unable to play ... d7-d6, though it is hard to say how important this is, given that Nisipeanu chooses not to bother with 10 ...d6 even when he has the chance. 9... Bb7 10 f4!? (Diagram 46)

Here we see an early attempt from White to play actively and aggressively while avoiding any mainstream theory. Before continuing we will briefly consider White's other main options (as the reader will recall, we have already looked at the weak moves 10 Rcl? and 10 Qd2?, as well as the main line of 10 £3). a) 10 Nxc6!? is a reasonable move, when 10... Bxc6 (10 ... dxc6 11 e5 followed by f2-f4 gives White a pleasant edge due to his space advantage and more active pieces) 11 f3 d6 12 Qd2 Nd7 has been played in several games, and leads to something resembling the Classical Maroczy from Chapter Six, in which the black bishop usually ended up coming to c6 after an exchange of knights on d4.

0

NOTE: Compared with the main lines of the Classical Mar6czy, in the present position ... b7-b6 has slightly weakened the black queenside. This may not seem very important at the moment, but it could become a telling factor.

For example, in the event of Nc3-d5, the reply ... Bxd5 may be met by c4xd5 followed by Ba6 to gain control of the c-file, or Bb5-c6 occupying a powerful outpost. b) White can also try 10 Nc2!?, reaching a position that should be compared with Chapter Nine. c) 10 Ndb5?! has scored very highly, though if White wishes to avoid a knight exchange it is hard to believe that the text can be an improvement over 10 Nc2. Black should proceed with 10... d6 and perhaps ... Nd7-c5 with a solid position. WARNING: When White plays a move like Ndb5, he is probably trying to provoke Black into ...a7-a6 which would weaken the b-pawn. This may not prove to be a problem, but one should always think carefully before playing such pawn moves. 10... Nxd4!?

10... d6 would be the normal move. In M.Rodin-V.Kron, Moscow 1991, Black played more creatively with 10... e5!? 11 Ndb5 exf4 12 Bxf4 Ne8 13 Qd2 a6 14 Nd6 Nxd6 15 Bxd6 Re8 and

313

Starting Out: The Accelerated Dragon while he eventually won the game, 16 Rad1 would have left White clearly better at this stage. 11 Bxd4 ReS!?

Black continues to play provocatively. Most players would have opted for ll ... d6, but Nisipeanu is not concerned about the advance of thee-pawn. 12Qd3

12 e5 could be met by 12... Ne4!, exchanging knights and intending to follow with the undermining move ... d7-d6. 12 ... d6 13 Rad1 Qc7 14 Kh1 Rfe8?

This is pushing the provocation a step too far. 14... Rfd8 would have discouraged the following pawn advance. 15 es! dxes 16 fxes Nhs (Diagram 47)

Diagram 47 (W)

Diagram 48 (B)

Extremely provocative play

An elegant mating combination

16... Nd7 would have been met by 17 e6. 17 NdS?

White is better after this, but he could have obtained an overwhelming advantage with 17 Bxh5! gxh5 18 e6!, when 18.. .f619 b3leaves White with a huge positional plus. But the really elegant point is seen after 18.. .fxe6?! 19 Bxg7 Kxg7 20 Rf7+! (Diagram 48) 20 ... Kxf7 21 Qxh7+ Kf6 22 Qh6+ Kf7 (the king has to go back, as 22 ... Ke5? 23 Qg5 is mate) 23 Rfl+ Kg8 24 Qg6+ Kh8 25 Qxh5+ Kg7 26 Qg5+ Kh8 27 Rf7 Bxg2+ 28 Qxg2 Qc6 29 Ne4 and mate is imminent. 17 ... Bxds 18 cxds Bxes 19 Bxhs Bxd4 20 Qxd4 gxhs 21 Rd3! (Diagram 49)

Black's position looks highly precarious, but Nisipeanu manages to hold on with some computer-perfect defence. 21... es1 The only move to survive. 22Qh4

314

Mar6czy Bind: Sidelines After 22 dxe6 Rxe6 Black is in little danger, as his rook will find a perfect home on g6 from where it will combine attack and defence 22 ...Qd6!

