Sicilian Kan

Citation preview

I

I

I

SICIIan an

by John Emms EVERYMAN CHESS Everyman Publishers pic www.everymanbooks.com

First published in 2002 by Everyman Publishers plc, formerly Cadogan Books plc, Gloucester Mansions, 140A Shaftesbury Avenue, London WC2H 8HD Copyright© 2002 John Emms The right of John Emms to be identified as the author of this work has been as­ serted 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 1 85744 302 0 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, Gloucester Man­ sions, 140A Shaftesbury Avenue, London WC2H 8HD tel: 020 7539 7600 fax: 020 7379 4060 email: dan@everyman. uk.com website:

www.everyman.uk.com

To Daniel

EVERYMAN CHESS SERIES (formerly Cadogan Chess) Chief advisor: Garry Kasparov Commissioning editor: Byron Jacobs Typeset and edited by First Rank Publishing, Brighton. Production by Book Production Services. Printed and bound in Great Britain by The Cromwell Press Ltd., Trowbridge, Wiltshire.

Everyman Chess Popular opening books: 1 1 1 1 1 1 1 1 1 1 1 1 1

85744 2 1 8 0 85744 253 9 85744 256 4 85744 232 6 85744 281 4 85744 292 X 85744 290 3 85744 242 3 85744 262 8 85744 291 1 85744 252 0 85744 257 1 85744 276 8

Unusual QG Declined Alekhine's Defence Queen's Gambit Declined French Classical Modern Defence Symmetrical English c3 Sicilian Offbeat Spanish Classical Nimzo-Indian Sicilian Grand Prix Attack Dutch Stonewall Sicilian Kalashnikov French Winawer

Chris Ward Nigel Davies Matthew Sadler Byron Jacobs Speelman & McDonald David Cummings Joe Gallagher Glenn Flear Bogdan Lalic James Plaskett Jacob Aagaard Pinski & Aagaard Neil McDonald

Books for players serious about improving their game: 1 1 1 1 1 1 1 1 1

85744 226 1 85744 231 8 85744 236 9 85744 241 5 85744 246 6 85744 223 7 85744 228 8 85744 233 4 85744 238 5

Starting Out in Chess Tips for Young Players Improve Your Opening Play Improve Your Middlegame Play Improve Your Endgame Play Mastering the Opening Mastering the Middlegame Mastering the Endgame Simple Chess

Byron Jacobs Matthew Sadler Chris Ward Andrew Kinsman Glenn Flear Byron Jacobs Angus Dunnington Glenn Flear John Emms

Books for the more advanced player: 1 1 1 1 1

85744 233 4 85744 233 4 85744 219 9 85744 224 5 85744 273 3

Attacking with 1 e4 Attacking with 1 d4 Meeting 1 e4 Meeting 1 d4 Excelling at Chess

John Emms Angus Dunnington Alexander Raetsky Aagaard and Lund Jacob Aagaard

I

I

I

SICIIan an

by John Emms EVERYMAN CHESS Everyman Publishers pic www.everymanbooks.com

CONTENTS

I

Bibliography

6

Preface

7

Introduction

9

1 e4 c5 2 l2lf3 e6 3 d4 cxd4 4 l2lxd4 a6 5 �d3

1

5 �d3 l2lf6 6 0-0 'f!ic7 7 'ii'e2 d6

2 5 �d3 l2lf6 6 0-0 'flic7: Seventh move alternatives 3

5 �d3 l2lf6 6 0-0 d6

4 5 �d3 .tc5 5 5 .td3: Fifth move alternatives for Black

12 46 60 82 112

5li:lc3

6

5 l2lc3 -vJ1c7

129

7

5 ltJc3 b5 6 �d3 'i'b6!?

156

8 5 l2lc3 b5: Sixth move alternatives

173

5 c4

9

5 c4

182

Index of Complete Games

191

I

BIBLIOGRAPHY

I

Books

Enryclopaedia of Chess Openings Volume B, 3rd Edition (Sahovski Informator 1997) Enryclopaedia of Chess Openings Volume B, 4th Edition (Sahovski Informator 2002) Nunn's Chess Openings, John Nunn, Graham Burgess, John Emms and Joe Gallagher (Every­

man/Gambit 1999) Siiflianisch im Geiste des !gels, Frank Zeller (Schachverlag Kania 2000) Winning with the Kan, Ali Mortazavi (Batsford 1996) Trends in the Paulsen VoL 2, John Emms (Trends 1997) Beating the Sicilian 3, John Nunn and Joe Gallagher (Batsford 1995) Periodicals

Chess Informants 1-84 The Week in Chess 1-405 Chesspublishing.com Databases

Mega Database 2002 Mega Corr 2

PREFACE

I

This book is a study of the Sicilian Kan (1 e4 cS 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4), an ideal opening for those playing Black who wish to learn the intricacies of the ever popular Sicilian Defence. I've been looking forward to writing this book for a long time. The Kan was the first Sicilian variation I learned to play and, because of this, I've always had a soft spot for it. I first became aware of its potential as a possible defence for Black in a typical fash­ ion: struggling to find an advantage against it with the white pieces! From the start I was very much struck by its flexibility and simplicity. A major selling point, which cannot be overstressed, is that the Kan is one of the easiest variations of the Sicilian to learn. Unlike some more high profile lines I could mention (the Dragon, the Na­ jdorf and the Sveshnikov), the onus is not on the player with the black pieces to memorise reams of opening theory simply to stay on the board. Of course Black still has to play good moves (!), but is much less likely to be at a disadvantage simply down to a memory loss. The Kan is more of a 'system' in that Black tends to react similarly regardless of how White plays it. This brings us to another positive fea­ ture; how does White play it? When I began

to use the Kan as my main weapon against 1 e4 I was struck by the number of strong and experienced white players who would become flummoxed and slump into deep thought early on. My basic theory is that the majority of those playing open Sicilians have budgeted just enough time to study the ins and outs of the 'trendy' Dragon, Najdorf and Sveshnikov, leaving them a little short against the 'less fashionable' lines. So who plays the Kan? Going back to the 1970s you would find games from the likes of Karpov, Petrosian, Portisch, Miles, Andersson, Hubner, Ljubojevic and Gheor­ ghiu. Of the new generation of top class players, there are games by Ivanchuk, Ru­ blevsky, Smirin and Judit Polgar, while Kas­ parov and Kramnik have also used it occasionally. Other players who have con­ tributed to its theory include Bologan, Vyz­ manavin, Milov, Epishin, Eingorn, Kengis, Landa, Ilya Gurevich, Miezis, Movsesian, Romanishin, Farago, etc. I could go on but there's enough quality on display there to dispel any doubts about the soundness of the defence. To what type of player does the Kan es­ pecially appeal? Anyone who believes that Black's sound Sicilian structure can over­ come White's early initiative. For a Sicilian, 7

Sicilian K a n

there is relatively little theory and still much uncharted territory. For example, a few years ago a new move was discovered for Black which has completely altered how a particularly important line was perceived. And when does this idea occur? As early as move six! The Kan is more likely to appeal to a player who is keen to think for himself from an early stage, rather than one who is reliant on the comfort zone twenty or so moves of solid theory. To this I would add to this that the Kan player generally plays on the counter-attack and that patie-nce, especially in certain variations, is a very use­ ful quality. Despite my fondness for the Kan, I've endeavoured to produce a balanced and objective study of the variation. This is not intended as an ail-in-one repertoire book and it's not filled with ninety-nine per cent of black wins! My idea was to include both the popular and the less fashionable lines, recommending variations for both colours where it's merited. In general the results of the games I've selected reflect the success rate of the lines at grandmaster level. This is not to say, however, that they produce simi­ lar results at lower levels. I've assumed that the reader doesn't have access to other material on the Kan in such publications as ChessBase Maga'.{jne, Infonnator and ChessPublishing.com and I've collected and checked relevant analysis and assess­ ments from these sources. I've also endeav­ oured to attribute such analysis and assess­ ments correctly, except perhaps when they have been blindingly obvious. Naturally I have been assisted in my task of checking and providing new analysis by various com­ puter engines - it would be reckless to con­ sider writing certain chess books these days without one. Computers do have some weaknesses but they are excellent at both checking analysis and prompting the author to consider unusual (or sometimes blatantly obvious) ideas. Virtually all the world's top 8

players study openings with the aid of com­ puters and many of the major opening nov­ elties over the last few years have been computer inspired. I decided on using illustrative games in­ stead of the traditional method of variation trees. In this particular instance, with both White and Black having such a free rein regarding move orders and with both sides employing 'systems' rather than just 'moves', it's easy to see that illustrative games have a major advantage. I will, how­ ever, be concentrating mainly on the open­ ing moves . and the early middlegame plans and tactics, and will only linger on the later parts of the games if they have characteristic 'Kan' qualities or if they are of particular interest. To provide as much useful infor­ mation as possible, I have forced myself to be quite ruthless about this. Once you be­ come involved in studying a particular game, it's all too easy to be seduced by all the tactics of the late middlegame and end­ game, even if they have nothing to do with the actual opening. Before you know it, a game covering one small sideline turns into a ten-page epic! I have deliberately fought against this; this is first and foremost a book on the K.an, not a games selection. Except for a brief explanation in the in­ troduction, I will be dealing with the various positional and tactical ideas for both players as they occur in the games, rather than sepa­ rately. I've found that some games have come out as a labyrinth of difficult tactical variations, while others much more posi­ tional and wordy; that's the way it is with the Sicilian. Many thanks go to Byron Jacobs and Dan Addelman for their patience and ex­ tended deadlines, and to Christine for her support and proof-reading. John Emms, Kent, September 2002

INTRODUCTION

I

Let's begin by taking a look at the initial moves of the Kan. 1 e4 c5 2 tt:lf3 e6 3 d4 cxd4 4 lt:Jxd4 a6

Black's fourth move underlines the un­ doubted flexibiliry of the Kan. Black waits for White to commit himself before decid­ ing where to develop his pieces. The advan­ tage is that Black has more 'information' to go on before deciding what set-up to use. The disadvantage is that White also has more options at his disposal. More specifi­ cally, White can choose if he wishes to erect the 'Maroczy Bind' with an early c2-c4 (this is avoided by Black in many other Sicilian defences by the insertion of an early ...l2:\f6, inducing White to play l2:\c3) . Onto the actual attributes of the move

4 ... a6. In some ways it could be construed as a 'high-class' waiting move, but it does have positive features relevant to the position. Firstly, it protects the bS-square and thus rules out any early lbbS by White. This is particularly important because 2 ... e6 weak­ ens the d6-square, while Black also often plays his queen to c7 at an early stage. An­ other positive feature of ... a7 -a6 is that it can support the rypical Sicilian lunge ... b7b5, which Black may play as early as move five! This strike on the queenside is nor­ mally much more e ffective if White plays an early l2:\c3, as then a timely ...b5-b4 forces the knight to move again. This can be im­ portant as often Black bases his counterplay on attacking White's very slightly vulnerable e4-pawn and the knight on c3 is a natural defender of this pawn. Kan/Paulsen/Taimanov

There has often been some confusion when chess players talk about these openings. The subject of this book is the Kan Variation (1 e4 cS 2 l2:\f3 e6 3 d4 cxd4 4 lLlxd4 a6) . However, in the past this move order has also been referred to as the Paulsen Varia­ tion. Indeed, at my editor's insistence, my 1997 booklet on this line was entitled Trends in the Paulsen Volume 2. Looking at other 9

Sic ilia n Ka n

literature, The Oxford Companion to Chess, a good guide, calls it the Paulsen but does mention that it can be called the K.an too. More recently people have come to accept the Kan as the main name, but of course it's how well you play it that really matters! In most circles, the very similar variation 1 e4 cS 2 t2Jf3 e6 3 d4 cxd4 4 tbxd4 tbc6 is known as the Sicilian Taimanov (although just to confuse things this line is also some­ times referred to as the Paulsen!). Naturally the Kan can transpose into the Taim.anov and vice versa. The concept that really sepa­ rates the two openings is that in the K.an Black either delays or forgoes the move ...tbc6 unless it is really beneficial. In this book, on the whole, we will be concentrat­ ing on pure 'Kan' positions and so we will see more of ...t2Jbd7 than of ...t2Jc6.

advantage, but i n the 1970s many top grandmasters, including the likes of Ulf Andersson and Ljubomir Ljubojevic, dis­ covered many new defensive resources for Black and these are seen throughout the book. It was shown that Black's 'Hedgehog' pawn structure (e6, d6, b6 and a6) is in fact very difficult to break down. Furthermore, Black's structure is very flexible and White sometimes has to use a lot of piece power ensuring that Black cannot break with either ...dS or ...bS. Finally, White has chances to attack on the kingside, in the centre and on the queenside but must be very careful not to overextend himself in doing so, other­ wise he runs the risk of being impaled on one of those Hedgehog's spikes! It's true to say that those playing White, as well as those with Black, need a lot of patience to play these positions.

Maroczy Bind versus the Hedgehog Scheveningen Structure

The diagram above shows White with employing the so-called 'Maroczy Bind', a structure which dominates the first few chapters of this book. The prongs on c4 and e4 promise a pleasant space advantage for the first player. Furthermore, Black's important Sicilian pawn break in the centre with ... d6-d5 is now very difficult to achieve, while ...b7-b5, another typical Sicilian lunge, is also prevented for a very long time. It was once thought that this type of pawn struc­ ture promised White a clear or even decisive 10

Later in the book we deal with a similar pawn structure but with a subtle difference. In the diagram above White's c-pawn is back on c2, normally blocked by a knight on c3. This is known as the Scheveningen structure (the Scheveningen Variation arises after 1 e4 cS 2 t2Jf3 d6 3 d4 cxd4 4 tbxd4 t2Jf6 5 tbc3 e6). Purely from a structural point of view, the Scheveningen structure is less advantageous for White than the Ma­ roczy Bind (some would say it's actually

Intro duction

advantageous to Black). The point is that there is now much less to prevent Black both advancing in the centre with ... d6-d5 and expanding on the queenside with ... b7b5. White can use some method of con­ tainment (with a2-a4, for example) but gen­ erally this is not his main aim. The advantage White has here over the Maroczy Bind is one of time. By missing out c2-c4 White gains an extra tempo and in open Sicilian positions a tempo is often the difference between success and failure. Generally White will try to use this extra time to generate a quick attack against the black kingside. Talking simply of pawn moves, this will normally involve the ad­ vance f2-f4, after which Black has to con­ stantly be on guard over f4-f5 and e4-e5 advances. Sometimes White also throws in g2-g4-g5 for good measure. Of course these plans are also possible in Maroczy Bind positions, but in Scheveningen positions they are really crucial to the success of the white player. Many games follow the same formula: either White wins in barnstorming fashion in under thirty moves, or Black sur­ vives the onslaught and later on feasts on the remains of White's overextended posi­ tion.

Chapter Order

Chapters 1 -5 deal with 5 .id3. This solid move continues to be by far the most popu­ lar choice against the Kan. The chief posi­ tive feature of 5 .id3 is its flexibility. White keeps the option open of erecting the Ma­ roczy Bind with c2-c4, as well as simply forgoing this in favour of lLlc3. Perhaps the only disadvantage of 5 .id3 is that Black has so many different responses at his dis­ posal, so the white players needs to be well booked up. Possible black responses include 5 ...lLlf6, 5 ... .ic5, 5 ...'ir'c7, 5 ...'ir'b6, 5 ...g6, 5 ...lbc6, 5 ...lbe7, 5 ... d5. The list goes on and on! Chapters 6-8 deal with the more tactical 5 lLlc3, immediately reaching the Scheveningen structure. Black has fewer choices here, the main 'Kan' responses be­ ing 5 ...'ir'c7 and 5 ... b5. Finally, in Chapter 9 we deal with 5 c4, whereby White immediately sets up the Maroczy Bind. You would think that this 'no-nonsense' move would be popular, but in fact the reverse is true and it's actually much less common than both 5 .id3 and 5 lbc3. The main reason for this is that Black has many more options than simply to ac­ cept a Hedgehog structure.

11

CHAPTER ONE

I

5 i.d3 '2Jf6 6 0-0 'Wic7 7 'Wie2 d6

1 e4 c5 2 lt:lf3 e6 3 d4 cxd4 4 lt:lxd4 a6 5 .id3 lt:lt6 6 0-0 �c7 7 �e2 d6

In this chapter we will be dealing with what is generally known as 'the main line' of the Kan. Let's begin by considering the opening few moves.

tively (perhaps a t c5 o r even d6), instead of boxing it in with the move 6... d6 (sec Chap­ ter 3).

1 e4 c5 2 lt:lf3 e6 3 d4 cxd4 4 lt:lxd4 a6 5 .id3 lt:lf6

5.. /2Jf6 is by far the most popular move. Black develops his king's knight and pre­ pares to castle. Alternatives to 5...4Jf6 will be discussed in Chapters 4 and 5. 6 0-0

Moving the king into safety. It's ex­ tremely uncommon for White to castle queenside in the 5 .td3 variations of the Kan so there is no particular advantage in delaying castling. For alternatives, however, see Chapter 3. I will mention here that 6 eS? loses a pawn to 6...'it'a5+!. 6 . . . 'iic 7

This little queen move is the subject of the first two chapters. Black takes control over the long b8-h2 diagonal and protects the crucial e5-square - White was already threatening e4-e5. The queen also keeps an eye on events down the c-file and directly attacks White's c-pawn, which normally ends up on c4. For the moment Black keeps the option of developing his f8-bishop ac12

7 �e2!

This clever move is again the most popu­ lar choice for White. He renews the 'threat' of e4-e5 and keeps his options open regard­ ing pawn structure. There are, however, quite a few alternatives (including the im­ mediate 7 c4) and these are discussed in Chapter 2. 7 . . . d6

On first sight this looks a little inconsis­ tent with Black's previous move as the f8bishop is now blocked anyway. However, most Kan players are loath to allow White to play an early e4-e5, dislodging the impor-

5 ii d3 t:Dt6 6 0 - 0 'IWc 7 7 '1We2 d 6

tant defensive knight on f6. Alternatives, including 7...i.c5 and 7....td6 are studied in Chapter 2. 8 c4

With this move White sets up the Ma­ roczy Bind. 8 tLlc3 leads to positions discussed in Chapter 6, while also important is the ag­ gressive 8 f4 (see Game 13). 8

. . .

g6

The idea of fianchettoing the king's bishop in this line was discovered in the 1980s as a way of strengthening Black's kingside in anticipation of a white offensive in that area, and 8...g6 is now Black's most popular response to 8 c4. Classical development with 8...i.e7 is the traditional way of playing the Hedgehog and is discussed in Games 9-11, while 8...tt:Jbd7, leaving the option open as to where to de­ velop the dark-squared bishop, is the sub­ ject of Game 12.

bat Black's fianchetto. The bishop o n g7 provides an excellent shield for the black king but, as opposed to when the bishop is developed on e7, it doesn't protect the slighdy vulnerable d6-pawn. White immedi­ ately makes this his main target by posting a rook on the half-open d-fi.le. 10 tt:Jf3 can simply transpose into the main line after 10...0-0 but White docs have some tricky move order options: 11 i.f4 tbc6 (11...tLlbd7 12 l::tacl!? tLlg4 13 l::tfdl transposes to the note to Black's 12th move in Game 5 without allowing the possibility of what Kobalija played against Adams) 12 h3 (12 l::tfd1 transposes to Game 4) 12...tLld7 13 l::tacl tt:JccS 14 lLlxeS tLlxeS 15 l::tfd1 and we have reached a position from Game 3. White certainly has other plans, including a quick kingside attack involving the lunge f2-f4 (discussed in Game 7). Another, more positional, idea is to bolster the e4-pawn with f2-f3 and aim for a gradual attack on the qucenside (see Game 6). 1 0 . . . 0-0 1 1 t:Df3

9 t:Dc3

White continues to develop naturally. In most cases the knight's best square is c3, from where it keeps an eye on the critical bS- and ciS-squares. However, White can also contest the long diagonal by fianchetto­ ing his own dark-squared bishop with 9 b3 (see Game 8). 9 . iig7 1 0 l:!.d1 .

.