Once again this is the only good move. 23 Qxhs

White's attacking ambitions are hampered by his weak back rank, so for the time being Sasikiran settles for recapturing his missing pawn while preparing to target Black's h-and f-pawns. 23 ... Kh8 24 Rg3 f6 25 Qh6 Qf8 26 Qh4 RedS 27 h3 (Diagram SO)

Diagram 49 (B)

Diagram so (B)

Black can just hang on

Black has survived

Following Black's accurate defence the game now peters out to a level ending. 27 ... Rxd5 28 Rg6 Rd4 29 Rgxf6 Rxh4 30 Rxf8+ Rxf8 31 Rxf8+ Kg7 32 Rf3 Rb4 33 b3 as

Usually the players might have agreed a draw here, but in Sofia draws are forbidden until an expert arbiter agrees that there is absolutely no play left in the position. Personally I rather like this rule and would love to see draw offers banned from chess altogether! 34 Kh2 e4 3S Rfs a4 36 bxa4 Rxa4 37 Rbs Rxa2 38 Rxb6 Ra3 39 Re6 Re3 40 Kgl Kf7 41 Kf2 Rxh3 42 Rxe4 Yz-Yz

Summary and Conclusions As I explained in the introductory text, once White has opted for the Mar6czy Bind he has very few good ways of deviating from the usual paths. The Mar6czy is very much a classical system, and it is obvious that White's best strategy is to develop his pieces on their natural squares and castle short, with the only major decision concerning the possible retreat of the knight from d4 to c2. I felt it was worthwhile to take a brief look at the 'Mar6czy-Yugoslav' attack, purely for the purposes of demonstrating how Black should react. Under no circumstances would I recommend playing this way with White. If you want to play a Yugoslav attack then turn back to Chapters Three and Four!

315

Starting Out: The Accelerated Dragon The story for Black is much happier, at least insofar as having the freedom to choose from a number of interesting alternatives to the 'big three' systems. We have examined a full spectrum ranging from the immediately shocking 5... Bh6!?, to the relatively mainstream Double Fianchetto, to the risky counterattack with ... Nh6 and ... f7-f5. There really is something for everyone. Of the lines covered in this chapter, the one carrying the greatest surprise value is undoubtedly 5... Bh6!?. This is positionally well motivated, as Black often attempts to exchange these bishops in the Mar6czy. The only drawback is that he tends to end up in a slightly passive position, and can struggle to generate winning chances. Still, if the reader is looking for a way to surprise an opponent while reaching a solid middlegame, then 5 ... Bh6!? would be the perfect choice. The system featuring an early ... Nh6 and ... f7-f5 is playable, and may well appeal to those who like to disrupt the equilibrium at an early stage. Ultimately, however, I do not feel I can recommend it as a fully respectable alternative to the main lines, and the notes to this section and the illustrative game Hiibner-Afifi highlight the potential drawbacks to Black's approach. While the ... f7-f5 system might make for a decent surprise weapon, I would advise the reader to pick a sounder line as the mainstay of his repertoire. Of the three systems mentioned, the Double Fianchetto is the most respectable and could easily have formed the subject of a whole chapter. This set-up is certainly a worthwhile choice for players looking to base their repertoire on a sound yet non-mainstream opening system. It is always nice to know that there are a couple of traps lurking, and the weak moves 10 Rcl? and 10 Qd2? are seen time and time again. After the superior 10 f3 Black is slightly worse, but there are plenty of ways to reach an acceptable position and I hope to have covered enough ground to enable most players to find something that appeals to them.

Final Word Thus we reach the end of the present chapter and the book. I hope that I have succeeded in demonstrating the Accelerated Dragon to be a fun and stimulating opening to play with either colour. Thank you for reading, and I wish you great success in your games on both sides of this great opening.