The plan beginning with this move was quickly installed as the critical way to com-

Continuing the basic plan. White clears wood from the d-fi.le and in some cases prepares to add pressure to d6 with i.f4 (note that the immediate 11 .tf4? would simply lose a piece after 11...e5!). The stage is set for a complex positional battle. Move sequences, ideas and tactics from this posi­ tion are covered in Games 1-5. 13

Sicilian Kan

Game l Z .Aimasi-Farago Linz 1995

1 3 .i.e3

13 �f4 tt::ld7 transposes to the note to Black's 13th move in Game 3. 1 3 . . . tiJd7

5 .i.d3 ttJt6 6 0-0 �c7 7 �e2 d6 8 c4

Black now has many options with his knights, including ...ltJcS, ...tt::lceS and ...tt::ldeS.

g6 9 lUc3 .i.g7 1 0 �d 1 0-0 1 1 ttJt3 ttJc6

1 4.li!.ac1 ttJc5

1t...tt::lbd7 is also possible (see Game 5) but I believe that 1t...tt::lc6 gives Black more options, especially against 12 �f4 (see Game 4).

14...tt::lce5?! is a typical move in this type of position but in this instance, with White's bishop on e3, it leaves Black struggling to cover all his weaknesses: 15 tt::lxeS and now: a) 15...dxe5?! has been played here and in similar positions but I don't like this move at all.

1 e4 c5 2 lUt3 e6 3 d4 cxd4 4 tUxd4 a6

1 2 h3

This refinement was introduced after black players found a good way of combat­ ing the previously popular 12 .tf4 (see Game 4). With 12 h3 White doesn't rush the bishop to f4. He reasons that with a black knight probably on its way to eS, the bishop will be blocked on f4 and may be more happy on e3, at least for the time being. Instead White cuts out both ...tt::l g4 and, if Black plays...e6-e5, the idea of ...�g4. 1 2 . . . b6

So that Black can develop his light­ squared bishop on b7. Black normally plays 12...tt::ld7 13 �e3 and only then 13...b6 (or 13...tt::lde5 - see Game 2), reaching the same position as in the text. The move order cho­ sen here can only make a difference if White decides to play 13 �f4 (see Game 3), in which case Black is already committed to playing ...b7-b6. 14

Black has voluntarily blocked his own dark-squared bishop and has presented White with a dangerous queenside pawn majority which can be galvanised immedi­ ately with 16 b4!, intending c4-c5. In gen­ eral, the only time Black can justify playing ... dxeS is if there is a knight on c6 ready to jump into the newly-acquired d4 outpost (see Game 2). b) 15... tt::lxe5 16 b3 �b7 17 �b1 l:tfd8 (17 ...�ab8 18 f4 tt::lc6 19 �d2 �fd8 20 �f2 reaches the same conclusion, while 17....l:ife8 18 �d2 �f8 19 tt::la4 tt::ld7 20 tt::lxb6! tt::lxb6 21 �aS tt::ldS 22 l:txdS! left White a clear pawn up in Luther-Von Gleich, Bonn 1993) 18 �d2! (Von Gleich) 18....Uab8 (18...�£8 19 tt::la4 tt::ld7 [Z.Alrnasi­ Vogt, Altensteig 1993] 20 tt::lxb6! wins a

5 i.. d3 li:Jf6 6 0 - 0 �c 7 7 �e2 d6

pawn in a similar fashion) 19 f4 lL'lc6 20 �f2 i.a8 21 ltJa4 and the b-pawn drops off the board. 14...ltJde5?! allows the tactic 15 ltJd5! (Black always has to be wary of this idea and a similar one with ltJb5) 15...exd5 16 cxd5 and the pin on the c-ftle allows White to regain his piece with interest, for example 16...lL'lxf3+ 17 �xf3 lL'le5 18 l:Ixc7 lL'lxf3+ 19 gxf3 i.xh3 20 �h2 i.c8 21 b4 b5 22 ..if4 and Black is struggling in this ending. 14...i.b7 15 ..ib1 lDc5 transposes to the main line. 1 5 i.. b 1

A typical move for this line. The bishop vacates the d-ftle (so that the d6-pawn feels the full effect of the rook on d1), while keeping an eye on the slightly vulnerable e4pawn. The move ..ib1 is favoured over i.c2 because on b1 the bishop is out of harm's way and blocks neither rook. 1 5 . . . i.. b 7

1 6 �d2!

A good all-purpose move. White both threatens d6 and sets up the possibility of exchanging dark-squared bishops with ..ih6. However, White does have alternatives: a) 16 a3!? prepares to chase the knight away with b2-b4 but allows the black knight to settle after 16...lL'lb3 17 l:lc2 and now it's a case of whether the knight on b3 proves to be a nuisance for White or a liability for Black (or perhaps both):

a1) 17...ltJe5 18 ..ia2 ltJc5!? (18...lL'lxf3+ 19 gxf3! lL'lc5 20 �d2 ..ie5 21 b4 lL'ld7 22 f4 ..ixc3 23 l:lxc3 lL'lf6 24 f3 ..ixe4 25 fxe4 lL'lxe4 26 �d3 lL'lxc3 27 �xc3 e5 was agreed drawn in Tsuboi-Urday Caceres, Sao Paulo 1998, but I like White's bishop pair in the final position) 19 lL'lxe5 ..ixe5 20 ..id4 ..ixd4 (20....:.fd8 21 b4 lL'ld7 22 ..ixe5 lDxe5 23 f4 ltJd7 24 l:Icd2 looks a bit better for White) 21 l:lxd4 e5! 22 l:td1 lL'le6 23 �e3 lL'ld4 24 .l:tcd2 .Uad8 25 lL'le2 lL'lxe2+ 26 �xe2 ..ic8! and Black will equalise with .....ie6. a2) 17...lL'lca5 18 ..ia2 l:tac8 19 ltJd4 ..ixd4 20 i.xd4 e5!? (20...ltJxd4 21 l:lxd4 is a touch better for White - d6 is more vul­ nerable than c4) 21 ..ie3 b5 and now in­ stead of 22 �f3 f5!, which was very unclear in Luther-Mattick, Seefeld 1996, White should play 22 cxb5! .i.xe4 23 b6 �c6 24 b7 �xb7 25 lL'lxe4 'it'xe4 26l:Ixc8l:Ixc8 27 'ii'xa6, when Black's in serious trouble. b) 16 b3 is a solid move, lending extra support to the c4-pawn before commencing further operations. In the game Kersten­ Chuchelov, Willingen 1999, Black suc­ ceeded in achieving a reasonable position after 16...�e7 17 �d2 nfd8 18 lL'ld4 lDe5! (as is so often the case in hedgehog posi­ tions, the weakness of d6 becomes more prominent as pieces are exchanged; hence Black's reluctance) 19 ..ig5 f6! 20 ..ie3 g5. Black still has weak pawns but his two knights have acquired excellent outposts and neither can be removed without White creating weaknesses in his own camp. 16

. . .

�ad8 1 7 b3

Again a useful move - White gives the c4-pawn some protection. However, this is hardly life threatening for Black and so it's unsurprising that White has also tried more direct methods. a) 17 i.h6 f5! is a suggestion of Istvan Almasi and I believe that this is the way that Black should treat this position. With both ...b6-b5 and ...d6-d5 looking extremely un15

Sic ilia n Kan

unlikely in the short term, the ... f7-f5 lunge is Black's only realistic pawn break. Of course there is always some risk advancing pawns in front of your king, but in this case this is counterbalanced by the activity it gives to the black position. Instead 17...lbe5 18 ..i.xg7 lbxf3+ 19 gxf3 �xg7 20 b4 lbd7 21 ifxd6 'ir'xc4 22 ..i.d3 'ir'c8 23 lba4 was better for White in Lane-Chuchelov, Ant­ werp 1999. b) 17 lDh2 (planning lDh2-g4) 17...'i!fe7 18 lbe2?! (18 ..i.h6 is stronger, when I still like 18...f5) 18...a5 19 b3 lbb4! 20 lbc3 dS! and Black, having achieved the desirable ...d6-d5 break, took over the initiative in Van den Doel-Shaked, Wijk aan Zee 1998. c) The more I look at 17 lDdS!, the more I like it. It's the sort of move which should always be considered, even with Black hav­ ing an extra knight blocking the c-ftle:

17...exdS 18 cxdS (White regains his piece as if the c6-knight moves White plays b2-b4) 18...lbe5 (18...l::tfe8 19 dxc6 ..i.xc6 20 i.d4 looks uncomfortable for Black) 19 lDxeS ..txeS 20 b4 1fe7 21 bxcS bxcS 22 ..igS ..tf6 23 ..txf6 ifxf6 24 f4 and White's pawn centre promises him an edge. If Black has no improvement here then there is cer­ tainly a case for 17 lbdS over 17 b3. 1 7 .'i'b8? . .

Black defuses lDdS ideas and prepares .....i.a8 and ...b6-b5. However, Alrnasi's sub­ sequent play casts doubt upon the wisdom 16

of removing the queen from the action and it's here that Black should be looking for improvements. a) Another more recent game involving Ivan Farago saw the Hungarian GM trying the slightly clumsy-looking manoeuvre 17.....ta8 18 lbh2 (18 lbdS is again possible although on this occasion Black is not forced to caprure) 18...1fb7!?, once again angling for ...b6-b5. However, at least in this instance the queen still covers the second rank and the advantage of this can be seen in the sequence 19 i.h6 fS, when Black is not forced to recaprure on g7 with the king. Instead the game Graetz-Farago, Latschach 2001 continued 19 1fe2 fS! 20 f3 (20 exfS? allows Black to demonstrate another point of ....i.a8 and ...1Vb7: 20.....txc3! 21 lhc3 lbd4 22 'ir'g4 lbe2+ and Black wins) 20...f4 and Black had gained some useful space on the kingside, while securing eS as an out­ post. b) The immediate 17...f5!? looks to me like the most consistent continuation (Zeller also mentions this move).

Black immediately strives for counterplay with this logical move, for example: b1) 18 lbd4?? loses material to 18...f4!. b2) 18 lDdS isn't so effective now. After 18...exd5 19 cxdS Black has the choice of two ways to play it: 19...lbe7 20 b4 fxe4 21 i.xe4 lbxe4 22 l::txc7 lbxd2 23 lbxd2 ltJxdS 24 l::txb7 l::tfl with an equal ending, or

5 i. d3 0, (6 6 0 - 0 � c 7 7 � e 2 d 6

19...lDe5 20 lDxeS dxeS 21 b4 f4 22 i-xcS bxcS 23 bxcS nf6 which is difficult to as­ sess. In this second line I think I would prefer to play the black side - at least he will probably have the option of returning the piece for White's two dangerous passed pawns. b3) 18 exfS lDeS!? (18...gxf5 19 i-gS �d7 [19...i-f6!?) 20 %:te1 lDeS also looks okay for Black) 19 lDd4 (or 19 lDxeS dxeS 20 'ir'e2 gxfS and Black has a strong-looking pawn centre) 19...gxf5 and Black can even con­ sider following up with the aggressive ...�f6-g6. Critical now is 20 b4 'ir'f7! 21 i.gS lid7 22 'ir'e2 lDe4 but Black looks to be okay in these complications. 1 8 .i.h6!

A very natural move, underlining Black's weaknesses on the dark squares. This idea has increased in strength now that the black queen finds herself further away from the action.

2 3 . . . 0,d7

The only way to hang on to the d6-pawn was to play the extremely ugly 23...l2Jb7 but unsurprisingly Farago could not bring him­ self to play like this! 24 �xd6 0,e5 25 �xb8 l:!.xb8 26 0,e2 l:!.fd8 27 0,d4 �f7 28 f3 l:!.d7 29 0,b3 l:txd 1 + 30 l:!.xd1 0,xc4 31 l:!.d7 + �e8 32 l:!.xh7 l:!.d8 33 l:!.h8 + �d7 34 l:!.xd8 +

1 8 . . . i.a8

�xd8 35 i.d3 0,a3 36 e5?

Istvan Almasi suggests 18...-ic8 as an improvement, but I don't really see how this helps that much after 19 lDh2.

I don't understand this move at all. Why allow Black's terrible bishop on a8 back into the game? 36 �£2 (I. Almasi) keeps total control.

1 9 0,h2!

An excellent idea. The knight plans to hop into g4, where it highlights Black's dark-squared problems.

36 . . . fxe5 37 .i.xg6 e4!

1 9 . . . f6

38 fxe4 0,c2 39 b5 axb5 40 e5 0,e3 41

19...b5 amounts to nothing more than a demonstration on the queenside. After 20 i-xg7 �xg7 21 cxbS axbS 22 b4 Black's weaknesses (bS, d6, f6 and h6) will become very difficult to defend. I would still be tempted to play 19...f5!?, if nothing else to stop lDg4, although I do admit this is not consistent with Black's previous play.

g5 0,c4! 42 .i.d3 0,xe5 43 .i.xb5 �e7

20 .i.xg7 �xg7 2 1 0,g4 0,e5 22 b4

g6 9 0,c3 ..ig7 1 0 l:!.d1 0-0 1 1 0,t3 0,c6

0,xg4 23 hxg4

1 2 h3 0,d7 1 3 i.e3

White has triumphed in the middlegame batde and has acquired a virtually winning game. Black's position is full of weaknesses and he has no real counterplay.

Black seizes his chance and reaches a drawn ending.

Y,-Y,

Game 2 Z .Aimasi-Anand

FIDE World Ch., Groningen 1997 1 e4 c5 2 0,t3 e6 3 d4 cxd4 4 0,xd4 a6 5 i.d3 0,t6 6 0-0 �c7 7 �e2 d6 8 c4

13 i-f4 transposes to Game 3. 1 3 . . . 0,de5 14 l:!.ac1 i.d7!

This deployment goes well with playing ...ttJdeS. When compared to 14...b6?!, this 17

Sicilian Kan

set-up leaves Black much better placed to cope withlLldS ideas. 1 5 lt:\xe5

With this move White clarifies the posi­ tion in the centre, but there are some play­ able alternatives: a) 15 b3!? lLlxB+ (15... f5 16lLlxe5lLlxe5 17 f4lLlfl looks like a better bet) 16lixB liaS gives Black reasonable counterplay according to Almasi, but I suspect that after 17 .id2 White still has an edge, for example 17...f5?! 1S exfSlLleS (1S...gxf5 19lig3) 19 lig3lLlxd3 20lixd3lixfS 21 �xfS gxfS 22 .if4!. b) 15lLle1 tries to avoid an exchange of knights, but this looks too artificial to me: 15...f5! 16 f4 lLlfl! (a good square for the knight; here it protects the d6-pawn) 17 exfS gxfS 1Sli£2 �acS 19 lLlBlidS 20 b3 lLle7 21 .ib1lLlg6 and Black was fine in the game Andriulaitis-Brooks, correspondence 1999. c) 15 ltJdS? doesn't hit the mark here as Black has c6 well covered: 15...ltJxf3+ 16 lixf3 exdS 17 cxdS .ixb2 1S �c2 .ieS 19 dxc6 bxc6 and Black is a safe pawn up. 1 5 . . . dxe5!?

terbalanced by the fact that Black can utilise the outpost on d4. The alternative 15...lLlxe5!? is also not bad: 16 f4!? (16 b3 fS transposes into note 'a' to White's 15th move) 16...lLlxd3 17 lixd3 .ic6 1S ltJdS!? (1S lixd6 lixd6 19 �xd6 .ixc3 20 �xc3 .ixe4 is equal - Ribli) 1S...exd5 19 cxdS (with an edge, according to Ribli) 19...llfeS! (targeting e4; 19....ixb2 20 �c2 .ig7 21 dxc6 �fdS 22 lib3! looks strong for White) 20 .id4 .ixd4+ 21lixd4 �acS and Black looks okay. 1 6 f3!

So that the queen can nudge herself to £2 and the c3-knight can contest the d4-square from e2. The previously played 16 ltJdS caused Black no problems after 16...exd5 17 cxdS �fdS 1S �c2 �acS, Arnason­ Vyzmanavin, Manila Olympiad 1992. 1 6 . . . lt:\d4 1 7 'i't2 i.c6 1 8 lt:\e2 .l::.f dB 1 9 lt:\xd4!

Giving Black a protected passed pawn, but it will be well blocked and White will have chances to use his pawns on both sides of the board. 19 .ib 1 apparently adds more pressure on d4, but Black can reply calmly with 19...�d7!, answering 20 lLlxd4 exd4 21 .ixd4?? with �adS (Almasi), pinning and winning the bishop. 1 9 . . . exd4 20 i.d2

Recapturing with the pawn is much more playable when, as on this occasion, the knight is on c6 rather than d7 (compare with note 'a' to Black's 14th move in Game 1). It's true that White has an active pawn majority on the queenside but this is coun18

Almasi assesses this position as slightly better for White and I agree with him.

5 i. d3 0,t6 6 0 - 0 'Wif e 7 7 fie2 d6

White is threatening to expand with either f3-f4 or b3-b4 and it's impossible for Black to prevent both. Still, with careful defence Black shouldn't be in too much trouble. 20 . . . e 5

O r 2 0...a 5 2 1 f4! and now ...e6-e5 i s an­ swered by f4-f5. White can slowly arrange kingside play by re-racking his rooks on f1 and el. 21 b4!

Launching the queenside pawn majoriry before Black has time for ...a6-a5. 2 1 . . .i.d7?!

Missing a win. Almasi gives the following convincing analysis: 33 aS! .1d6 34 eS bxaS 35 exd6!! .l:.xb 1+ 36 'ith2 and now: a) 36...a4 37 cS a3 38 c6 a2 39 .l:.xa2!! (but not 39 c7? .l:.ht+! 40 'itxhl alii'+) 39 ....txa2 40 .tfS!! and White wins. b) 36....l:.b3 37 cS with a further split: bl) 37....tf5 38 c6! .l:.xd3 39 c7 ii'c8 (or 39 .'iVf8 40 ii'xd3 .1xd3 41 c8ii' .1xc2 42 d7) 40 d7+. b2) 37....l:.xd3 38 ii'xd3 .tfS 39 ii'd2 .1xc2 40 if'xc2 and the two connected passed pawns are decisive. . .

Almasi prefers 21...ii'd7, for example 22 ii'e2 (threatening to win a piece with b4-b5) 22...b5! 23 cS ii'b7 (dissuading White from playing a2-a4) and it is difficult to see how White makes any real progress.

Perhaps in time trouble, Almasi begins to drift quite badly and soon he is worse de­ spite the material advantage.

22 'ii'e 2 aS?!

.l:!.b7 .l:!.cB! 3B .l:!.f2 .l:!.c7 39 .l:!.b5 a4 40

And this is too panicky. Black succeeds in splitting White's queenside but a pawn is too high a price to pay for this.

'ife4 l:.bB 44 .!:!.aS 'Wife7 45 'ii'f3 'Wifc7 46

23 bxa5 i.e6 24 �h 1 .!:!.deB 25 l:.c2 i.fB 26 .l:!.b1 ..tcs 21 'iff2 .!:!.ebB? 2B f4!

This now works due to tactical reasons. If Black does nothing then White will push with f4-f5. Therefore Anand opted to sacri­ fice an exchange.

33 . . . l:!.aB 34 .l:!.a2?!

34 . . . 'Wifd7 ! 35 aS? bxa5 36 'iff3 'ifdB! 37 .l:!.a5 a3 4 1 l:!.f1 l:lcB 42 .l:!.a6 'ii'e B 43 l:tb5 'ifxe5 47 'iff6 'ii'c 7 4B l:!.a 1 .l:!.cB 49 'iff2 i.fB 50 l:lc 1 'Wifc6 0-1

Game3 Dix-Edwards

Comspondence 1993

2B . . . exf4 29 i.xf4 'ifxa5

1 e4 c5 2 {jjf 3 e6 3 d4 cxd4 4 0,xd4 a6

Or 29....td6 30 .txd6 ii'xd6 31 cS ii'eS 32 .l:.bS! and White targets the b7 -pawn.

5 i.d3 {jjf6 6 0-0 'Wifc7 7 'ii'e 2 d6 B c4 g6 9 0,c3 i.g7 1 0 l:ld 1 0-0 1 1 {jjf 3 0,c6

30 ..txbB .l:!.xbB 3 1 'ii'g 3! 'ii'd B 32 a4 b6

1 2 h3 0,d7 1 3 ..tf4

33 e5?

The main alternative to 13 .1e3. White 19

Sicilian Ka n

adds immediate pressure to the d6-pawn and induces Black to move a knight to e5.

the weaknesses in his position.