316

Index of Variations Chapter One: The Hyper-Accelerated Dragon 1 e4 C5 2 Nf3 g61? 10 3 d4 (3 h4!? 17; 3 Bc4 19; 3 c3 25 d5!? 4 exd5 Qxd5 5 d4 Bg7 26 6 Na3! 27) 3... cxd4 (3 ... Bg7!? 12) 4 Qxd41? 34 (4 Nxd4 Nc6- 2 ... Nc6) 4... Nf6 5 Nc3!? 44 (5 Bb5!? 34; 5 e5 38) 5... Nc6 6 Qa4 d6 (6 ... Bg7!? 45) 7 e5 45 Chapter Two: 5 Nc3- Classical and Other Lines 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 50 (6 Nde2!? 50; 6 Nb3 59) 6... Nf6 7 Be2 (7 Bc4- Chapter Three; 7 Nxc6- Chapter Five) 7...0-0 8 0-0 (8 Nb3!? 61; 8 £4!? 66 d5!? 67) 8...d5! 9 exd5 (9 Nxc6 bxc6 10 e5 53) 9... Nb4!? 54 (9 ... Nxd5 54) Chapter Three: Yugoslav Accelerated Dragon: Introduction and 7...Qa5!? 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! (7 £3?! 0-0 8 Qd2 d5! 75; 8 Bc4 Qb6! 79) 7...Qa5!? 83 (7 ... 0-0- Chapter Four) 8 0-0 0-0 9 Bb3 86 (9 Nb3 85 Qc710 £4 86; 10 Bg5!? 90) 9... d6 10 h3 Bd7 11 Re1 94 (11 £4 97) Chapter Four: Yugoslav Accelerated Dragon: Main Line with 7... 0-0 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 Bg7 6 Be3 Nf6 7 Bc4! 0·0 8 Bb3! 104 (8 £3?! Qb6! 102; 8 Qd2?! Ng4! 102; 8 0-0 Nxe4! 103) 8... a5!? (8 ...Qa5!? 104; 8 ... e6!? 106; 8 ... d6 109 9 £3 Bd7 10 h4! 110 h5!? 111) 9 f3 (9 a4 Ng4! 114; 9 0-0!? 115) 9... d5! 11610 Bxd5! 119 (10 Nxd5? 117; 10 exd5 Nb4 117) 10... Nxd5 11 exd5 (11 Nxd5 £5! 123) 11... Nb412 Nde2 Bf5 13 Rc1 b5! 14 0-0 (14 a3 Bxc2! 125) 14... Rc815 Nd4 Bxd41? 120 (15 ... Rxc3 127) Chapter Five: Lines with Nxc6 1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 Nc3 (5 Nxc6 bxc6 6 Qd4 132) 5... Bg7 6 Be3 Nf6 7 Nxc6!? bxc6 8 e5 133 Ng8140 (8 ... Nd5!? 134 9 Nxd5 cxd510 Qxd5 Rb8 11 0-0-0 134; 11 Bxa7!? 135; 11 Bc4 136) 9 Bd4! 141 (9 £4 Nh6! 140) 9...Qa5!? 143 Chapter Six: Maroczy Bind: Classical Variation 1 e4 C5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 (8 ... b6!?Chapter Ten) 9 0·0 148 Bd7 (9 ... Nxd4!? 152 10 Bxd4 Be6?! 154; 10... Bd7 11 Qd3!? 152; 11 Qd2 Bc6 12 £3 a5- 9 ... Bd7) 10 Qd2 (10 £3 154; 10 £4!? 155; 10 Nc2!?- Chapter Nine) 10... Nxd411

317

Starting Out: The Accelerated Dragon Bxd4 Bc6 156 12 f3 (12 Bd3!? 158) 12... as 13 b3! 162 (13 Kh1!? 157) 13 ... Nd7 14 Be3 (14 B£2 173) 14... Ncs 15 Rab1 Qb6 162 (1S ...e6!? 162) 16 Rfc1 Rfc8 17 Rc2! (17 a3? 163) 17...Qd8 (17 ...Qb4!? 170) 18 Bf1167 (18 a3 164) Chapter Seven: Mar6czy Bind: Gurgenidze System 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Nf6 6 Nc3 d6!? 178 (6 ... Nxd4!? 7 Qxd4 d6181) 7 Be2 (7 £3!? 182 Nxd4 8 Qxd4 Bg7 9 Be3 0-0 10 Qd2 Be6 183; 10... aS!? 185 11 Be2 a4! 12 0-0 QaS 186; 12... Be6! 188) 7 ... Nxd4 8 Qxd4 Bg7