1 3 . . . lbce5

Again Black must make the decision over which knight, if any, should block on e5. a) 13...b6 14 .!:tact .ib7 15 lLlb5!? (15 ltJd5 'ii'b8!; 15 b3!?) 15...axb5 16 cxb5 .l:txa2 17 bxc6 .ixc6 (Ribli), and now 18 .ixd6! 'ii'xd6 19 .ic4 is better for White. b) 13...lL:lde5!? 14 .!:tact .id7, playing as in the previous game, is very possible, for example 15 b3 .l:tfc8!? (15... lL:lxf3+ 16 'ii'xf3 ltJd4 17 'ii'e3 e5 18 .ih6 .ie6 19 .ixg7 xg7 20 lLle2 lLlxe2+ 21 .ixe2 was a tiny bit better for White in Palac-Galego, Lisbon 2001 but Black should still hold comforta­ bly) 16 .ib1 .ie8 and now: b1) 17 lL\et 'ii'b8 (Black is aiming for ...b7-b5) 18 .ig3 lL:la7! 19 f4 lLlec6 20 lLlf3 (20 1i'd2!? .l:td8 21 .ih4 .l:td7) 20...b5 21 c5?! (21 fS!? looks more threatening) 21...b4 22 cxd6?! .ixc3 23 d7 .ixd7 24 .l:txd7 lL:ld4! 25 'ii'd 1 lLlxf3+ 26 gxf3 lLlb5 and Black was doing well in Moberg-Grabliauskas, Man­ hem 1999 - White's kingside strucrure is a little shaky. b2) 17 .ie3!? looks stronger as now 17...1i'b8 18 lLlxe5 dxe5?! 19 lL:la4! exploits the weakness of b6. c) 13....ixc3!? is a radical idea, giving up the dark-squared bishop in order to cripple White's queenside pawns. However, this concept has acquired some credibility re­ cently (it has been tried in similar positions) and here it gives Black a playable game ac­ cording to Ernst: 14 bxc3 e5 15 .ih6 (the dark squares around the black king are look­ ing vulnerable but it's quite difficult for White to exploit them) 15...l:td8 16 lLlh2 lLlc5 17 .ic2 f6 18 f4 exf4 19 .ixf4 lLle5 20 lLlfl .ie6 21 lL:le3 .l:tac8 22 .l:td4 b6 see

following diagra m

with an unclear position in Kotronias­ Johannessen, Istanbul Olympiad 2000. Black's well-placed knights compensate for 20

1 4 lbxe5

It makes sense for White to clear the ten­ sion in the centre before deciding what plan of action to take. That said, no harm can be done with the flexible move order 14 .!:tact b6 and now: a) As I mentioned before, 15 lLlxeS dxe5?! works less well when Black has a knight on d7 rather than c6 as White tends to get a free hand on the queenside (15...lL:lxe5 transposes to the main game): 16 .ie3 .ib7 17 b4! f5 18 f3 l:.f7 19 c5 bxc5 20 lL:la4 fxe4 21 fxe4 .if8 22lLlxc5

and White was clearly better in Xie Jun­ Movsesian, Hastings 1996/97. b) 15 b3 .ib7 16 .ib1 with a further split: b1) 16... f5 17 lLlxe5! lLlxe5 transposes to Palac-Farago below (17...dxe5 18 .ig5 looks

5 i. d3 liJ f6 6 0 - 0 W c 7 7 'if e 2 d6

slightly better for White - 1 8... h6 can be answered by 1 9 'il'd2!). b2) 16 ... ltfe8! (the importance of this move is explained in the note to Black's 17th move) 17 lt:Jh2 (17 lt:Jxe5! lt:Jxe5 trans­ poses into the main game) 1 7 ...l:tad8 1 8 'i!i'd2 ( 1 8 i..g 5!?) 1 8. . .lt:Jc5 19 ..ig3?! f5! 20 f4 lt:Jf7 and Black had reached a very comfort­ able position in Wahls-Lau, Munich 1 992. Note what a good job the knight does on f7, where it defends the d6-pawn. 1 4 . . . liJxe5

14 ... dxe5?! 15 ..ie3 b6 16 ltac1 trans­ poses to Xie Jun-Movsesian above. 1 5 l:!.ac1 b6 1 6 b3 i.b7 1 7 i.b1

Another possibiliry for White is 17 'il'd2!?, planning to retreat the light-squared bishop along the d3-f1 diagonal. This should be seriously considered as, although ..id3-b 1 is a very popular retreat in this line, the bishop can end up being rather passive on this square.

lt:Jf7 21 f4 f5 22 exf5 gxf5 23 'il'£2 ..ic6 24 i..d4 and the exchange of dark-squared bishops favours White. Golubev-M.Ivanov, Delzlsau 1 997 continued 24 ... ..ixd4 25 'il'xd4 lt:Jh8? 26 lt:Jd5! (again this move!) 26 ...'i!i'g7 27 'il'xg7+ �xg7 28 liJc7 and White won a crucial pawn. b) 1 7 ... f5 tends to work less well when a pair of minor pieces have been exchanged the weaknesses on d6 and b6 are more dif­ ficult to cover: 1 8 'il'd2 ltfd8 1 9 ..ig5! ltd? 20 ..ie3 (targeting b6 now that the black knight has lost access to d7) 20 ...ltb8 21 ..id4 ltdd8 22 f4 lt:Jf7 23 ..ixg7 �xg7 24 exf5 exf5 (24...gxf5 looks more consistent, but 25 lte1 lte8 26 'il'd4+ still looks prom­ ising for White: 26 ... e5 27 'il'£2 e4 28 lt:Jd5 i..xd5 29 cxd5 'i!i'd7 30 l1c6 and White can eventually open Black up by arranging g2g4) 25 lt:Jd5 i..x d5 26 cxd5 'il'd7 27 ltc6 and White held a clear advantage in Palac­ Farago, Oberwart 200 1 . 1 8 'ikd2 �fB 1 9 �e3

1 7 . . JUe8!

An important move. Black creates the possibiliry of a d6-defending ... ..if8, while he also takes steps against a possible 'il'd2/..ih6 plan for White. Other ideas in­ clude: a) 17 ...ltfd8 defends the d6-pawn but now Black must always be wary of an an­ noying ..ig5: 1 8 'ir'd2 l:tab8 19 ..ig5! f6 (19 ...ltd7? 20 lt:Jd5! exd5 21 cxd5 lt:Jc6 22 ltc2 and Black is in big trouble) 20 ..ie3

Preparing f2-f4 and hitting the b6-pawn. If White is looking for an alternative here, then I quite like the look of 1 9 ..ih6!?. In general I believe that the exchange of these bishops helps White as Black has problems defending his dark-squared weak­ nesses. With this in mind, Black's most logi­ cal move looks to be 1 9 ... ..ie7. White can continue in the same vein with 20 i..g5 and now: a) 20 ... l1ad8? 21 i..xe7 'ir'xe7 22 'ir'd4! 'il'c7 23 lt:Ja4 and White wins either the b­ or the d-pawn. b) 20 ... f6 21 i.e3 and White will follow up with f2-f4. c) 20 ...i.. f8 21 f4 (21 i.. h6 repeats the po­ sition; black players would have to take this into account) 21 ...lt:Jc6 (21 ...lt:Jd7 22 e5! looks good for White) 22 lt:Jd5!? (22 i.. f6 also look interesting) 22...exd5 23 cxd5 'ifd7 (23 ... ltac8 24 dxc6 i.xc6 25 i.. f6 'ifb7 26 i..b2 i..g7 27 i..xg7 �xg7 28 lie 1 gives Black problems with his d6-pawn) 24 dxc6 21

Sicilian Kan

�xc6 2S eS dS 26 l:tel and I like White's kingside pawn majority. Going back to Black's 1 9th move, 19 ...l:tad8 20 �xf8 l:txf8 21 f4 tt:'lc6 22 tt:'ldS!? is again promising for White. Per­ haps Black's best is 19 ... �xh6!? 20 'tixh6 and only then 20...l:tad8. 1 9 . . . .!:!.ad8

Edwards later suggested prophylactic defence of the improvement. Given the eventually reaches, it's hard the text move.

19 ... l:tab8, with b6-pawn, as an position Black to be critical of

20 f4! ?

Black wins - Edwards. 28 l:te t looks better, although I still prefer Black after 28 ... l:tce8 29 l:txeS dxeS 30 l:td 1 tt:'lcS 31 tt:'lxcS 'i!VxcS+ 32 lif2 e4 33 'tixcS bxcS. 28 . . ..1:!.ce8 29 lLlc3 lLlf6 30 .!:!.f1 lLlh5 3 1 fxg6 hxg6 3 2 .!:!.f2 'ilt'e7!

The balance of power has shifted very much in Black's favour; the dark squares around White's king are looking vulnerable and Black's last move shows real ambition to exploit this. 33 �g1 'i'h4 34 lLld5 b5 35 .l:!.cf1 ..txd5 36 cxd5 lLlg3 37 .l:!.d 1 lLle2 + 38 �f1 .l:!.e3!

Pushing the f-pawn down the board is double-edged as it creates weaknesses in White's camp. Even though it loses a tempo over the previous note, there is still some­ thing to be said about playing 20 �h6 here. 20 . . . lLld7 21 f5 .!:!.c8

21 ...�a8 (Edwards), so that ...l:tb8 will defend b6, is another sensible move which may be slightly stronger than the text. 22 'ilif2 i.g7 23 .i.f4

23 tt:'la4!? is more ambitious. After 23 ... tt:'lcS (23 ... exfS!? 24 exfS tt:'lcS) 24 f6 �f8 2S tt:'lc3 tt:'ld7 26 l:tfl it's not clear whether the pawn on f6 will turn out to be a strength or a weakness. 23 . . . ..te5 24 .i.xe5 lLlxe5 25 lLla4 lLld7 26 'ilid2?!

White again targets the d6-pawn but Black has some tactical resources. Probably White should prevent Black from opening the e-ftle with 26 fxg6 hxg6 and now 27 'tid4 tt:'lcS (27...�c6 28 tt:'lc3!) 28 lixd6 tt:'lxa4 29 bxa4 'ticS+ 30 'tixcS l:txcS 31 aS! is better for White (3 1 ...l:txaS 32 l:td6!). 26 . . . exf5! 27 exf5 .l:!.e5!

39 .l:!.xe2

This desperate measure is forced. 39 �d3 fails to 39 ...l:txh3! 40 gxh3 'i!Vxh3+ 41 'ite1 (or 41 l:tg2 'tif3+ 42 'itel tt:'lf4+) 41...'i'ht + 42 l:tft 'i'h4+ 43 l:t£2 tt:'ld4+ 44 �e2 tt:'lf3+ 4S 'itft 'i'h1 mate. 39 . . . .l:!.xe2

40

'i'xe2

.l:!.xe2

41

�xe2

'i'h5 + 42 �e 1 'ilt'e5 + 43 �f2 'i'b2 + 44 �f3 �g7 0-1

Game4 Shaposhnikov-Karttunen

Suddenly Black has some freedom and his pieces soon become very active.

Athens 200 1

28 �h 1 ?

1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4 a6

This just seems to accelerate Black's pro­ gress on the kingside. The d6-pawn cannot be captured: 28 'tixd6?? l:tel + 29 'it£2 'tixd6 30 l:txd6 l:txcl 31 l:txd7 �c6 and

5 .i.d3 lLlf6 6 0-0 'i'c7 7 'i'e2 d6 8 c4

22

g6 9 lLlc3 .i.g7 1 0 lLlf3 0-0 1 1 .i.f4 lLlc6 1 2 .l:!.fd1

Or 11 l:tdt tt:'lc6 12 �f4. This line (for-

5 il.. d 3 ttJf6 6 0 - 0 �c 7 7 � e 2 d 6

going h2-h3) was initially White's most popular response, but Black's next move has been responsible for its decline. Note that Black can also answer 12 .l:tacl with 12 ... e5, for example 13 i.e3 i.g4 14 h3 4Jd4 1 5 i.xd4 .ixf3 1 6 'ifxf3 exd4 17 4Jd5 'ifd8 with equality.

12

. . .

e5!?

An amazing idea, which I believe was the invention of Ivan Farago. At first it looks as though Black, who is blocking in his g7bishop and gifting White the d5-square, is committing positional suicide. However, this is not the whole story. Black soon achieves control over the d4-square, which will be reinforced after ... i.g4. It seems that this compensates Black fully and he has excellent chances of equalising. 1 2...4Jd7 reaches positions similar to ones discussed in Game 3, except that White hasn't expended a tempo on the 'semi-useful' h2-h3. You would think that this should be enough to sway the balance of the position in White's favour, but it's not so clear, for example 1 3 lhc1 4Jde5 and now: a) 14 b3 'ife7 1 5 i.b1 see followin g dia gram

with a further split: a1) 1 5 ...l:tb8 1 6 Wh1 !? (or 1 6 i.e3 b6 1 7 4Jd4 4Ja7 1 8 f4 4Jd7 1 9 tLl f3 .ib7 with an edge in Mainka-Chuchelov, Senden 1 998)

1 6. . .b 6 1 7 i.g5! f6 1 8 i.e3 Wh8 1 9 4:'Jd4 4Jxd4 20 .ixd4 4Jc6 21 .ie3 f5 22 'ifd2 was better for White in Ernst-Hjartarson, Ostersund 1 992.

a2) 15 ... .id7 16 4Je1 l:tad8 17 i.e3 f5 1 8 f4 and now, instead o f 1 8 . . .4:'Jg4 1 9 i.b6 .l:tde8 20 exf5 gxf5 21 4Jf3 which was better for White in Hendriks-Chuchelov, Dieren 1 997, Black should play the thematic 1 8 ...4Jf7! lending support to the d6-pawn. b) 1 4 'ife3!? and now: b1) 14 ... 'ife7 1 5 .ie2! .id7 16 4Je1 (pre­ paring i.h6 and f2-f4; at the moment the knight on e5 has no retreat square) 1 6 ... .l:tad8 1 7 i.h6 i.xh6 1 8 'ifxh6 g5 (18 ... f5 1 9 'ifd2 i.c8 20 f4 tLlf7 21 exf5 gxfS 22 .if3 with some advantage to White [Milos]; note that White's light-squared bishop is more active on f3 than it would be on b 1) 1 9 4Jc2 f6 20 4Je3 4Jg6 21 g3 tLlce 5 22 .l:tfl and White, who prepares £2-f4, stood better in Milos-Motwani, Manila Olympiad 1 992. b2) 14 ... 4Jxf3+ (it makes sense to effec­ tively gain a tempo - White has played 'ii'e2-e3xf3) 1 5 'ii'x f3 tLle5 (1 5 ...4:'Jd4 1 6 'ifg3 e 5 1 7 .ie3 .ie6 1 8 4Je2 is better for White - Zeller) 1 6 'ifc3! (a suggestion from Zeller; the queen is better placed on e3 than e2 as it facilitates a possible .ih6 and allows the light-squared bishop to retreat along the d3-f1 diagonal) 1 6 ... .id7 (grabbing on c4 loses: 1 6 ... 4:'Jxc4? 1 7 .ixc4 'ii'xc4 1 8 .ixd6 23

Sicilian Kan

�d8 1 9 l2Jd5) 1 7 i.e2! (again it should be said that the bishop is more active here than on b 1) 17 ...�fc8 18 b3 i.e8 19 'il'd2 and White maintains an edge, Gandalf 4.32g­ Nimzo 8, Cadaques 200 1 . Smooth posi­ tional play from the machine with the white pieces!

f5 with good counterplay, Stefansson-Lutz, Manila Olympiad 1 992. 1 5 h3 i.xf3 1 6 'i'xf3 llld4 1 7 i.xd4 exd4

1 3 i.e3

13 ..ig5?! ..ig4! leaves White with some problems over the threat of ... l2Jd4. Luther­ Farago, Budapest 1991 continued 14 �acl l2Jd4 1 5 'il'e3 .i.x£3 1 6 gx£3 h6! 1 7 i.xh6 ..ixh6 18 'il'xh6 lLix£3+ 19 �g2 l2Jd4 20 lLid5 'il'd8 21 �g1 lLixd5 22 exd5 'il'f6 23 'it>fl 'il'g7 and with both a stronger minor piece and pawn structure, Black stands very well. 1 3 . . . i.g4!

1 8 llle 2

It's unusual for this bishop to be devel­ oped on g4, but on this occasion I believe it's mightily effective. Black begins to lay claim to the d4-square.

18 .I:.xd4 'iVb6! hits b2 and d4, regaining the pawn. However, this might be White's best continuation, as in the game I believe Black might even be slightly better when the players agreed a draw.

1 4 ..tc2

14 h3 leads to early simplification into a totally level position: 14 ...l2Jd4 1 5 i.xd4 exd4 1 6 l2Jd5 i.x£3 1 7 'il'x£3 1!2-1/2 Z.Almasi-Farago, Hungarian Championship 1 992.

1 8 . . . llld 7 1 9 lllx d4 llle 5 20 'i'e2 lllxc4 21 ..tb3 :tea 22 :ac 1 b5 23 'i'd3 Y:z-Y:z

Game 5 Adams-Kobalija

FIDE World Ch., Las Vegas 1999 1 e4 c5 2 lllf3 e6 3 d4 cxd4 4 lllx d4 a6 5 ..td3 lllf6 6 0-0 'i'c7 7 'i'e2 d6 8 c4 g6 9 lllc3 ..tg7 1 0 lllf3 0-0 1 1 i.f4

1 4 . . . :ac8! ?

Also perfectly acceptable for Black is 14 ... l2Jd7 15 l2Jd5 'il'd8 1 6 h3 ..ix£3 17 'il'x£3 lLic5 1 8 'il'e2 lLie6 (taking complete control over d4) 19 'il'd2 tLicd4 20 i.d3 �b8 21 f3 24

If White chooses to play 1 1 �d1 first then Black has to be slightly careful if he wishes to play ...t2Jbd7 and ...l2Jg4 (although it has to be said that this line looks good for White in any case - see below) . a) The immediate 1 1 ...lLibd7 can be an­ swered by 1 2 h3!?, preventing Black from playing ...l2Jg4 and thus giving Black less options. 12 ... b6 13 ..if4 (1 3 ..ie3!?) and now: a 1) 13. .. i.b7?! 14 l:tacl l2Je5 1 5 lLixe5 dxe5 16 i.e3 and Black has no compensa­ tion for White's active queenside pawn rna-

5 � d3 ti:J f6 6 0 - 0 ike 7 7 ii e 2 d6

jority, Hamdouchi-L.Marin, Sitges 1 999. a2) 1 3...ltJh5 14 ..1d2!? (preventing the pawn-crippling ... ..ixc3; 14 ..1e3 ..1xc3 1 5 bxc3 is similar to the main game) 1 4.....1b7 1 5 l:.acl l:.ac8 1 6 b3 l:r.fe8 17 ..1e3 (17 ..ib 1 is safer) 17 ...ltJc5 1 8 ..ibt and now, instead of 1 8 ... ..1ffi 19 ltJh2! ltJf6 20 ltJg4 ltJxg4 21 hxg4 giving White a pleasant position in Kalod-Chuchelov, Cappelle Ia Grande 2002, I believe Black should grab the pawn with 1 8 ... ..1xc3 1 9 l:.xc3 ltJxe4. White certainly has compensation for the pawn, but whether this is enough to give him the ad­ vantage is another matter. b) t t ...ltJg4 12 ..1f4 (12 h3 ltJe5 13 ltJxe5 dxe5! and Black can follow up with ... ltJc6d4) 1 2...ltJd7 transposes to the note to Black's 1 2th move. 1 1 . . . ti:Jbd7

This move used to be Black's most popular choice but it has now been super­ seded by the more flexible 1 1...ltJc6. Again Black can continue with 1 1 ... ltJg4 if he is worried about White preventing this with h2-h3; 1 2 l:.fd t ltJd7 transposes into the note to Black's twelfth move. Also pos­ sible is 1 t ...ltJh5!?, hitting f4 and c3 before White has the chance to defend the c3knight with l:.ac l . 1 2 ltfd1

In his notes to this game in Informator, Adams suggests 1 2 l:.acl !?, giving the c3knight some added support and thus taking

some of the sting out of ...ltJh5. a) 12 ... ltJg4 13 l:r.fd 1 leads to the note to Black's 1 2th move. b) 12 ...ltJh5 may be the best move in any case; at least the pressure on d6 is relieved for the time being: 1 3 ..ie3 b6 14 l:.fdl ..ib7 15 'tid2 ltJhf6!? (15 ...l:.fe8 may be more accurate; Black can always defend with ... i.ffi) 1 6 ..1e2!? ltJxe4 (16 ... ltJe8? 1 7 b4 l:tc8 1 8 ..if4 ltJe5 1 9 ltJxe5 dxe5 20 ..1e3 again left Black with no compensation for White's queenside advantage in Gromotka­ Podzielny, correspondence 1 996) 17 ltJxe4 ..1xe4 18 'tixd6 l:.a7! and Black is just about okay. 1 2 . ti:Jh5!? . .

The older move is 1 2 ...ltJg4 but White has good chances to secure an advantage against this. The game continues with 1 3 l:tac 1 and now: a) 13. .. ltJde5 1 4 h3 ltJxf3+ 1 5 'tixf3 ltJe5 16 'tie3 transposes to note 'b2' to Black's twelfth move in Game 4, except that White has the extra move h2-h3. This was already quite pleasant for White and the extra move is a bonus. b) 13 ...ltJge5 14 ltJxe5 and now: bl) 14 ... dxe5?! has been suggested in at least one source, but I don't see the point. After 1 5 ..1e3! Black has no active plan and White will push on the queenside with the b2-b4 advance; this is obviously better for White. b2) 1 4...ltJxe5 1 5 'tid2 l:.e8 16 ..1e2 ..iffi (De Vreugt-Bosboom, Dieren 1 999) and now I like 1 7 'tid4! ltJc6 18 'tie3 and Black is under some pressure from threats of ltJa4 and c4-c5. c) 1 3. .. b6 14 'tid2! (planning to retreat the bishop to e2 or ft ; this plan of Lastin's looks stronger than the previous 1 4 b3 and 1 5 ..ibl) 14 ...ltJde5 (14.....1b7 1 5 .1ft gives Black serious problems with his d-pawn: 1 5 ... ..1xc3 16 .ttxc3 e5 1 7 'tixd6 'tixd6 1 8 l:.xd6 ltJc5 1 9 ltJxe5! ltJxe5 20 ..1xe5 ltJxe4 21 l:.xb6 ltJxc3 22 l:.xb7 is very strong) 1 5 25

Sicilian Ka n

�e2 l:td8 1 6 �g5 f6 (or 1 6 ...l:td7 1 7 lL'ld4 h6 1 8 .th4 g5 1 9 .tg3 .tb7 20 h4 gxh4 21 .txh4 Wh7 22 l2Jd5! exd5 23 cxd5 'tlib8 24 f4 'ii'g8 25 lL'lf5 1 -0 Lastin-Shaposhnikov, Moscow 1 999) 17 lL'lxe5 lL'lxe5 18 �e3 18 ... l:.b8 19 b3 lL'lf7 20 f4 f5 21 exf5 gxf5 22 .th5! and Black has trouble developing his c8-bishop, Lastin-Kobalija, Elista 200 1 . For example, 22...�b7 is met by 23 'ii f2 .tc6 24 lL'ld5! (Ribli). 1 3 ..te3

1 4 bxc3 e 5

Naturally Black attempts to regain some control over the dark squares. Adams also mentions the sensible-looking 1 4...b6, pre­ paring to develop the bishop on b7. 1 5 ..th6 .l:!.e8 1 6 �e3 lt:\c5 1 7 h3

Preventing ... .tg4. 1 7 . . . ..te6

17 ... .td7, preparing . . ..tc6 (Adams), may be stronger. 1 8 ..te2 lt:\f6 1 9 lt:\d2 'it>hB

13 .tg5 invites Black to do the same: 15 ... .txc3 14 bxc3 e5 15 'ii'd2 lL'lc5 1 6 .tc2 f6 17 .te3 l:td8 18 'iie 2 .te6 with an equal position, Kulaots-J.Bellin, Gausdal 2000. The safe 13 �d2 leads to positions simi­ lar to the those studied in note 'b' to White's twelfth move, but White's bishop is less active on d2 and Black has a reasonable chance to equalise: 13 ... b6 14 !tact .tb7 1 5 �e3 ltac8 1 6 .tb1 l:tfd8 1 7 h3 'iib 8 1 8 b3 b5!, Nijboer-Bosboom, Wijk aan Zee 1 991 is another example of a successful black strategy. 1 3 . . . ..txc3! ?

Black has completed his plan that began with 1 3 ....txc3 and has achieved a very solid position, despite the absence of the dark­ squared bishop. Nevertheless, White still holds a slight edge. The problem is that, although Black is extremely solid at the moment, at some point the position is likely to open up and then Black may feel the pinch on those dark squares. This is indeed what happens in the game. 20 ..tg5 lt:\gB 21

�g3 �c6 22 ..te3

.!:!.adS 23 .l:!.ab1 .l:!.d7 24 .U.b4 f6 25 f3 �cB 26 �f2 lt:\e7 27 lt:lb3!? lt:\xb3 28 axb3 b5 29 cxb5 a5 30 .l:!.c4! ..txc4 3 1

Again we meet this outwardly-shocking idea of giving up control of the dark squares. The compensation, as always, is the infliction of doubled isolated c-pawns on the white camp and the chance to find use­ ful outposts for the knights. 13 ... b6 14 !tact .tb7 transposes to note 'b' to White's twelfth move. 26

..txc4 d5 3 2 exd5 a 4 3 3 .i.c5 axb3 34 ..txb3 lt:lf5 35 b6 �a6 36 g4 �b5 37 .l:!.b1 lt:\d6 38 ..tc2 �a6 39 .U.b4 lt:lb7? 40 ..ta4 l:!.xd5 4 1 ..txeB .l:!.xc5 42 .l:!.a4! �d3 43

.l:!.aB

�b 1 +

44

'it>g2

�xb6

45

..txg6 + 'it>g7 46 ..txh7 .l:!.b5 47 �xb6 .l:!.xb6 48 .i.e4 .l:!.b2 + 49 'it>g3 lt:\d6 50 .l:!.a7 + 'it>fB 51 .i.d5 .U.c2 52 .l:!.d7 1 -0

5 i. d3 &iJ f6 6 0 - 0 ik c 7 7 ik e 2 d6

Game 6 Johnson-West

Comspondence 1997 1 e4 c5 2 &iJf3 e6 3 d4 cxd4 4 &iJxd4 a6 5 i.d3 &iJf6 6 0-0 ikc7 7 ike2 d6 8 c4 g6 9 &iJc3 i.g7 10 i.e3 0-0 1 1 .l:!.ac1 &iJbd7 1 2 l:tfd1 b6

Black has the opportunity here to change the character of the position by trading his d7-knight for White's light-squared bishop: 1 2...tt:'!eS!? 13 f3 i.d7 14 b3 :ac8 1 S 'ii'd2 (1S i.b1 !?) 1 S ... tt:'!xd3 1 6 'lixd3 1:tfd8 1 7 tt:'!de2 tt:'!e8 1 8 'it>h 1 'ii'aS 1 9 i.d2 'lieS 20 tt:'!f4 'Wa3 21 i.e3 'WaS 22 i.d2 'lia3 23 .U.c2 bS! and Black was fine in the game Vogt­ Bischoff, Baden-Baden 1 992. Considering the ease with which Black gains equality here, it's surprising that this idea hasn't been seen more often. 1 3 f3 i.b7

Here we see a totally different approach from White, who very much adopts an 'anti-Hedgehog' formation. The first thing to notice is White's particularly solid king­ side structure. The e4-f3-g2 pawn formation very much blunts the power of the b7bishop and it's unlikely that White will be provoked into changing this structure with­ out good reason. With his kingside well protected, White will aim to attack Black on the queenside. A typical and dangerous plan

involves b2-b4, tt:'!d4-b3 and a2-a4-aS. With the exchange of a-pawn for b-pawn, Black's queenside structure is more vulnerable and White also has a chance to create a powerful passed pawn with c4-cS. If Black finds no answer to this plan then his position can soon become critical. He has to sit tight and look for opportunities to break with either ... b6-bS or, more likely, ... d6-dS. If Black can arrange a successful advance then usu­ ally his problems are over, but White will not allow this to happen without a fight. Sometimes Black even has to consider play­ ing ... d6-dS as a pawn sacrifice. 1 4 'i'f2 l:tac8 1 5 i.f1

1 4 'lif2 and 1S i.ft is typical prophylaxis from White. The light-squared bishop moves out of range from attack via ...tt:'!cS or ...tt:'!eS and White removes some wood from the d-fJ.!e, uncovering the potential of the d 1 -rook. 1 5 . . JUe8! ?

A good square for the rook as after . . .d6dS the e-fJ.!e may well be opened. If Black wishes to avoid the next note, then he can play 1 S ...'Iib8 16 b4 and only then 1 6 ...:fe8. Given the precariousness of Black's posi­ tion in this game, it's certainly worth explor­ ing different ways of creating counterplay against White's super-solid structure. One idea is 1 S ... tt:'!hS!?, for example 1 6 b4 i.eS (with the idea of ... d6-dS) 1 7 g3 l:tfe8 18 f4 (18 tt:'!b3, keeping the pawn on f3, looks stronger) 1 8...i.g7 and Black can be very happy with his work. White has been lured forward on the kingside and is now sof­ tened up along the h1 -a8 diagonal. Black's b7 -bishop has come to life and suddenly Black has ready-made counterplay against c4 and e4 (Kalod-Movsesian, Zlin 1 997). 1 6 b4

Black's move order allows White the tac­ tical possibility of 1 6 tt:'!dbS!? axbS 1 7 tt:'!xbS 'Wb8 1 8 tt:'!xd6 when White will obtain a rook and two pawns for two knights. If 27

Sicilian Kan

White can create passed pawns on the queenside then Black may struggle, but this is all very double-edged as Black certainly benefits from having an extra piece in the middlegame. The game Browne-Petrosian, Milan 1 975 is rather unhelpful: after 1 8... .ic6 1 9 b3 .if8 20 tt'lxe8 .l:txe8 21 �d2 the players decided to call it a day. I t seems that most white players are reluctant to re­ linquish their positional edge to reach such an unclear position. 1 6 . . .'i'b8

Removing the queen from the c-ft!e and thus lining up the option of ... d6-d5. 1 7 lLlbJ

tt'lxc8 .l:txc8 was a speculative exchange sacrifice which worked in the game Kaeser­ Podzielny, Dortmund 1 992, but shouldn't give Black enough compensation. c) 17 ....ia8 18 cS! again looks strong. d) 1 7 ... .if8 has been played a few times but surprisingly I can't find any examples of the obvious 1 8 .ixb6. West believes accept­ ing the sacrifice is critical and calls Black's compensation 'nebulous'. I agree with him - I don't think Black's position is strong enough to be so bold with this sacrifice: after 18 ... tt'lxb6 19 ifxb6 both 19 ... d5 20 exdS tt'ld7 21 'i!fd4 exdS 22 tt'lxdS and 1 9 ... .ih6 20 .l:tc2 dS 21 exdS .if4 22 g3 .ic7 23 iff2 exdS 24 cS seem to fall short for Black. 1 8 a4

White continues with the logical plan of a4-a5. I also think that the direct 18 tt'la4 .ia8 1 9 cS!? is worth considering: 1 9 ... b5 (or 1 9 ...bxc5 20 bxcS dxcS 21 tt'laxcS tt'lxcS 22 tt'lxcS aS 23 ife1 and Black's a-pawn will be picked off) 20 tt'lb6 tt'lxb6 21 cxb6 .l:txct 22 .l:txcl and White's passed b-pawn looks dangerous. 1 8 . . . d5!?

White's queenside play is beginning to look threatening. Already the b6-pawn is attacked. 1 7 . . J:tc6 !?

'[fhis move] looks ugly (and downright bad!) but actually it gives nothing away and has some points that are not immediately obvious' - Guy West. I imagine the Austra­ lian IM finally got round to this move after discovering faults with Black's alternatives. a) A fter 17 ....ic6 West likes the direct 1 8 cS, highlighting a problem with Black's pre­ vious move: it leaves the a-pawn unde­ fended. 1 8 ...bxc5 (or 1 8... dxc5 19 .ixa6 and Black cannot avoid material loss) 19 bxcS tt'lxcS 20 tt'lxcS dxcS 21 .ixa6 .l:tcd8 22 .ixcS and White is just a clear pawn up. b) 17 ... tt'le5 1 8 tt'la4! dS 19 tt'lxb6 dxe4 20 28

Imaginative play. With Black's rook blocking the bishop on b7, this thematic move looks unplayable here, but Black does have some resources. The continuation 1 8 ...tt'le5 19 bS .l:tcc8 20 .ixb6 tt'lxc4 21 .ia7 ifa8 22 .id4 .l:ted8 23 .ixc4 .l:txc4 24 tt'laS .l:tcc8 25 b6 .l:td7 26 �b2 tt'le8 27 .ixg7 tt'lxg7 28 eS left Black in big trouble in Branding-Haufe, corre­ spondence 1 998. 1 9 b5!?

1 9 cxdS ('fantastically complicated' West) must be critical. West believes that Black can hold the balance but keeps his analysis to himself (and why not?): 1 9 ... exd5 20 exdS (20 tt'lxdS tt'lxdS 21 exdS .l:txcl 22 tt'lxcl 'i!fd6 and Black has typical compensa­ tion in the form of targets at dS and b4 plus a more solid structure on the kingside; 20

5 � d 3 tl:J f6 6 0 - 0 'ike 7 7 'ii e 2 d6

b5 axb5 21 .txb5 reaches the game posi­ tion) 20 .. Jhc3!? (20 .. J:td6 21 i.f4 lt'le5 22 lt'ld2! looks good for White) 21 �xc3 lt'lg4 22 fxg4 .ixc3 looks like the most important line. Does Black really have enough for the pawn? I don't believe so.

to many players this is an enticing strategy but Black, with his fianchettoed bishop providing extra cover, is ready for White to throw in the proverbial kitchen sink.

1 9 . . . axb5 20 cxd5

After 20 axb5 Black can choose between 20... �d6!? and 20 ...l::tcc8 21 cxd5 exd5 22 lt'lxd5 lt'lxd5 23 cxd5 1Wd6. In the latter case I certainly agree with West that Black has enough counterplay. He follows up this assessment with the line 24 i.c4? 1Wa3!, when it is difficult to deal with the threat of ...�xc4. 20 . . . exd5 21 �xb5 dxe4! 22 tl:Jxe4!

White can grab the exchange for a pawn with 22 .ixc6 i.xc6 but Black always has good practical chances in this type of posi­ tion. 22 . . . tl:Jxe4 23 fxe4 tl:Je5!

Another offer of the exchange. 24 h3

Again White declines. 24 i.xc6 .ixc6 again promises Black good compensation both a4 and e4 are attacked while ...lt'lg4 is also in the air. 24 . . Jbc1 25 llxc 1 :d8

Threatening to exchange knight for bishop with ...lt'ld3. The b6-pawn remains very weak but this is counterbalanced by the vulnerability of the e4-pawn - the position is more or less equal. 26 :t1 .be4 27 .txb6 :c8 28 :c 1 ! l:txc 1 + 29 tl:Jxc 1 'ikd6 % - %

Game l Borngaesser-Lau

Essen 1996 1 e4 c5 2 tl:Jf3 e6 3 d4 cxd4 4 tl:Jxd4 a6 5 i.d3 tl:Jf6 6 0-0 'ikc7 7 fie2 d6 8 c4 g6 9 tl:Jc3 �g7 1 0 f4

Raw aggression! What happens if White goes for glory on the kingside against Black's solid structure? It must be said that

1 0 . . . 0-0 1 1 tl:Jf3

White retreats the knight for kingside operations and prepares a possible e4-e5. But what happens if White lunges forward in 'caveman' fashion? Kan players should not be unduly worried by such shows of aggressiOn: a) 1 1 e5?! is premature: 1 1 ...dxe5 12 fxe5 lt'lfd7 shows a big advantage in developing the bishop on g7 rather than e7. Black's kingside is rock-solid (there are certainly no worries about threats on h7 as there would be without ...g7-g6) . Furthermore, White's e5-pawn is a major weakness which could fall at any moment: 13 .if4 lt'lxe5 14 �ae1 lt'lbd7 1 5 �h 1 1Wd6 and White has run out of good moves. b) 1 1 f5!?, however, deserves more re­ spect. White activates the rook on f1 and adds pressure to the e6-pawn. He docs, though, give away the important e5-square. White players should never give away this square lightly as a black knight will find it to be a wonderful outpost: 1 1 .. ..l::te 8 1 2 fxe6 fxe6 13 'iti>h 1 lt'lfd7! (heading for e5) 14 .ie3 lt'lc5 15 �acl b6 16 b3 .ib7 and the players rather unhelpfully agreed a draw in Luther­ Kochycv, Leningrad 1 989. Black is com­ fortable in the final position. White has no 29

Sicilian Kan

entry squares on the f-Ele, the knight on e5 is a powerful beast and Black can follow up with ...lt:Jb8-d7-c5. 1 1 . . . b6 1 2 �d2 �b7

to do anything constructive without allow­ ing Black to break in the centre or on the queen side. b) 14 .. .l:he8!? (Black counter-attacks where White believes he is strongest - on the kingside!) 1 5 b3 lt:Jh5! 16 l:tfe 1 (White can push the knight back with 1 6 g4 lt:Jhf6 but it's very possible that White will later feel the draught around his king, especially if the a8-h1 diagonal opens up) 1 6... ..i.h6 1 7 g3 f5! 1 8 ..i.b1 lt:Jdf6 and Black's bishop on b7 is a marvellous piece, Matulovic-Cvitan, Yugoslav Championship 1 988. 1 4 . . . l:tac8 1 5 1Vf2

A typicai manoeuvre. White's queen is heading for h4, where it hopes to participate in a kingside attack. 1 3 l:tac1

1 5 . . . 1Vb8 1 6 1Vh4

A prophylactic move. White wishes to play b2-b4 (to prevent ...lt:Jd7-c5) and thus takes his rook off the long a 1 -h8 diagonal to avoid tricks. More direct attempts are not dangerous for Black. a) 13 e5? is again too rushed and White may simply end up a pawn down: 1 3 ...dxe5 14 fxe5 lt:Jfd7 15 if4 ..i.xf3! 16 :xf3 lt:Jc6 and Black threatens both the e5-pawn and ...lt:Jd4. b) 13 l:he1 lt:Jbd7 14 e5?! isn't much bet­ ter: 14 ... dxe5 1 5 fxe5 lt:Jg4 16 if4 ..i.xf3 1 7 l:txf3 lt:Jgxe5 1 8 l:te3 11ad8 and White has litde to show for his pawn.

1 6 . . . b5!

1 3 . . . lt:lbd7 1 4 b4

14 �h 1 is a typical prophylactic move for White in the Kan: in one go White solves any problems with tactics involving a check along the g 1 -a7 diagonal. Whether this is worth expending a tempo over is a long-running debate in many positions. a) 14 ... l:tac8 (the traditional way) 1 5 b4 'iib 8 (vacating the c-flie and unleashing the power of the c8-rook; Black is eyeing up opportunities to break with either ... b6-b5 or ... d6-d5) 16 ..i.e3 'ifa8! 1 7 lt:Jd2 l:tfe8 with a finely balanced position, Zude-Hulak, Bundesliga 1 990. It's very difficult for White 30

Striving for the initiative on the queen­ side. Tactics against the e4-pawn and White's king on the long diagonal support this lunge. 1 7 cxb5 axb5 1 8 ..txb5?

White cannot resist the pawn, but after this move his centre collapses. 1 8 i.e3 looks stronger. 1 8 . . . ..txe4!

1 9 �xd7 1Va7 +

20 Wh 1

1Vxd7

Black possesses the centre pawns and the control they give is far more important than any endgame potential of White's passed pawns on the queenside. It's safe to say that

5 JJ.. d3 0, (6 6 0 - 0 'i'c 7 7 'i' e 2 d 6

Black is already clearly better. White's next move blunders a pawn, after which he lost the will to continue.