-9 o-o 191 o-o 10 Qe3 191 (10 Qd3 193) -9 Be3 194 0-0 10 Qd2 Be611 0-0 (11 Rcl QaS 12 £3 Rfc8 13 b3 a6 14 Na4 Qxd2+ 1S Kxd2 Nd7 219) 11... Qas 12 Rac1 (12 Rfcl Nxe4! 197; 12 £3 Rfc8 197; 12 Rab1!? Rfc8 13 b3 Ng4!? 199; 13 ... a6 199) 12... a6! 195 (12 ... Rfc813 b3 a614 £4!? 195; 14 f3 -12 ... a6) 13 f3 (13 £4 bS! 196) 13 ... Rfc8 14 b3 bs! 180 - 9 Bgs 203 o-o 10 Qd2 Be6 11 Rc1 206 (11 0-0 QaS 205) 11...Qas (1l...a6!? 206) 12 f3! Rfc8 13 b3 a6 (13 ... Rab8!? 214; 13 ... Kf8!? 215) 14 Na4! 208 (14 NdS 209) 14... Qxd2+ (14 ... Qd8!? 209 1S cS?! Nd7! 209; 1S 0-0! 210) 15 Kxd2 212 Rc6 215 (1S ... Nd7 219) Chapter Eight: Mar6czy Bind: 7... Ng4 1 e4 cs 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 6 Be3 Nf6 7 Nc3 Ng41? 227 8 Qxg4 (8 Nxc6? 227) 8 .•. Nxd4 (8 ... Bxd4?! 228) 9 Qd1 Ne6 242 (9 ...eS!? 229 10 NbS!? 0-0! 11 Nxd4? 230; 11 Qd2 Qh4 231; 1l...Qe7!? 234; 10 Bd3! 235 0-0 11 0-0 d6 236 12 a4! 237 a6 239) 10 Rc1! 244 (10 Be2?! Bxc3+! 243; 10 Qd2 242) 10...Qas!? 248 (10 ... b6 245; 10 ... d6 246) 11 Bd3! 258 (11 Qd2 249; 11 Be2 254) Chapter Nine: Mar6czy Bind: Avoiding Exchanges with Nc2 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 c4 Bg7 6 Nc2!? (6 Be3 Nf6 7 Nc3 0-0 8 Be2 d6 9 0-0 Bd710 Nb3!? 281; 10 Nc2!? 281 a6!? 11 f3 ReS 12 Qd2 282; 12 Rcl!? 284) 6 ... d6 (6 ... Nh6!? 266) 7 Be2 Nf6 8 Nc3 Nd7 (8 ... 0-0 9 0-0 Be6!? 263) 9 Bd2 (9 0-0!? Bxc3! 10 bxc3 NcS 265) 9 ... 0-0 10 o-o Ncs (10 ... aS!? 271) 11 b4! 268 Ne6 (1l...Nd7!? 278) 12 Rc1 as!? (12 ... Ned4 275) 13 a3 axb4 14 axb4 Ned4 (14 ... Nf4!? 270) 15 Nxd4 Nxd416 Be3 269 Chapter Ten: Mar6czy Bind: Sidelines 1 e4 cS 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 g6 5 C4 Bg7 (S ... Bh6!? 292 6 Bxh6 Nxh6 7 Nc3 0-0 8 Be2 d6 9 0-0 293; S... N£6 6 Nc3 d6 7 f3 Nxd4 8 Qxd4 Bg7 9 Be3 0-0 10 Qd2 aS!? 11 g4?! 291; 11 Bh6?! 291) 6 Be3 - 6... Nf6 7 Nc3 0-0 8 Be2 (8 f3 d6 9 Qd2 Bd710 g4?! 291) 8 ... b6!? 304 9 0-0 Bb7 10 f3 (10 Rcl? 305; 10 Qd2? 305; 10 Nxc6!? 313; 10 £4!? 313) 10... e6!? 310 (10 ... NhS!? 307; 10 ... Qb8!?