21 b5? i.d3 0-1

Game 8 Short-Sax

Amsterdam 1983 1 e4 c5 2 0,t3 e6 3 d4 cxd4 4 0,xd4 a6 5 JJ.. d 3 0,t6 6 0-0 'i'c7 7 'i'e2 d6 8 c4

Note that the actual move order to this game was 8 b3 ll'lbd7 9 c4 g6 10 i.b2 ..tg7 1 1 lt:'lc3 0-0 1 2 �ad1 b6 1 3 ..tb1 i.b7 but I have tinkered with it to include other possi­ bilities. B . . g6 9 b3 .

When discussing move orders, it should be mentioned that 9 ll'lc3 ..tg7 10 b3 is slightly inaccurate as it gives Black the op­ tion to break with 10 ... b5!?, taking advan­ tage of the undefended c3-knight. 9 . . . JJ..g7 1 0 JJ.. b 2 0-0 1 1 0,c3

This is another popular set-up for White. In my opinion, the fianchetto of the c l ­ bishop gives a harmonious 'feel' t o the White position - his pieces look to be on nice squares. White will try to attack in the centre (pressure down the d-flie) and the kingside. Nevertheless, Black's solid king­ side structure, aided by the fianchetto, is well set up to oppose White's ideas. This queenside fianchetto is more successful

when employed against classical black de­ velopment (see Games 10 and 1 1 ).

Another option for White is to develop his knight on d2: 1 1 lL'ld2 lL'lbd7 1 2 �act b6 13 �h1 ..tb7 14 ..tb1 �feB 1 5 f4 and now 1 5 ... e5! shows a disadvantage in White's choice of square - there is no knight ready to jump into the ourpost on d5. Gyimesi­ Farago, Pula 2000 concluded 1 6 fxe5 dxe5 17 ll'l4f3 ll'lh5 18 'ii'f2 1/2-1h. One can say that Black is certainly not worse in the final position. 1 1 . . .b6 1 2 :ad1

White continues to improve his pieces the rook can add pressure to the slightly vulnerable d6-pawn. The 'caveman' approach is too crude to succeed: 12 f4 ll'lbd7 1 3 �act i.b7 and now: a) 1 4 e5? dxc5 1 5 fxe5 ll'lh5 and the e­ pawn just drops off the board. b) 14 f5?! �ae8! 1 5 fxe6 fxe6 and White has achieved nothing except to give away the e5-ourpost. 1 2 . . . JJ.. b 7 1 3 JJ.. b 1 0,bd7

The typical Kan move, but 13 ...ll'lc6 is also not bad: 14 ll'lxc6 ..txc6 1 5 �d2 l:lfd8 1 6 �fd1 (Hi.ibner-Lutz, Baden-Baden 1 992) and now Hi.ibner suggests 1 6 ...lt:'ld7!?, an­ swering 17 l:.xd6 with 17 ... ..te5. 14 �h 1

Note that the immediate advance 14 f4 should be answered by 1 4...e5!, as in the 31

Sicilia n Kan

game.

Game 9

1 4 . . JHe8 1 5 f4 e5!

Trapl-Tompa

Decin 1977 1 e4 c5 2 lt:'lf3 e6 3 d4 cxd4 4 lt:'lxd4 a6 5 ..id3 lt:'lf6 6 0-0 'i'c7 7 'i'e2 d6 8 c4 i.e7

Finally we come to the classical devel­ opment of the f8-bishop. Advantages over the fianchetto include the saving of a tempo, no weakening of the dark squares and better protection of the d6-pawn. The main disadvantage is that Black's kingside is less well protected. An important defensive thrust in the Kan which highlights another positive feature of the kingside fianchetto: ... e6-e5 becomes much more playable as White's knight no longer has access to the fS-square, such an important point for White's attacking ambi­ tions. As for the qualities of ... eS, it's true that Black gives away the dS-square, but this is compensated for by the acquisition of the eS-square and pressure against White's iso­ lated e-pawn. Black is certainly not without counterplay. Strangely enough, 1 5 ... i.f8 would trans­ pose to the note to Black's 1 4th move in Game to (which was fine for Black), al­ though of course it looks ridiculous to play it here! 1 6 fxe5 lt:'lxe5 1 7 J:t.de1 b5!?

Now the game explodes into unclear complications. If Black wishes to play more quietly, then 1 7 ...l::te7 with a possible dou­ bling on the e-ftle makes sense. 1 8 cxb5 d5! 1 9 J:t.xf6! ? i.xf6 20 bxa6 i.xa6 21 lt:'lxd5 'i'd6 22 lt:'lb5 ..ixb5 23 lt:'lxf6 +

'i'xf6 24 'i'xb5 'i'f2 25 ..ic3

lt:'lg4 26 'i'g5 l:lac8 27 i.b4 'i'd4 28 'i'xg4 'i'xb4 29 J:t.f1 'i'd2 30 'i'f3 J:t.e7 31 h4 J:t.c3 32 'i'f6 l:td7 33 'it>h2 l:tc 1 34 ..id3 .l:!.xf1 35 i.xf1 'i'e 1 36 'i'f4 J:t.e7 37 ..id3 h5 38 'i'g5?? l:t.d7 ! 39 i.b5 J:t.d 1 40 'i'f4 'i'g1 + 41 'ith3 l:te1 0-1 32

9 lt:'lc3 0-0

1 0 'ith 1

Once again we see this prophylactic idea, clearing the king from possible tactics along the long gl -a7 diagonal. John Nunn has also described such moves as 'lazy' - White can't be bothered to work out the tactics on every move so he gets rid of the possibility. On this occasion there is a case for deleting it, or at least postponing it for the time being. The extra tempo could, after all, prove to be very useful in building up a strong attack. More aggressive is 10 f4!? and now: a) 10 ...lt::lc 6?! 1 1 lt::lx c6! bxc6 (or t t . . .'ilt'xc6 12 eS dxeS 1 3 fxeS lt::ld7 14 i.f4 and White can freely build up a kingside attack; note that Black has much less pressure on eS without the bishop on g7) 1 2 eS lt::ld7 (or

5 i. d3 lD f6 6 0 - 0 �c 7 7 � e 2 d6

1 2... dxe5 1 3 fxe5 lLld7 14 i.f4!) 1 3 exd6 i.xd6 1 4 lLle4 1l..e 7 1 5 b3! c5 1 6 1l..b 2 i.b7 17 f5! exf5 1 8 .l:txf5 and White's pieces point ominously towards the kingside, Ryt­ shagov-Hellsten, Asker 1 997. b) 10 ... b6 (I can find no games with this move, which is hardly surprising as it looks very risky at best) and now: b1) 1 1 i.d2 i.b7 1 2 .l:tae1 lLlc6! (if Black still has the option, ...lLlc6 is often very ef­ fective against i.d2, .l:tae1 plans as it leaves the d7-square available for the other knight; of course, Black is often already committed to playing ...ltJbd7) 13 lLlxc6 11..xc6 looks okay for Black. Note that 14 lLld5? doesn't work here: 14 ... exd5 15 exd5 i.d7 16 'tlr'xe7 .l:tae8 and the queen is trapped. b2) 1 1 e5! (why not?) 1 1 ... dxe5 1 2 fxe5 lLlfd7 13 i.f4! (13 'tlr'e4?! wins a rook but after 13 ... g6 14 'tlr'xa8? i.b7 15 'tlr'a7 i.c5 1 6 i.e3 i.xd4 1 7 i.xd4 lLlc6 White's queen is trapped; this resource is definitely worth remembering) 1 3 ... i.b7 1 4 lLle4 and White's pieces flood into the centre. Black already has to watch out for ideas such as lLlf6+, while 14 ...lLlxe5 15 'tlr'h5 f5 1 6 lLlxe6 looks very dangerous. c) 10 ...lLlbd7 is the most sensible move; Black prevents e4-e5 for the time being: 1 1 i.d2 b6 1 2 l:he1 (renewing the idea o f e4e5) with a further split: c1) 12 ... .l:td8!? provides an antidote against e4-e5; 13 e5? dxe5 1 4 fxe5 'tlr'xe5 1 5 'tlr'xe5 lLlxe5 leaves the d4-knight e n prise; this is a common resource for Black. White can transpose into the main game with 1 3 'it>h1 i.b7, when Black has already commit­ ted his f8-rook to d8. c2) 1 2 ...i.b7 see following diagram

1 3 e5!? (fhis is absolutely critical, al­ though I can find no games from this posi­ tion! 13 'it>h 1 would transpose into the main game.) 13 ... dxe5 14 fxe5 and now we have the following:

c21) 1 4...'tlr'xe5 is less effective here than in the note 'e3' to Black's 1 3th move: 1 5 'tlr'xe5 lLlxe5 1 6 .l:txe5 .l:tad8 1 7 lLlce2 i.c5 (17 ... lLlg4 18 .l:th5 g6 19 .l:th3 i.cS 20 .l:tf4! h5 21 b4 and White wins) 1 8 i.c3 lLld7 1 9 .l:txc5 lLlxc5 20 i.c2 and White's two pieces outweigh the rook and pawn. c22) 1 4... i.c5! 1 5 .l:tf4! (15 exf6 i.xd4+ 16 'it>h 1 lLlxf6 gives White nothing).

In this position both sides must tread ex­ tremely carefully: c221) 1 5 ...lLle8 is 'safe' but not the way to play; it does nothing to oppose White's attack. After 1 6 lLla4 White has a clear plus. c222) 15 ... 'tlr'xe5 16 'tlr'xe5 lLlxe5 17 .l:txe5 .l:tad8 1 8 lLlce2 lLld7 19 ne3 e5 20 .l:th4 g6 21 lLlb3 f5 22 g3! and White is better. c223) 1 5 ...l::. fd8 16 exf6 lLlxf6 17 i.e3 (17 'tli'f2 e5 18 nxf6 .l:txd4! is good for Black) 1 7 ...e5 1 8 l::.x f6 exd4 1 9 lLld5! 'tlr'e5! (other 33

Sicilian Ka n

moves lose: 1 9 ... j_xd5 20 j_f4 �d7 21 �h5 h6 22 cxd5 gxf6 23 �xh6, or 1 9 ...l:txd5 20 j_f4 'iid7 21 cxd5 gxf6 22 'iih5) 20 l:tf5 l:txd5! 21 cxd5 'iix e3+ 22 �ft g6 and Black has a pawn and reasonable compensation for the exchange. c224) 1 5 ...l:tad8 (there are some small differences if Black chooses this rook) 1 6 exf6 ttJxf6 1 7 j_e3 e 5 1 8 l:txf6 exd4 1 9 ttJd5! �e5! ( 1 9. . .j_xd5 2 0 j.h6! j_e6 2 1 j_xg7! �xg7 22 'iih 5 �xf6 23 'iih4+ �g7 24 'iix h7+ �f6 25 'iih 4+ �g7 26 'iig 5+ �h8 27 �h6+ �g8 28 �h7 is mate; 19 ... l:txd5 20 j_h6! l:te5! 21 'iixe5 'iix e5 22 l:txe5 gxf6 23 l:txc5 bxc5 24 j_xh7+ �xh7 25 j_xf8 d3 26 �£2 j_xg2 27 j_xc5 i.e4 28 j_e3 f5 29 b4 is probably a winning ending for White) 20 j_xh7+!? (20 l:tf5 l:txd5 21 cxd5 'iix e3+ is again unclear) 20 ... �xh7 21 �d3+ �g8 22 i.d2 l:txd5 23 l:txe5 l:txe5 and this is still difficult to assess - Black has quite good compensation for the queen. Crazy sruff, but it hammers home one point: Black must be very well prepared and resourceful against these quick-fire attacks from White. White's 'gain' of a tempo by avoiding �g1 -h1 very much raises the stakes - one small mistake from either side could prove to be devastating. Generally in the i.d3 lines White aims for an attack on the king, but it should be pointed out that White can also adopt the anti-Hedgehog set-up with, for example, 10 j_e3 tiJbd7 11 l:tacl b6 1 2 f3 j.b7 1 3 l:tfd1 l:tac8. This position is more often reached via 5 c4 and will be discussed in Chapter 9. 1 0 . . . b6 1 1 f4 lLlbd7

More sensible than 1 1 ...j.b7, which al­ lows 12 e5!. 1 2 i.d2

For 12 b3, see Game 10. The immediate 12 f5 is answered simply by 1 2 ...ttJe5!. I n general White should wait for Black to commit himself to ...j_b7 be­ fore playing f4-f5, so that the light-squared bishop no longer defends e6. 34

It's important for Black to know how to react to the hyper-aggressive 1 2 g4!?, despite the fact that I can find hardly any games with this move.

Black always has to be careful of the g2g4-g5 thrust. This is even more dangerous if Black is already committed to ...ttJbd7 so that the d7-square is not vacant for a ... t2Jf6d7 retreat. Deprived of this square, the knight may have to do with the inferior e8 square, from where it is far less influential. Black's choices are: a) 12 ... d5? 13 cxd5 exd5 1 4 e5! with a clear plus for White. b) 12 ... ttJc5 13 j.b1 j.b7 14 g5 tDe8 (14...ttJfd7 loses a piece to 15 b4; this is the problem with having both knights on the same circuit) 1 5 f5 exf5 (or 1 5 ...e5 16 ttJf3 b5 1 7 tiJd5 j_xd5 1 8 cxd5) 1 6 ttJxf5 j_d8 17 ttJd5 i.xd5 18 exd5 and White has strong pressure on the kingside, Pallova­ Palkova, Chrudim 1 994. c) 12 ... h6!? 13 h4 (13 g5 hxg5 14 fxg5 tDe8 1 5 g6 tDe5 1 6 gxf7 + t2Jxf7 is okay for Black) and now: cl) 13 ...ttJc5 1 4 i.c2 e5 (or 14 ... j.b7 1 5 g5 hxg5 1 6 hxg5 tiJh7 1 7 l:t£2, followed by l:th2) 1 5 tiJf5 i.xf5 1 6 gxf5 and again the strucrure favours White. c2) 13 ... h5! (it's worth sacrificing the h­ pawn to negate White's kingside charge) 1 4 gxh5 ttJc5 1 5 h 6 g6! looks okay for Black. d) 12 ... g6! (this looks best) 13 g5 ttJh5 1 4

5 .1L d 3 ti'J f6 6 0 - 0 � c 7 7 � e 2 d 6

fS lLleS 1 5 f6 .i.dS and Black can be happy. The bishop is misplaced on dS but both knights are on strong outposts. White's pawns on f6 and gS look awesome but it's very difficult for White to make progress and indeed Black could eventually try to dismantle them with a timely ... h7-h6. 12

. . .

�b7

I f Black is concerned about White's pos­ sibility in the next note and doesn't mind transposing into note 'e' to Black's 13th move, then he could employ the move or­ der 1 2 ... .:.cs, answering 1 3 fS with 13 ....i.f8! keeping the structure intact. 1 3 l:tae 1

This is White's most aggressive set-up. Note that the bishop is better placed on d2 rather than e3 as it doesn't obstruct White's major pieces on the c-fJ..!e. It seems that the threat is very much e4-e5, or is it? Sec note 'c' to Black's 1 3th move. As an alternative, attacking with 13 fS!? certainly springs to mind, especially as Black doesn't have time to defend e6 with ... :reS and ....i.f8 (or ...:aeS and ....i.dS). Black can continue with: a) 13 ...e5 14 lLlc2 lL:lcS (14...b5 1 5 cxbS axbS 1 6 lLlxbS doesn't give Black anything like what he needs for the pawn) 1 5 lLle3 bS 16 lLledS .i.xdS 17 lLlxdS lLlxdS l S cxdS gives White a nice space advantage, Milos­ Zapata, Bogota 1 992. b) 13 ... lL:lc5 14 fxc6 fxe6 15 .i.c2! (15 b4?

lLlxd3 1 6 'ifxd3 nfcS! 1 7 lLlxe6 'ifxc4 is good for Black) 15 ... a5 (giving away the bS­ squarc but b2-b4 had to be prevented) 1 6 b3 :res 1 7 a3 .i.f8 t S b 4 axb4 1 9 axb4 lLla6 20 .i.a4 :e7 21 .l:tabt and Black was some­ what uncomfortable in Kudrin-Bakhtadzc, Ycrcvan 1 996. If Black cannot improve in the above lines, then it may well be that 1 3 fS is stronger than the 'automatic' 1 3 :ac t . Food for thought! 1 3 . . . l:tfd8

Black's rook indirectly hits the knight on d4 and thus prevents e4-e5 for the moment. Also, the f8-square is vacated in case of ... .i.f8 or ... lLlf8. However, there arc cer­ tainly other options for Black: a) 13 .. JhcS (this move is too routine) 1 4 cS! lL:lcS (Vehi Bach-Raurell Bernada, Olot 1 992) and now Zeller's suggestion of 15 b3 keeps White ahead - the knight on eS is a poor p1ece. b) 13 ... 'ifc5?!, with the idea of 14 lL:lb3 'ifhS, looks very risky - the black queen is committed to hS and is lacking safe squares, assuming White avoids an exchange of queens. c) 13 ... g6 14 fS! (14 cS lLlhS looks okay for Black) 14 ... exf5 1 5 exfS lL:leS 1 6 .i.gS 'ii'd S 1 7 .i.xf6 .i.xf6 t S .i.e4 .i.xc4 1 9 'ifxc4 and White has an excellent outpost for the knight on dS, Luther-Perdomo, Cali 2000. d) 13 ...:adS (Black wants to follow up with ....:.feS and ....i.f8) 14 fS! eS (1 4...lL:lc5 15 fxe6 fxe6 1 6 .i.c2 aS 17 a3! is good for White) 1 5 lLlc2 lL:lcS (1 5 ... b5 is a sacrificial suggestion from Zeller; after 1 6 cxbS axbS 17 lLlxbS 'ifbS t S lL:lc3 lL:lcS 19 lL:lb4 I don't believe Black has enough for the pawn) 1 6 lLle3 b S 1 7 lL:ledS lLlxdS t S lLlxdS .i.xdS 1 9 cxdS is similar to note 'a' to White's 1 3th move - White has a pleasant space advan­ tage. e) 1 3 ...:fcS!? looks suicidal, but is it really so straightforward? White has the following possibilities: 35

Sicilian Kan

e1) 1 4 fS can be answered by 1 4... �£8!, keeping the structure intact. e2) White should consider the rook lift 1 4 no, for example 1 4... nadS (or 1 4...g6 1 S ng3 .:adS 1 6 fS �f8) 1 S nh3 .tf8 1 6 ltJO g6 17 lDgS lDcS (or 17. ..�g7!? 1S eS dxeS 19 fxeS 'il'xeS 20 'il'xeS lDxeS 21 :xeS h6!) 1 S i.c2 h6 19 eS!? led to great complica­ tions in Toth-Trincado, Mar del Plata 1 990. e3) 1 4 eS (Of course! Doesn't this just win a piece?) 14 ... dxeS 1 S fxeS 'il'xeS! 1 6 'il'xeS lDxeS 1 7 :xeS .:adS

this line. 1 4 ll'lf3

Preparing e4-eS or ltJO-gS. Alternatively: a) 1 4 b4 (expanding on the queenside and preventing ...lDcS) and now: a 1) 1 4...g6?! 1 S fS! exfS (1S ...ltJf8 16 fxe6 fxe6 1 7 lDxe6! lDxe6 1 S ltJdS i.xdS 1 9 exdS and Black can already resign) 16 exfS looks dangerous for Black, Adla-Paramos Dom­ inguez, Mondariz 1 994. a2) I believe Black should play 14 ... i. f8!. Now he is ready to play 1S ...g6, meeting 1 6 fS with 1 6. . .:es!. b) I can find no examples of 14 nO!? pre­ paring to swing the rook: 14 ... .tf8 1 S nh3 g6 16 fS and again Black should retract with 1 6 ... neS! to keep the structure intact. 1 4 . . . g6

So that Black can answer 1 S eS with 1 S ... lL:!hS 16 i.e4 (or 1 6 g4 lL:!g7 17 i.e4 i.xe4 1S 'il'xe4 ltJcS) 16 ... i.xe4 17 'il'xe4 lDcS 1 S 'il'e2 :acS and Black is fine. 1 5 lt'lg5 �fS

(White is a piece for a pawn up but is well and truly skewered on the d-file!) 1 S lDce2 lL:!g4 1 9 nhs g6 20 nh3 i.cS! and now, instead of 20 ... eS? 21 lDb3 which is winning for White (Zeller), Black should play 21 ...i.cS!: e31) 21 ng3 lL:!eS 22 i.b1 (22 �gS?! i.xd4 23 i.xdS nxdS 24 b3 �cS 2S lDf4 fS 26 h3 'li;f7 27 .tb 1 nd2 and White was in some trouble in the game H.Hunt-Ahrens, Bratislava 1 993) i.xd4 23 lL:!xd4 nxd4 24 i.c3 :edS 2S i.xd4 nxd4 and Black has excellent compensation for the slight mate­ rial disadvantage. e32) 21 nf4 hS (or 21 ...fS!?) 22 lDb3 eS 23 :n lDf2+ 24 nxf2 i.xf2 and I prefer Black's rook and pawn over the two knights. If the above analysis in note 'e' holds up, then both 1 2...:es and 13 ... nfeS are of substantial importance in the assessment of 36

1 S ... h6 also looks playable: 16 lDO (hav­ ing provoked a slight weakness, the knight returns; 1 6 lDxf7 'li;xf7 1 7 eS dxeS 1 S fxeS lL:!xeS looks unsound) 16 ... .tf8 17 nc1 .tg7 1 S i.b 1 :acS 1 9 b3 'il'bS (Kudrin-Dzindzi­ chashvili, Philadelphia 1 992) and here White should push on with 20 fS. In view of this, I prefer the prophylactic 1 9 ...:eS!. 1 6 'ii'f 2 �g7 1 7 'ii'g 3 lt'lc5 1 8 �c2

1 8 . . . d5?

5 � d3 f:D f6 6 0 - 0 � c 7 7 � e 2 d 6

Right idea, wrong timing. Stronger is 1 8 ... h6 1 9 lL\f3 and only then 1 9 ... d5!. I f anything, I then prefer Black. 19 cxd5 exd5 20 e5 f?Jfe4 21 f?Jcxe4 f?Jxe4 22 .i.xe4 dxe4 23 .i.c3 l:td3

After this Black is a pawn down with no compensation, but following 23 ...'ir'c4 White can play 24 fS!, for example 24... h6 25 lLlxf7! 'ir'xf7 26 e6 'ir'e8 27 ..txg7 �xg7 28 f6+ �h8 29 'ir'c7. 24 �h4 h6 25 f?Jxe4 �c6 26 �f2 l:tad8 27 f?Jd6 .i.a8 28 f5 g5 29 f6 .i.f8 30 lLlf5 l:te8 31 e6 l:txe6 32 lLle7 + .i.xe7 33 fxe7 f5 34 l:txe6 �xe6 35 �xf5 1 -0

Game 10 Plachetka-Ravikumar

Pofitiken Cup, Copenhagen 1980 '-------•

1 e4 c5 2 f?Jf3 e6 3 d4 cxd4 4 f?Jxd4 a6 5 �d3 f?Jf6 6 0-0 �c7 7 �e2 d6 8 c4 �e7 9 b3

This is the most accurate move order for White to reach the position after the 12th move. Instead 9 lL\c3 0-0 10 b3 allows Black to play 10 ...b5!, exploiting the undefended knight on c3: 1 1 ..tb2 b4 t 2 lLlbt lLlbd7 13 lLld2 ..tb7 1 4 l:tael lLlcS and White cannot shift the powerful knight on cS.

Again w e see the queenside fianchetto from White. I believe it's more dangerous here than against ...g6 and ... ..tg7; Black will find it more difficult to arrange a successful ... e6-e5 break against f2-f4. 1 2 l:tad1

Once more the most aggressive plan is 12 f4 ..tb7 13 .U.ael and now: a) 13. .. .l:tad8 14 fS!? (1 4 �hl l:tfe8! 1 5 ..tbt ..tf8 1 6 lLlf3 g6 1 7 e S lLlhS 1 8 lLlgS ..tg7 was okay for Black in Soylu-Cebalo, Budva 1 981) with a further split: a 1) 14 ...lLlc5 1 5 fxe6 fxe6 16 ..tc2 aS (1 6 ... ..tc8 17 b4 lLlb7 18 ..tb3) 17 a3!, in­ tending b3-b4, is uncomfortable for Black. a2) 1 4 ...e5 1 5 lLlc2 bS! (a suggestion from Zeller; this sacrifice works quite well here) 1 6 cxbS axbS 1 7 b4 (17 lLlxbS? 'ir'b6+ 1 8 �ht lLlcS is good for Black) 1 7 ... d5 1 8 exdS lLlxdS 19 lLlxdS ..txdS 20 ..txbS 'ir'a7+ 21 'ir'f2 'ir'xa2 22 ..txeS lLlxeS 23 l:txeS ..tf6 and Black's bishop pair provide good com­ pensation for the pawn. b) 1 3. .. g6 14 eS (14 fS exfS 1 5 exfS lL\eS 16 ..te4 lLlxe4 1 7 lLlxe4 l:tfe8 18 'ir'd2 ..txe4 19 l:txe4 ..tf8 was level in Zapata-Blanco Fernandez, Matanzas 1 995) 14 ...lLlh5! 1 5 fS?! dxeS (Erdogdu-Mastrovasilis, Antalya 2001) 1 6 fxe6 lLlc5!

9 . . .0-0 10 �b2 f?Jbd7 1 1 f?Jc3

The alternative knight development with 1 1 lLld2 can be seen in Game 1 1 . 1 1 . . . b6

and I believe the complications favour Black: 17 exf7+ l:txf7 1 8 l:txf7 �xf7 1 9 lLlf3 lLlf4 20 "i!VxeS (20 lLlxeS+ �g8 21 'ir'c2 ..td6!) 20... ..td6 21 'ir'd4 lLlcxd3 22 liJbS 37

Sicilian Kan

i.cS!.

20 .l:!.xf7! 'it>xf7 21 'i!fxg4 g6

1 2 . . . i.b7

2t ...lbxe5 22 �hS+ (Plachetka) 22 ... lbg6 23 lbe4 gives White a winning attack.

To include other possibilities I've again fiddled with the move order. The actual sequence was 6 ... d6 7 c4 i.e7 8 lbc3 0-0 9 b3 b6 10 i.b2 i.b7 1 1 �e2 lDbd7 1 2 :ad1 �c7.

22 liJe4 liJc5 23 liJf6 .l:!.ed8 24 .l:!.f 1 'it>g7 25 liJxh7 liJd3

Or 25 ...'it>xh7 26 i.xg6+ 'it>h8 27 �hS+ 'it>g8 28 i. f7 +.

1 3 i. b 1 .I::Ue 8 1 4 f4 .l:!.ac8?!

26 liJxf8 l:bf8 27 .l:!.xf8 l:!.xf8 28 i.xd3

Funnily enough, Black's best plan is to concede a tempo or two and fianchetto the dark-squared bishop in any case! This leads to positions similar to Game 8, although of course the lost tempi mean it's not such a good version: 14 ... i.f8! 1 5 'it>h 1 g6! 1 6 lbf3 :adS 17 lbgS i.g7 1 8 :d2 h6 1 9 lbh3 lbcs and Black had arranged a successful defen­ sive formation, Granda Zuniga-Milov, Vil­ larrobledo (rapid) 1 997.

'i!ff7 29 h4 .l:!.d8 30 i.e4 .l:f.d2 31 i.c1

1 5 liJf3 i.f8 1 6 'it>h 1 i.c6?

5 i.d3 liJf6 6 0-0 'i!fc7 7 'i!fe2 d6 8 c4

Now 16 ...g6 really was vital. 1 7 e5!

This advance is very strong here. Black can­ not resist White's forthcoming attack. 1 7 . . . i.xf3 1 8 .l:!.xf3 dxe5 1 9 fxe5

'i!ff1 + 32 'it>h2 'i!fxc1 33 'i!Vxg6 + 'it>f8 34 'i!ff6 + �g8 35 'i!fxe6 + 'it>g7 36 'ilff6 + 'it>g8 37 e6 1 -0

Game l ! Csoke-Lehoczki

Hungarian League 1998 1 e4 c5 2 liJf3 e6 3 d4 cxd4 4 liJxd4 a6 i.e 7 9 b3 0-0 1 0 i.b2 liJbd7 1 1 liJd2

1 1 f4 b6 1 2 lbd2 will transpose, but note that 1 1 ...:e8? 12 lbxe6! fxe6 13 eS l:tf8 1 4 exf6 lDxf6 1 5 lDd2 led t o a clear advantage for White in Moiseev-Cvitan, Sibenik 1 988. This is a trick worth remembering. 1 1 . . .b6 1 2 f4 i.b7

The actual move order of the game was 9 f4 lbbd7 10 b3 b6 1 1 i.b2 i.b7 1 2 lDd2 0-0. 1 3 l:!.ae 1

1 9 . . .liJg4

Or: a) 19 ... �xe5 20 lbe4 �aS (20 ... �c7 21 lDxf6+ lDxf6 22 l:.xf6! gxf6 23 �g4+ i.g7 24 i.xf6 wins - Plachetka) 21 i.c3 �a3 22 lbxf6+ lbxf6 23 �c2 i.e7 24 b4 llxc4 25 i.xf6 :xc2 26 :xa3 and White wins. b) 19 ... lbxe5 20 lDbS! axbS 21 i.xeS �e7 22 �c2 and White wins - Plachetka. 38

With the knight o n d 2 the presence o f the b2-bishop is felt more - there i s only one knight blocking it. On the other hand,

5 i. d 3 li:::. f6 6 0 - 0 'ii c 7 7 'ii e 2 d 6

White has less conttol over the d5-square. I f Black can arrange a favourable ... e6-e5 he does not have to worry about tLld5 ideas. 1 3 . . J:!.fd8

Again aimed at the advance e4-e5. On this occasion the knight on d4 is protected, but Black can still use tactics to disarm White's thrust. Other possibilities include: a) 13 ... e5?! 14 tLlf5 l:tfe8 15 g4 tLlc5 (1 5 ...'ilfc5+? 1 6 'it>hl i.f8 17 g5 exf4 1 8 l:txf4 g6 [Moiseev-Vasiukov, Sibenik 1 988] 19 gxf6 gxf5 20 l:tg1 + 'it>h8 21 'ilfh5 and Black will be very quickly mated, for exam­ ple 21...tLle5 22 'ilfxh7+ 'it>xh7 23 l:th4+ i.h6 24 l:tg7+ 'it>h8 25 l:txh6 mate) 16 g5 tLlfd7 17 tLlxe 7+ l:txe7 1 8 f5 f6 19 h4 and White has a sttong bind on the position, Anka-Csoke, Hungarian League 1 994. b) 13 ... g6!? (making the h5-square avail­ able) 1 4 e5 tLlh5 1 5 i.e4 (15 f5 dxe5 1 6 fxe6 i.c5! 17 exd7? 'ilfxd7 1 8 'ifxe5 l:tae8! is winning for Black) 1 5 ... tLlc5 1 6 i.xb7 'ilfxb7 (Zeller) and Black can be reasonably happy with his position. c) 13 ...l:tad8 14 e5?! (White should play more quietly with 1 4 i.b 1 or 1 4 'it>h 1 , al­ though this then gives Black the chance to play ... g7 -g6) 14 ... dxe5 1 5 fxe5 'ilfxe5! 1 6 'ilfxe5 tLlxe5 1 7 l:txe5 i.c5 1 8 l:tf4 tLld7 1 9 l:th5 g 6 20 l:th3 e 5 21 l:tfh4 tLlf6 was good for Black in Szieberth-Izsak, Budapest 1995.

i. b 1 , but this i s critical. 1 4 . . . dxe5 1 5 l'i:::. x e6!

15 fxe5 'ilfxe5! 16 'ilfxe5 tLlxe5 17 l:txe5 i.c5 18 l:tf4 tLld7! (Moiseev) and Black will regain the piece with ... e6-e5. 1 5 . . . fxe6 1 6 fxe5

1 6 . . . l'i:::. c 5?!

Black returns the piece and accepts a slight disadvantage. However, I can't find anything wrong with 1 6 ... tLle8! here. a) Moiseev gives 17 'ili'h5 but then Black has the clever defence 1 7 ...tLldf6! 1 8 exf6 'ilfc5+ 1 9 'ifxc5 (19 l:tf2 'ilfxh5 20 fxe7 tLlf6 21 exd8'if+ l:f.xd8 is good for Black) 1 9 ...i.xc5+ 20 'it>h 1 tLlxf6 21 i.e4 tLlxe4 22 tLlxe4 i.xe4 23 l:txe4 l:td2 with a better ending for Black. b) 17 i.xh7+ ! 'it>xh7 1 8 'ilfh5+ 'it>g8 1 9 'ilff7+ 'it>h7 looks best, but White has no more than a perpetual check. An earlier game saw 1 6 ... tLlxe5?! 1 7 i.xe5 'ilfd7 (Moiseev-Vyzmanavin, USSR 1 989) and now 1 8 i.e4! (Vyzmanavin) 1 8...tLlxe4 1 9 tLlxe4 i.xe4 20 'ilfxe4 leaves White with an advantage due to Black's weak e6-pawn. 1 7 exf6 l:ixd3 1 8 fxe 7

After 1 8 f7+ Black plays 1 8 ...'it>f8!, but not 1 8 ... 'it>h8? 19 'ili'g4 i.f8 20 'ilfxg7+!. 1 8 . . . 'iix e7 1 9 b 4 .:adS 2 0 i.c1

14 e5!?

Again White can play with 1 4 'it>h1 or 1 4

So far the players had been following analysis from Moiseev, who assessed this position as slightly better for White. This is correct - White reaches an ending with 39

Sicilian Kan

some chances to convert an extra pawn. In this game White actually loses his pawn advantage but ends up winning anyway after some inaccurate play from Black.

20 . . . ltJa4 21

'ii'x e6 +

'ii'x e6 22 �xe6

�c3 23 .l:!.e7 .l:!.f8 24 l:txb7 l:txf1 + 25 lt>xf1 .l:l.xc1 + 26 lt>e2 .l:!.c2 27 lt>e3 .l:!.xa2 28 .l:l.a7 .l:l.b2 29 .l:!.xa6 l:txb4 30 ltJe4 ltJc5 3 1 .l:l.a8 + �f7 32 .l:!.a7 + lt>g6 33 ltJd6 l:tb3 + 34 'it>f4 l:lb2 35 ltJe8 �f2 + 36 lt>e3 l:f7 37 l:xf7 'it>xf7 38 ltJc7 ltJd7 39 lt>d4 lLlf6 40 h3 h6 41 ltJa8 ltJd7 42 lt>d5 lt>e7 43 �c6 ltJe5 + 44 lt>b5 lt>f6 45 ltJxb6 lt>f5 46 c5 lt>f4 4 7 ltJc4 ltJg6 48 c6 ltJe7 49 c7 'iti>g3 50 ltJb6 lt>xg2 5 1 lt>c5 lt>xh3 52 lt>d6 g 5 5 3 lt>xe7 1 -0

Game 12 Sandor-Chernuschevich

Poznan 1995 1 e4 c5 2 ltJf3 e6 3 d4 cxd4 4 ltJxd4 a6 5 .i.d3 ltJf6 6 0-0 'ii'c 7 7 'ii'e 2 d6 8 c4 ltJbd7 ! ?

A flexible move. Black waits t o see how White plays it before deciding whether to fianchetto or not. In some cases Black can also play an early ... lbe5 or possibly ... tt:'lc5 instead. 9 f4!

By playing f2-f4 immediately, White cuts out any early ... tt:'le5 ideas for Black. I think this is White's strongest move order. Other 40

moves include:

a) 9 lL'lc3 lL'le5!? (9 ... g6 10 l:td1 i.g7 1 1 tt:'lf3 0-0 1 2 .i.f4 transposes to Game 5; for 9 ... b6 1 0 f4 .i.e7 see the note to Black's 9th move) 10 'ith1 .i.e7 1 1 f4 tt:'lxd3 12 �xd3 0-0 13 b3 b6 14 .i.b2 i.b7 1 5 l:tae1 l:tad8 16 'ir'g3 l:tfe8 occurred in Gavrilakis­ Velikov, Kavala 1 990. This defensive set-up looks quite satisfactory for Black. The ex­ change of the knight for White's light­ squared bishop seems to dull White's at­ tacking chances. The game continued 1 7 l:te3 �d7 1 8 �e1 g6 1 9 f5?! tt:'lg4 20 f6? .i.xf6 21 l:txf6 tt:'lxf6 22 �h4 �e7 23 l:th3 tt:'lh5 and Black went on to win. b) 9 b3 g6! (as we already know, the king­ side fianchetto works well against White's queenside fianchetto) 10 i.b2 .i.g7 1 1 tt:'lc3 0-0 transposes to Game 8. 9 . . . g6!?

Now White has committed himself to playing f2-f4, Black logically goes for a king­ side fianchetto. By now we know that this defensive set-up works well against a king­ side storm. However, on this particular oc­ casion Black has problems with the defence of e6 due to the early development of the d7-knight. If White proceeds energetically then Black cannot keep his pawn structure intact. 9 ... h5!? is a very ambitious move cham­ pioned by the German IM Karl Heinz Pod­ zielny.

5 i. d3 tiJ f6 6 0 - 0 �c 7 7 � e 2 d 6

10 'it>h 1 ! (10 b3 'ii'b6! 1 1 i.e3 lt:\g4 shows a point of Black's 9th move: 1 2 lt:\c3 g6 1 3 l:.adl lt:\xe3 1 4 'ii'xe3 i.g7 and Black was fine in Leko-Podzielny, Dortmund 1991) 10 ... b6 1 1 lt:\c3 i.b7 12 fS! lt:\cs (12... eS 13 ltJdS ltJxdS 14 exdS ltJf6 l S ltJf3 h4 16 ltJgS i.e7 17 i.e3 ltJd7 18 lt:\e4 left White with a massive knight on e4, Z.Szabo-Lehoczki, Solymar 1 996) 13 i.c2 exfS 14 ltJdS i.xdS l S exdS+ 'ii'e7 1 6 'ii'x e7+ 'it>xe7 1 7 i.gS and White has a terri­ fic ending, Quinteros-Debarnot, Sao Paulo 1 972. 9 ...i.e7 10 lt:\c3 b6 could transpose into lines considered earlier on in the chapter, but let's see what happens if Black delays castling: a) 1 1 'it>h 1 is important because it can be reached by transposition from lines where White plays a very early 'it>h 1 : l l .. . ..ib7 and now: al) 1 2 i.d2 g6!? (12... 0-0 transposes to Game 9) with a further split: al l) 13 l:tacl hS! (a Podzielny speciality; possible ideas include ... hS-h4, followed by ...ltJhS-g3+ or simply ... h4-h3) 1 4 b4 h4 l S a3 l:tc8 1 6 .U.cel 'ii'b 8 1 7 ..icl b S 1 8 ..ib2 bxc4 1 9 i.xc4 'ii'a8 20 ltJf3 h3 21 g3 0-0 22 ..id3 aS! 23 bS tLlcS and Black was doing well in Siemes-Podzielny, correspondence 1 998. a12) 13 l:tael (this looks more consistent than 13 l:.acl) 13. .. hS!? 14 eS (14 ltJdS?

exdS l S exdS 0-0 1 6 'ii'xe7 l:.ae8 wins for Black - Zeller; 1 4 fS allows Black to dem­ onstrate a point of ...g7 -g6: 14 ...gxfS! l S exfS e S 1 6 ltJb3 h4 and Black aims for counterplay against g2) 1 4 ... dxeS l S fxeS lt:\g4 16 tLlxe6!? (16 ..if4 is less ambitious) 16 ... fxe6 17 i.xg6+ 'it>d8 18 i.f4 (Zeller) with a very unclear position. White has two pawns and Black's king is not safe. On the other hand, Black is a piece ahead and is quite active himself. a2) I prefer the move 1 2 fS!, for example 1 2...eS (or 12 ...ltJcS 13 fxe6 fxe6 14 ..ic2 aS l S lt:\a4!) 13 lt:\c2 ltJcS 14 lt:\e3 looks better for White - compare to the note to White's 1 3th move in Game 9. b) 1 1 i.d2 (again White does without the 'safe but lazy' 'it>h 1) l l ...i.b7 1 2 l:.ae1 l:.d8?