308) - 6 ... Nh6!? 296 (6 ... d6 7 Nc3 Nh6 8 f3! 297) 7 Nc3 (7 f3 Qb6!? 298) 1 ...0-0 8 Be2 (8 f3!? 298) 8 ... d6 (8 .. .£S!? 300) 9 o-o fs 10 exfs Bxd4 301 (10 ... NxfS 301; 10... gxfS 301)

318

Index of Complete Games Amonatov.F-Inarkiev.E, Russian Team Championship 2007 ............................................. 97 Aschenbrenner.R-Krebs.J, Austrian League 2006 .............................................................. 136 Ashton.A-Williams.S, Hastings 2006/07 .............................................................................. 173 Bacrot.E-Tiviakov.S, Wijk aan Zee 2006 .............................................................................. 275 Bauer.C-Colin.V, French Championship, Besancon 2006 ................................................... .40 Ciuksyte.D-Greet.A, Southend 2006 .................................................................................... 170 Degraeve.J.M-Lerch.P, French Team Championship 2005 ................................................. 86 Drzymala.G-Caba.T, Polish Junior Championship 2004 ................................................... 254 Filippov.V-Nielsen.P.H, Minsk 1996 ..................................................................................... 164 Geller.J-Savchenko.B, Russian Junior Championship 2006 .............................................. 124 Ghisi.S-Giadis.D, Correspondence 2004 .............................................................................. 143 Grischuk.A-Gurevich.D, World Blitz Ch., Rishon Le Zion 2006 ....................................... 111 Gurevich.V-Chatalbashev.B, Cappelle la Grande 1998 ....................................................... 61 Hamdouchi.H-Kudrin.S, FIDE World Championship, Tripoli 2004 .................................. 29 Hamdouchi.H-Tiviakov.S, Wijk aan Zee 2004 ..................................................................... 157 Hasangatin.R-Vokarev.S, Maikop 1998 ................................................................................. 21 Hintikka.E-Koponen.A, Kuopio 1989 ..................................................................................... 67 Hubner.R-Afifi.A, Manila Interzonal1990........................................................................... 299 llczuk.J-Kristoi.L, Correspondence 2003 .............................................................................. 258 lvanovic.B-Cebalo.M, Yugoslav Championship 1989 ....................................................... 127 Jain.A-Greet.A, Crowthorne 2006 .......................................................................................... 54 Karjakin.S-Ivanchuk.V, European Championship, Warsaw 2005 ...................................... 90 Karpov.A-Kavalek.L, Nice 1974 ............................................................................................ 213 Karpov.A-Larsen.B, Brussels 1987 ........................................................................................ 249 Marin.M-Gashimov.V, Spanish Team Championship 2007............................................. .309 Minasian.Art-Malakhov.V, European Club Cup, Fuegen 2006 ........................................ 278

319

Starting Out: The Accelerated Dragon Mitkov.N-Perelshteyn.E, Chicago 2006 ............................................................................... 122 Nijboer.F-Tiviakov.S, Dutch Championship, Leeuwarden 2003 ...................................... 219 Pavlovic.M-Cireet.A, Port Erin 2004 ..................................................................................... 167 Pioch.Z-Toma.K, Rowy 2002 ................................................................................................... 94 Rowson.J-Malakhov.V, Selfoss 2003 .................................................................................... 198 Rowson.J-Turner.M, Edinburgh 2001 ................................................................................... .45 Sasikiran.K-Nisipeanu.L, Sofia 2007 ..................................................................................... 313 Shabalov.A-Baburin.A, Mermaid Beach 1998 ..................................................................... 239 Shaposhnikov.E-Malakhov.V, Moscow 2007 ........................................................................ 14 Short.N-Felgaer.R, Najdorf Memorial, Buenos Aires 2001 ............................................... 284 Van Wely.L-Van der Wiei.J, Dutch Championship, Leeuwarden 2004 ........................... 271 Volokitin.A-Jakovenko.D, Russian Team Championship 2007 ........................................ 188

320

starting out:

II II

978-1-85744-530-5

45305

52595

I

CHESS Andrew Greet - Starting Out - The Accelerated Dragon (Everyman 2008)

Recommend Documents