(Black should play 12 ... 0-0, transposing to note 'c2' to White's 1 0th move in Game 9, or 1 2 ...g6; 1 2...l:.d8 prevents e4-eS but...) 13 ltJdS!! exdS 14 exdS 0-0 (14 ...ltJcS l S ..ic2 l:.d7 1 6 b4!) l S ltJfS! and the rook on d8 is a real liability - the e7-bishop cannot escape: 1 S ...l:.fe8 16 lt:\xe 7+ 'it>f8 17 l:.f3! lt:\g4 (17 ...lt:\cs 1 8 .U.e3 l:td7 19 lt:\g6+! Zeller) 1 8 h3 ltJdf6 1 9 hxg4 lt:\xg4 20 l:.h3 and White won in Fogarasi-Erdelyi, Switzer­ land 1 992. 1 0 f5!

It's critical that White plays this now be­ fore Black has a chance to play ...i.g7 to defend f6. After 10 lt:\c3?! i.g7 1 1 fS Black 41

Sicilian Kan

keeps the order with 1 1 ...lDeS!.

32 'i'b4 .i.c6 33 .l::.g4 i.g5 34 i.xc6 + .l::.x c6 35 'i'xb7 'i'c2 36 'i'b8 + 'it>e7 37 .l::.f 1 'i'c5 + 38 'it>g2 .l::.c 8 39 'i'b7 + .l::. c 7 40 'i'f3 h6 4 1 h4 1 -0

Game 13 Schlosser-Vyzmanavin

Sochi 1989 1 e4 c5 2 tt'lf3 e6 3 d4 cxd4 4 tt'lxd4 a6 5 .i.d3 'i'c7 6 0-0 tt'lf6 7 'i'e2 d6 8 f4!?

1 0 . . . gxf5?!

1 0...cS looks stronger: 11 lLJc2 il..e 7 12 lLle3 lLJcS 1 3 il..c 2!? (13 lLJc3 looks safe for an edge) 1 3. .. lLlcxe4!? (13 ...gxfS 1 4 exfS left White with a solid bind over the dS-square in Varavin-Tunik, USSR 1 991) 1 4 il..xe4 lLlxe4 1 S lLJdS 'ticS+ 1 6 .ie3 'ii'c6 1 7 f6! (17 fxg6 hxg6 1 8 il..b 6 il..e 6! as 1 9 lLJc7+ 'it>d7 20 lLlxa8 .l:!.xa8 21 'ir'e3 il..gS is winning for Black) 1 7 ...il..d 8 1 8 'ir'f3 lLJcS 1 9 b4 lLJe6 20 lLld2 and White has good compensation for the pawn (analysis by Varavin). l O ... lLJeS? is of course impossible because of 1 1 fxe6.

tt'lxd5 1 6 cxd5 'i'xd5 1 7 .i.c4 'i'c6 1 8

A dangerous move. White delays c2-c4 and waits for Black to commit himself with the f8-bishop. After 8 ...il..e7 White will probably transpose back to normal lines with 9 c4, but what happens if Black tries to fianchetto? 8 lLlc3 will be discussed under S lLJc3 'ii'c 7 6 .id3 lLJf6 7 'iVe2!? d6 in Chapter 6. One further idea is 8 lLJd2 g6 9 a4!? il..g7 10 aS 0-0 1 1 lLJc4 lLJbd7 1 2 c3 dS! 1 3 exdS exdS 1 4 lLle3 lLleS and Black was more than comfortable in Geller-Vyzmanavin, Mos­ cow 1 989 - White's pieces are not well placed to battle against the IQP.

exd7 + i.xd7 1 9 i.f7 + 'it>d8 20 i.e3

8 . . . g6! ?

As far as I can see, Black is a pawn up but in total disarray. He does well to last as long as he does.

1 1 exf5 e5 1 2 tt'le6!

This idea is worth remembering; Black is already in some trouble. 1 2 . . .'i!i'c6?

More resolute is 1 2... fxe6 13 fxe6 JJ..g7 1 4 exd7 + il..xd7 1 S lLJc3 il..c 6 (Escobar­ Filgueira, Buenos Aires 1 989) although I think White is still doing well after 1 6 .igS!. 1 3 tt'lc3 i.e7 1 4 tt'ld5! fxe6 1 5 fxe6

20 . . . .l::.c 8 21 'i'f2 .l::.c 7 22 .i.b6 'i'e4 23

This looks risky, but Black has hidden re­ sources. Note that 8 ...'iVb6?! 9 il..e3 'ii'xb2?? loses to 10 lLlb3, trapping the black queen.

.l::.fe 1 'i'h4 24 g3 'i'f6 25 'i'c2 .i.c6 26

9 f5!

i.xc7 + 'it>xc7 27 i.d5 'i'g6 28 l:.e4 'it>d7

Direct and dangerous. Black must be careful not to be blown off the board.

29 'i'b3 .l::. c8 30 .l::. d 1 'it>e8 31 'i'c4 i.d7 42

5 i.. d3 Ci:J f6 6 0 - 0 'fie 7 7 'fl e 2 d 6

1 2...lt:lg4!?. I can only find one example o f this move, which leads t o immense compli­ cations: 13 'i1Vxg4 lt:lxd4

9 . . . i.g7

It's imperative for Black to seek devel­ opment. After 9 ...gxf5 10 exfS eS 1 1 c4! (getting a grip over dS) 1 1 ....ie7 12 lLlc2 lt:lbd7 1 3 lt:lc3 tLlcS 14 lt:le3 lLlxd3 1 5 'ii'xd3 'ii'c S 1 6 b4! 'ii'xb4 17 lLledS lLlxdS 1 8 lLlxdS 'ii'c S+ 19 .ie3 Black was already on the verge of losing, Adorjan-Honfi, Hungarian Ch. 1 973. 1 0 fxe6 fxe6 1 1 .i.c4

With the e6-pawn under pressure, Black already looks to be in some trouble. How­ ever, he has a tactical trick to ease his diffi­ culties. 1 1 .ie3 is less dangerous: 1 1 ... 0-0 1 2 lLld2 .id7! 1 3 .ic4 d S 1 4 .ib3 lLlc6 was level in Fleck-Chuchelov, Germany 2000. 1 1 . . . Ci:Jc6! 1 2 i.e3

1 2 lLlxc6 bxc6 1 3 lLlc3 0-0 14 .ib3 aS 1 5 'it>h1 .ia6! 1 6 .ixe6+ 'it>h8 1 7 .ic4 .ixc4 18 'ii'xc4 lLlxe4 was an early draw in De­ graeve-Chuchelov, Bethune 1 999. Let's continue a few moves: 19 lLlxe4 dS 20 'i1Ve2 l:txf1 + 21 'iiVx fl l:tf8 22 'i1Ve2 l:te8 and, if anything, Black is better. The point of Black's previous move is seen in the line 1 2 lLlxe6? .ixe6 13 .ixe6 ltJd4! 1 4 11Yc4 11Yxc4 1 5 .ixc4 lLlxc2 and Black wins material.

and now: a) 1 4 .id3 lLlxc2 1 5 .ixc2 (15 .l::tc l? lLlxe3 1 6 l:txc7 lLlxg4 1 7 l:txg7 ltJeS! 1 8 .ie2 lLlf7 and the rook is nicely trapped on g7) 1 5 ...11Vxc2 16 lLld2 (16 l:tc1 11Vxb2 17 l:txc8+ 'it>d7!! wins for Black) 16 ...'ii'd3 1 7 'i1Vg3 and White certainly has some compensation for the pawn. b) 14 lt:la3 lLlxc2 1 5 lLlbS!? with a further split: b1) 1 5 ...'i*'d7? loses to 16 .ixe6!. b2) 15 ... axb5 16 .ixbS+ .id7 1 7 11Yxe6+ '1t>d8 1 8 .igS+ '1t>c8 19 .ixd7+ 11Yxd7 20 'i*'xd7+ 'it>xd7 21 l:tf7+ 'it>e8 (21 ...'1t>c8? 22 l:tcl was good for White in Lind-Astrom, Helsingborg 1 991) 22 l:te 7+ 'it>f8 23 l:tfl + 'it>g8 24 .if6! .ixf6 25 l:txf6.

1 2 . Ci:Jxd4 .

.

Of major interest in this line is the fact that Black can even go onto an immediate counter-attack here with the enterprising 43

Sicilia n Ka n

Despite the minus piece, White has at least a draw, but maybe more, for example 25 .. .lha2 (25 ....1:tf8 26 .l:txf8+ �xf8 27 .l:txb7 can only be good for White) 26 .l:f.xd6 �a8 27 .l:txb7 lLle3 28 .l:f.dd7 lLlg4 29 g3 lLle5 30 .l:f.dc7 and White can push the b­ pawn. b3) 1 5...'ii'c 6! 16 i.xe6 (16 'ii' f4 .l:tf8 1 7 lLlxd6+ �d7!) 1 6. . .lLlxe3 1 7 i.f7+ �e7 1 8 'ii'h 4+ �d7 1 9 'ii'h 3+ �e7 is a draw by perpetual, as 1 9 ...lLlf5 looks like a very du­ bious winning attempt. 1 3 ..bd4 0-0 1 4 tt:ld2 �d7 1 5 c3

On 1 5 e5 Black should play 1 5 ...ltJh5!. 1 5 . . ..tt:lh5

15 ...d5?! is not to be recommended: 16 i.xf6 dxc4 1 7 i.xg7 �xg7 1 8 l:lxf8 .l:f.xf8 19 'ii'xc4 'ii'xc4 20 lLlxc4 i.c6 21 e5 .l:f.f5 22 .!:tel l::tg5 23 g3 left Black with no compen­ sation for the pawn in D.Belov-Hermans, correspondence 1 998.

Vyzmanavin criticised this move and suggested 1 6 ...�h8 as an improvement, assessing the position as unclear. 1 7 i.. b 3 .t>h8 1 8 l:!.ad 1 ? !

1 8 i.xg7+ 'itxg7 1 9 e5! dxe5 20 lLlf3 (Vyzmanavin) gives White a clear edge. 1 8 .. . . tt:lf6 1 9 h3

Now 1 9 e5? can be met by 1 9 ...lLlg4!. 1 9 . . . e5 20 �b6 'ilfb7!

Black has equalised due to the possibility of pressurising e4 with ... i.c6. 2 1 lLlf3?

21 i.a5!, intending i.b4, is stronger. 21 . . . i.. c6 22 .:!.xd6 tt:lxe4 23 l:!.xc6

The only move. 23 l::tdd 1 ltJg3 24 .l:tf2 lLlf5 25 'ii'c5 l::tac8! is winning for Black. 23 . . .. 'ilfxc6 24 l:!.d1

16 'ilfe3

1 6 'ii'g4!? i.xd4+ 17 cxd4 lLlg7! defends the vulnerable e6-pawn and guarantees Black equality, Nabours-Edwards, corre­ spondence 1 9 9 1 .

24 .. . . i.. h 6!

It would have been very easy to miss this resource. Now Black is clearly better and goes on to win quite comfortably. 25 'ilfxh6 'ilfxb6 + 26 .t>h2 'ilff6! 27 i.. c 2 'ilff4 + 28 'ilfxf4 l:!.xf4 29 tt:lxe5 l:!.e8 30 tt:lf3 tt:lc5 31 l:td5 tt:la4 32 .t>g3 l:!.f6 33 �xa4 bxa4 34 tt:ld4 a3! 35 bxa3 l:!.e3 + 1 6 . . . b5?!

44

36 .t>g4 .t>g7 37 l:!.d7 + l:!.f7 38 l:!.xf7 + 'it>xf7 39 tt:lb3 l:!.xc3 40 a4 l:!.c2 0-1

5 j_ d3 lD f6 6 0 - 0 'ii c 7 7 'ii e 2 d6 Summary

As far as I can see, Games 1 - 5 prove that Black is still very much alive and kicking in 'the main line'. It's true that White has some dangerous plans, but with expert treatment Black can hold his own. Of White's alternative plans, I would say that 1 0 i..e 3 (Game 6) has more chance of keeping an opening edge than either 10 f4 (Game 7) or 9 b3 (Game 8). Classical development with 8 ..i..e 7 (Games 9- 1 1) has perhaps been unfairly neglected in recent times (perhaps due to some high-proftle losses) and I believe these systems are also fully playable for Black. It must be added, though, that Black has to be especially careful not to be mowed down on the kingside in some lines. Special attention to correct move order is required here. .

1 e4 c5 2 lDf3 e6 3 d4 cxd4 4 lDxd4 a6 5 .i.d3 lDf6 6 0-0 'ii'c 7 7 'ii'e 2 d6 8 c4

(D)

8 f4 - Game 13 8 . g6 . .

8 ...i..e 7 9 ciJc3 - Game 9 9 b3 0-0 10 i..b2 lt::lb d7 1 1 ciJc3 - Game 1 0; 1 1 ciJd2 - Game 1 1 8 ...ciJbd7 - Game 12. 9 lDc3

9 b3 - Game 8. 9 . . . ..tg7 1 0 l:l.d1

1 0 f4 - Game 7; 10 i..e3 - Game 6 1 0 . . . 0-0 1 1 lDf3

(D) lDc6

1 1 ... ciJbd7 - Game 5 1 2 h3

12 i.. f4 - Game 4 1 2 . . . lDd7 1 3 ..te3

13 i.. f4 - Game 3 1 3 . . . b6

13 ... ciJde5 - Game 2 1 4 l:l.ac 1

(D) - Game 1

8 c4

1 1 ttJf3

14 'nac 1

45

CHAPTER TWO

I

5 i.d3 ltJf6 6 0-0 'JJIJc 7 : Seventh Move Alternatives

1 e4 c5 2 tt:lt3 e6 3 d4 cxd4 4 tt:lxd4 a6

5 ..td3 �c7 6 0-0 tt:lt6 7 �e2 ..tc5!?

5 ..td3 tt:lt6 6 0-0 �c7

In this chapter we deal with alternatives for White to 7 'ir'e2 and alternatives for Black after 7 'ir'e2. Black's choices are restricted after 7 'ii'e 2 as he generally feels obliged to do some­ thing about the 'threat' of e4-e5. The two options we shall look at here are 7 ...�c5!? (Game 1 4) and 7...�d6!? (Game 1 5). White's options at move 7 are much less restricted. The most important of these is the immediate setting up of the Maroczy Bind with 7 c4 (Games 1 6- 1 7). Of course Black can simply comply with 7 ... d6, most probably transposing at some point to lines considered in Chapter 1 . However, 131ack also has the important independent idea of 7 .. .'�Jc6! and it's this move we shall study here. Also in this chapter we shall take a look at 7 'it>h 1 (Game 1 8) and 7 f4 (Game 1 9), while less important options for White are discussed in the notes to Game 1 9.

Game 14 Vujanovic-Verdier

Comspondence 2000 1 e4 c5 2 tt:lt3 e6 3 d4 cxd4 4 tt:lxd4 a6 46

This move crops up an awful lot in the Kan (its most popular form is the 5 i.d3 .tc5 - see Chapter 4). The idea is to drive White's knight away from d4 to an 'inferior' square (usually b3) before deciding on where to put the bishop. We will also see many examples of a very similar idea involv­ ing ... ifd8-b6 and, after lbb3, simply drop­ ping the queen back to c7. The only reason this particular move or­ der is not more popular is that it allows White to play an early e4-e5, hitting the f6knight. 8 tt:lb3

In general this is the most popular re­ sponse to an early ... i.c5. White regains the

5 1J.. d3 li:J f6 6 0 - 0 1Wc 7 : S e v e n th M o v e A lt e r n a tives

tempo lost on moving the knight again by attacking the cS-bishop. The downside to this move is that the knight is slightly less influential on b3 than it is on d4; the advan­ tage is that it is also much less vulnerable on b3. 8 �e3 gives Black fewer problems. At some point White usually cannot make any progress without offering the exchange of dark-squared bishops and this eases Black's defensive task. After 8 ...d6 9 ltJc3 we have transposed to a line srudied in Chapter 6. B . . . i.a7

8... �e7?! 9 eS! ltJdS 1 0 c4! ltJb4 11 �f4 ltJxd3 1 2 'iWxd3 ltJc6 1 3 ltJc3 left Black pas­ sively placed in Tong Yuanming-I.Gurevich, Singapore 1 990. The game continued 13 ...gS!? 14 �g3 hS 1 S h3 h4 16 �h2 ltJxeS 17 'iWe2 f6 18 cS! b6 19 cxb6 'iWxb6 20 �xeS fxcS 21 'iWxeS �f8 22 �ad1 �b7 23 ltJe4 and Black was left with many weak­ nesses. 9 e5!?

The only way to 'punish' Black for his move order. Black has allowed c4-cS and so White goes ahead. 9 �e3 �xe3 10 'iWxe3 d6 leads to similar lines to those discussed in Chapter 4, as does 9 c4 and 9 ltJc3. 9 . . li:Jd5 1 0 c4 .

White continues to harass the knight. I f White i s looking for a n alternative way to play, I quite like the look of 10 ltJa3!? with

the intention of ltJc4 and preparing to an­ swer lO ... bS with 1 1 c4!. 10 . . . li:Jb4 1 1 i.f4 li:Jxd3

1 1 ...ltJ8c6?! 12 ltJ1d2 ltJxd3 13 'tixd3 fS 1 4 'iWg3 0-0 1 S cS! left Black in an uncom­ fortable position in Kudrin-I.Gurevich, Philadelphia 1 990 - Black's bishops arc finding it hard to breathe. In the game the American GM Ilya Gurevich tried to rem­ edy the siruation with l S ... bS but after 1 6 cxb6 �xb6 1 7 :act 'tib8 1 8 ltJc4 �c7 1 9 h 4 White maintained a strong positional bind. In a later game, Gurevich tried to im­ prove on Black's play with 1 1 ...fS!? although after 12 ltJ 1 d2 (12 ltJc3 may be even stronger; after 12 ... ltJxd3 13 'tixd3 0-0 14 �acl ltJc6 1S �fc 1 White is ready to play c4-cS without giving up control of the ciS­ square) 1 2...ltJxd3 13 'ii'xd3 0-0 14 cS ltJc6 1 S ltJc4 ltJe7 1 6 ltJd6 ltJg6 1 7 'iVg3 b6 1 8 cxb6 �xb6 1 9 �acl White was better, Favaro-I.Gurevich, Maringa 1 991 . 1 2 11i'xd3 f6

Black must react quickly or else the eS­ pawn will promise White a continual bind on the position. 1 3 11i'f3

After 1 3 'ii'g3 Black can hit out with 1 3 ...gS! (Gurevich), for example: a) 14 �d2 'ii'x eS 1 S 'ii'x cS fxcS 1 6 �xgS ltJc6 and Black's centre pawns control some important squares. b) 14 ltJc3!? d6! (14...gxf4 isn't men­ tioned by Gurevich, but 1 S 'ii'g7 �f8 16 exf6 ltJc6 1 7 ltJdS! 'ii'd6 1 8 �fc 1 , intending �ad 1, looks very dangerous for Black) 1 S ltJe4 gxf4 and Gurevich assesses this as unclear. Let's take things a bit further: 1 6 ltJxf6+ Wf7!? 1 7 'ii'xf4 dxeS 1 8 ltJdS+ (18 'ii'h6!?) 18 ...exf4 1 9 ltJxc7 b6 20 ltJxa8 �b7 21 ltJc7 l:tc8 22 ltJxe6 Wxc6 23 �fcl + Wf7 24 �acl ltJd7 and I prefer Black's two bish­ ops to the rook and two pawns. Black's dark-squared bishop can emerge via b8. 1 3 . . . fxe5 1 4 11i'h5 + g6 1 5 11i'xe5 47

Sicilian K a n

1 5 i.xe5?? d6 wins material for Black. 1 5 . . . 'i'xe5 1 6 .txe5 0-0 1 7 lt:\c3

An improvement over 17 tLl 1 d2 bS 1 8 c5 i.b7 19 i.d6 (Ulibin-I.Gurevich, Santiago 1 990), and now Gurevich gives 1 9 ...l:tc8! 20 a4 b4 as being equal. Black can play ... i.dS and can follow up with the manoeuvre ...tLlc6-d8-f7.

Gurevich assesses this position as slightly better for White and I agree with this. Po­ tentially Black's bishop pair could prove useful, but in the short term he has prob­ lems developing on the queensidc. 1 9 . . . b5

Black must try to develop the c8-bishop . 20 �ac1

After this move Black manages to equal­ ise. I prefer 20 c5!?, blocking in Black's dark-squared bishop. 20 . . . lt:\d4 21 lt:\xd4 .txd4 22 i.a3 �f4 23 lt:\d6 b4! 24 �xb4 �xb2 25 �c2 i.g7 26 g3 l:td4 27 i.c5 l:td3 28 l:tb 1 i.e5 29 l:tb6 a5 30 f4 l:txd6! 31 l:txd6 .txd6 32 .txd6 i.a6 Vz - Vz

Game 15 Magomedov-Khouseinov

Dushanbe 1999 1 7 . . . lt:\c6

1 e4 c5 2 lt:\f3 e6 3 d4 cxd4 4 lt:\xd4 a6

Gurevich considers the pawn sacrifice 17 ... b5!? but rejects it on account of 18 cxb5 axb5 19 tLlxb5 i.a6 20 a4 (Nunn and Galla­ gher stop here, saying that Black doesn't have enough for the pawn) 20 ... l::t f5 (20...tLlc6 21 i.d6 followed by lLlxa7 [Gure­ vich] simply leaves White a pawn up) 21 i.d4. However, I believe that 21 ...i.xb5 22 axb5 l::txb5 is at least okay for Black as White cannot really exploit the pin on the a­ file: 23 tLld2 tLlc6 24 i.c3 d5 and Black will follow up by moving the aS-rook. Going back a bit, 20 tLlc7 is also impor­ tant: 20 ... i.xf1 (20...tLlc6 21 i.d6 i.xfl 22 l::txfl leads to the same thing) 21 l:txfl tLlc6 22 i.d6 (22 lLlxaB?! lLlxe5 23 tLlc7 tLld3 and Black regains the pawn) 22...l:tab8 (22 ... l:txf2? 23 l:tx£2 i.x£2+ 24 ..ti>x£2 l:txa2 25 i.a3! traps the rook on a2: 25 ...tLle5 26 tLlb5 lLld3+ 27 ..ti>e2 lLlxb2 28 tLlc t !) 23 i.xf8 ..ti>xf8 and Black has good compensa­ tion for the pawn - White's p1eces arc somewhat uncoordinated.

5 �d3 lt:\f6 6 0-0 'i'c7 7 'i'e2 i.d6! ?

1 8 �d6 �f5 1 9 lt:\e4 48

A n idealistic move. Black activates his dark-squared bishop along the h2-b8 diago­ nal but on this occasion it walks into a not­ too-subtle reply by White. 7 ... d5 is the typical Sicilian freeing break. However, Black is behind in development so opening the position like this is asking for trouble. Play continues 8 tLlc3 (8 e5 tLlfd7 9 f4 may also be good for White but it makes sense to keep the tension) 8 ... tLlc6

5 i.. d3 li'J f6 6 0 - 0 'ii c 7: S e v e n th Mo v e A lt e rn a tives

and now: a) 9 l2Jxc6 bxc6 transposes to a position which is normally reached via the Taimanov Variation: 1 e4 c5 2 l2Jf3 e6 3 d4 cxd4 4 l2Jxd4 l2Jc6 5 l2Jc3 a6 6 l2Jxc6 bxc6 7 .id3 'ilic7 8 0-0 l2Jf6 9 'ii'e 2 d5 and can also be reached by the Kan in another way (1 e4 c5 2 l2Jf3 e6 3 d4 cxd4 4 l2Jxd4 a6 5 l2Jc3 'ii'c7 6 .id3 l2Jc6 7 l2Jxc6 bxc6 8 0-0 d5 9 'ii'e 2 l2Jf6) . This will be studied in Game 57. b) 9 .ie3!? with a further split: b1) 9 ... .ie7 10 exd5 exd5 1 1 l2Jxc6 bxc6 12 .ic5 .ig4 1 3 'ii'e3 .ie6 1 4 f4 .ixc5 1 5 'ii'xc5 'ifa7 1 6 l2Ja4 ..t>d8 1 7 :fe1 l::.e 8 1 8 b4 left White with a pleasant grip on the dark squares in G.Hernandez-Fernandes, Bu­ caramanga 1 992. The immediate 12 ....ie6 may be stronger. b2) 9 ... dxe4 10 l2Jxe4 .ie7 1 1 l2Jxc6 'ifxc6 1 2 .id4! and White's development advantage IS quite apparent, Fercec­ Haldemann, Areo 1 999. 7 ...l2Jc6?! allows White to play e4-e5: 8 l2Jxc6 dxc6 9 eS! l2Jd7 10 .if4 .ie7 1 1 l2Jd2 g5?! 12 .ig3 h5 13 h3 l2Jb6 1 4 l2Je4! and, with l2Jf6+ to follow, White was well on top in the game Kalesis-Karatzas, Aegina 1 993. 8 f4!?

Of course! Now the threat of e4-e5 forces Black to move the bishop again. If White is looking for a quieter way to an advantage then 8 ..t>h1 isn't bad: 8 ...l2Jc6 9 l2Jxc6 dxc6 10 f4 e5 1 1 f5

and White has an edge, Arnason­ Dizdarevic, Manila Olympiad 1 992. I'll talk more about this set-up later, but I will say something here. Note that when compared to Game 1 6, here White is not hampered by his own pawn being on c4. There is no out­ post for Black on d4 and the White's light­ squared bishop is more active. 8

. . .

i.c5 9 i.e3 'iib 6

All games I found continued with this pawn-grabbing idea but as far as I can see it's not at all clear that this is Black's best option. I can't find any examples of 9 ... d6!? but it's certainly not a bad move. a) 10 l2Jb3 .ixe3+ 1 1 'ii'xe3 would trans­ pose into an reasonable version of the 5 .id3 .ic5 line (Black has lost a move with ... .ie7-d6 but gained it back by playing ... .ic5xe3 rather than ....ic5-a7xe3) b) 10 l2Jc3 0-0 and we reach a variation similar to the one seen in Game 54, except that White is a tempo up. It's not clear whether this is enough to sway the balance in White's favour. One point to note is that 1 1 ..t>h 1 b5 12 f5? is answered by the sur­ prisingly strong 1 2 ...'ifa7!. c) White can also consider ideas such as 10 l2Jd2!? and 10 c4!?, but these are hardly a refutation of 9 ... d6. 1 0 c3 li'Jc6 1 1 'it>h 1 ! li'Jxd4 1 2 cxd4 i.xd4 1 3 i.xd4 'iix d4 1 4 li'Jc3

Black has won a pawn but that is where 49

Sicilia n Kan

the good news ends. He is heavily behind in development and has trouble defending all the weak dark squares in his position. I would go as far as saying that this line is only to be recommended for White. 1 4 . . . 0-0

1 9 .. J:tb8 20 'ifd3 g6 21 ltJc5 White is in complete control. 20 exd6 �d7 21

lLlc5 �b5 22 �d3

'ifxd6 23 �xh7 + �hB 24 'i!Ve3 'ife7 25 'ifh3 lLlf6 26 lLle4! �fdB 27 �fe1 �c6 2B lLlg5 �eB

How much trouble Black is in can be gauged from the following two sidelines. a) 14 ... d6 15 �ad1 'ifc5 16 e5 dxe5 1 7 fxe5 ltJd7 1 8 ltJe4 'ifxe5 1 9 i.b5!! axb5 20 ltJd6+ 'it>f8 21 l:1xf7+ 'it>g8 22 �xd7! 'iff6 23 'ifxe6 +! 1 -0 Steinbacher-Auchenberg, Copenhagen 1 994 b) 14 ... h5!? (you cannot be serious!) 1 5 e5 ltJg4 1 6 ltJe4 f6 17 �ad1 'it>e7 1 8 ltJd6 'ii'e3 19 'ii'c 2 fxe5

There are two very surpnsmg things about this position. The first is that it had all been seen before in a previous game. The second is White's next move ... 29 .:ta1 ! !

(Chiburdanidze-Palos, Baden-Baden 1990) and now I like 20 i.f5! exf4 (or 20 ... ltJh6 21 i.xe6! dxe6 22 'ifg6) 21 i-xg4 hxg4 22 'ifg6 and Black suffers horribly due to his lack of development.

This 'novelty' decides the game. The idea behind moving the rook back to its home square is that White requires both rooks to participate in the attack against the black king. Thus 29 �a 1 rules out any exchanges! 29 �d3?! �xd3 30 i.xd3+ 'it>g8 31 i.h7+ 'it>f8 32 i.f5 'it>g8 33 i.h7+ 'it>f8 34 i-f5 'it>g8 35 �xe6 'ifc7! allowed Black to survive in Luther-Barges Mateos, Elista Olympiad 1 998.

1 5 �ad 1 'i!Vb4 1 6 e5 lLleB

29 . . .�d6 30 �e5! .:.adS 31 �ae 1 'ifc7

1 6 ... ltJd5 ruins Black's strucrure and leaves him devoid of any defenders on the kingside: 17 a3 'ifc5 1 8 ltJxd5 exd5 1 9 'ii'h 5 g6 20 'ifh6 and the threat of f4-f5 is hard to deal with.

32 lLle4! �d 1 33 lLlxf6 �xe 1 +

1 7 a3 'i!Ve7 1 8 lLla4!

Continuing to attack the dark squares. 1 8 . . . 'i!Vd8

Narurally 1 8 ... b5? is met by 1 9 'ife4!. 1 9 �b1 d5!?

A desperate bid for freedom. After 50

Other moves also lose: a) 33 ...'ifxe5 34 i.f5 mate. b) 33 ...gxf6 34 i.g6+ 'it>g7 35 'ii'h 7+ 'it>f8 36 �xe6! fxe6 37 'ifh8+ 'it>e7 38 'ifg7+ 'it>d6 39 �xd 1 + 'it>c6 40 �cl + and White Wins. 34 lbe 1 gxf6 35 i.f5 + �gB 36 'ifh7 + 'it>fB 37 �xe6 'ifxf4 38 'ifhB + �e7 39 �xf7 + ! �xf7 40 'i!Vh7 + �fB 4 1 'i!Ve7 + 1 -0

5 i. d3 Ci:J f6 6 0 - 0 � c 7: S e v e n th Mo ve A l tern a tives

Game 16 Smirin-Goldin

could put his c-pawn back on c2, then these problems would disappear. Alas, pawns cannot move backwards!

Philadelphia 1998 1 e4 c5 2 Ci:JfJ e6 3 d4 cxd4 4 Ci:Jxd4 a6 5 i.dJ Ci:Jf6 6 0-0 �c7 7 c4! ?

The main alternative to 7 'ii'e2. White immediately sets up the Maroczy Bind and keeps options open regarding the placement of his queen. If Black now plays 7 ... d6 then we reach positions very similar to those in Chapter 1 , except that White has the slight advantage of being able to choose whether to play 'ii'e2 or not. This may be particularly useful for White if Black fianchettoes as White's queen may then be better placed at home, where it can bear down on the d6pawn. 7 . . . Ci:Jc6!

The negative side of White's previous move is seen in Black's clever reply. The move ... l2Jc6 rarely blends in well in a Kan system but this is an important exception. Normally in these lines White meets ...l2Jc6 with an exchange of knights, but the fact that Black has yet to move his d-pawn is a very positive feature for him. 7 ... �c5!? 8 l2Jb3 �e7 (8....ia7!?) 9 l2Jc3 d6 reaches positions similar to those in Chapter 4. 8 Ci:Jxc6

For other white possibilities, see Game 1 7. 8 . dxc6! . .

The point! Black breaks with tradition by capturing 'away from the centre' but the ensuing pawn structure holds little fear for the second player. Black will inevitably fol­ low up with ...e6-e5, leaving a symmetrical pawn structure with some attractive points for Black. The first is that he has a ready­ made outpost on the d4-square while White has no similar one himself. The second fac­ tor is that White's bishop on d3 is 'bad' and hemmed in by its own pawns. If only White

White does, however, have compensating features: he has a slight lead in development and (after the 1 0th move) a space advantage on the kingside. 9 f4!

The critical move. White threatens the space-gaining e4-e5 and forces Black's hand in the centre. 9 l2Jc3 e5! solves all of Black's problems, for example: a) 10 h3 .ic5 1 1 l2Ja4 �d4! 12 c5 .ie6 1 3 .id2 0-0 1 4 'ii'c2 h 6 1 5 l:tad1 l:tad8 1 6 l:tfe1 'ii'e 7 was equal in Maskova-Emms, Menchik Memorial, Maidstone 1 994. b) 10 .ie3 l2Jg4! 1 1 .id2 .ic5 12 l2Ja4 �d4! 1 3 c5 .ie6 1 4 'ii'c 2 'ii'e7 1 5 h3 l2Jf6 and Black was happy enough in Carlsson­ Salm, correspondence 1 984. The continua­ tion to this game is worth seeing: 1 6 'iti>h2?! h5! 17 g3? h4! 18 g4 .ixg4! 19 hxg4 l2Jxg4+ 20 'iti>g2 (20 'iti>g1 'ii f6!) 20 ... h3+ 21 'iti>h1 l2Jxf2+ 22 l:txf2 .ixf2 and Black went on to wtn. c) 10 'ii'e2 .ie6 (10 ... .ic5 1 1 �e3 .ixe3 12 'iixe3 0-0 is safe and equal) 11 f4 j_c5+ 12 'iti>h 1 .ig4! 1 3 'ii'c 2 (Griinfeld-Hulak, Za­ greb 1 987) and now the general opinion is that Black should play 13 .. l:td8! (instead of Hulak's 13 ... 0-0-0), for example: cl) 14 fxe5 l2Jd7! and ... l2Jxe5. 51

Sicilian Ka n

c2) 14 f5 'ii'd 6 1 5 i.e2 (or 1 5 ll'la4 i.a7 1 6 i.e2 i.xe2 17 'ii'xe2 'ii'd3!) 1 5 .. i.xe2 1 6 lixe2 lid3 and Black can b e happy. c3) 14 h3 ll'lh5!! 1 5 hxg4 lie7! (Zeller) and Black has a strong attack, for example 16 g5 ll'lg3+ 17 'it>h2 ll'lxfl + 18 i.xfl exf4 1 9 i.xf4? h6!. 9

. . .

e 5 10 f5

The most obvious and only testing move at White's disposal. White obtains a space advantage on the kingside and, for the mo­ ment at least, Black's light-squared bishop is sealed out of the game. 1 0 . . . ..tc5 + 1 1 �h 1

1 1 . . h5!? .

This is certainly Black's most ambitious move. For the moment Black does nothing to address the problem of his bishop on c8. Instead he plans an attack on the dark squares on the kingside with c5-bishop, knight and queen. A more restrained strategy can be seen with 1 1 ...i.d4 1 2 ll'lc3 h6 and now: a) 13 i.d2 b6 14 l:tb1 i.b7 1 5 b4 l:td8 1 6 lie2 0-0 was level i n Anagnostopoulos­ Bischoff, Bad Worishofen 1 993. b) 13 ll'le2 c5 1 4 l:tb1 i.d7 (14...b6!?) 1 5 ll'lxd4 cxd4 1 6 b4 and White was slighdy better in Pyhala-Bischoff, Thessaloniki Olympiad 1 988. Depending on where Black puts his king, White has chances to attack on either side of the board. It should be noted, though, that Bischoff was happy to 52

repeat this line as in note 'a' five years later. Maybe he has something up his sleeve. 1 2 lbd2

White's knight comes to help out on the kingside, where White is devoid of pieces. The main alternative is 12 ll'lc3 tt'lg4 1 3 li e ( 1 3 lie2 'ii'd8 1 4 .ic2? ll'lxh2!, Taylor­ Mortazavi, British League 1 998 is an obvi­ ous example of the tricks Black has in this line) 13. .. lid8! (a neat switchback; Black threatens ...lih4 and ... ll'lf2+) and now: a) 14 lig3!? .if2 1 5 'ii' e (or 1 5 lih3 .td4) 1 5 ... i.d4 16 lig3 i.f2 17 lie i.d4 18 'ii'g3 was a draw by repetition, Surak­ Edwards, correspondence 1 986. b) 1 4 ll'ld1 lih4 15 h3 b6 16 i.d2 .ib7 17 b4 i.d4 with a further split: b 1) 18 .ic3 0-0-0? 19 c5! was good for White in Agnos-Emms, St Albans 1 993, so Black should instead activate his b7-bishop with 1 8 ... c5!. b2) 18 l:tb1 c5! 19 bxc5 i.xc5 20 i.b4 i.d4 and here it's best for White to accept a draw by repetition after 21 i.c3 .ic5 22 .ib4 as 21 c5? bxc5 22 i.xc5 .ixc5 23 l:txb7 0-0 was good for Black in McDonald­ Emms, British Championship 1 993. Also possible is 12 i.g5 ll'lg4 13 lie2 lid6 14 .ih4 i.d7 1 5 h3 ll'lf6 1 6 ll'lc3 0-0-0 17 ll'la4 .id4, which was unclear in Casares Ripoli-Matt, correspondence 1 991 . Fe­ dorowicz suggests 12 ... h4, intending ... ll'lh5. 1 2 . . . lbg4 1 3 'i'e2

5 i. d3 l"i'J f6 6 0 - 0 � c 7 : S e v e n th Mo v e A lterna tives 13 . . . �d8

comfortably.

A major alternative for Black here is 13 ...'ii'b 6 and now: a) 14 g3?! lLlf2+ 1 5 Wg2 lLlxd3! 1 6 'ii'xd3 ..id4 1 7 l:lb 1 c5 1 8 a3 ..id7 1 9 b4 ..ia4! and Black was doing very well in the game Wolff-Gheorghiu, New York 1 987. b) 14 h3 lLlf2+ 15 Wh2 'ii'd B 1 6 lLlb3 (16 l2Jf3!?) 16 ...'ii'xd3 1 7 l:hf2 'ii'xe2 18 .l:txe2 ..ie7 was equal in Riemersma-Jukic, Graz 1 987 c) 14 lLlf3!? is a suggestion from Byrne and Mednis. After 1 4... f6?! 1 5 h3 lLlf2+ 1 6 Wh2 lLlg4+ 1 7 Wg3! Black i s forced back, while 14 ... l2Jf2+ 1 5 l:lxf2 ..ixf2 16 lLlxe5 gives White serious attacking chances for the small material investment.

1 5 .ic2 �b6 1 6 l"i'Jh4

Now after 16 h3 lLlf2+ (but not 16 ...l2Je3 1 7 ..ixe3 ..ixe3 1 8 b4!) 1 7 Wh2 l2Jg4+ 1 8 Wg3 lLle3! the difference i s that the knight forks rooks and bishop - Black has success­ fully lost a move to induce White's bishop to the unfavourable c2-square. Maybe White should consider 1 5 ..ib1!? . 1 6 . . . l"i'Jf2 + 1 7 l:.xf2 i.xf2 1 8 l"i'Jg6 l:.h7

1 4 l"i'Jf3! f6

Not a move Black really wants to play but the e5-pawn needs some protection. The alternative is 14 ...l2Jf2+ 1 5 l:lxf2 ..ixf2 16 lLlxe5, which again gives White tremen­ dous compensation for the exchange. Play continues with 16 ... ..ic5 17 ..if4 and now: a) 17 ... ..id6 1 8 c5! ..ixe5 (or 1 8 ... ..ixc5 1 9 ..ic4) 1 9 ..ixe5 0-0 20 ..id6 l:le8 21 'ii'xh5 is virrually winning for White. b) 17 ... 'ii'e7 1 8 l2Jf3 g6 with a further split: b 1) Estrada Nieto gives the line 1 9 b4 ..ixb4 20 'ii'b 2. Now it's true that 20... 0-0? 21 ..ih6 f6 22 ..ixf8 Wxf8 23 lLlh4 is win­ ning for White but, although I would be reluctant to take the black pieces, I can't quite find a killer blow for White after 20...l:lg8. b2) 1 9 fxg6 fxg6 20 e5 'ii' f7 21 ..ig5 ..ig4 22 ..if6 l:lg8 23 l:lft ..ie7 24 h3 ..ixf3 25 l:lxf3 0-0-0 26 c5 'ii'e 8 27 ..ixa6 ..ixf6 28 exf6 bxa6 29 'ii'x a6+ e8 1 9 liJd6+ 'it>d8 20 l:t£2 liJf5 21 liJxf5 (21 liJf?+? 'it>c7 22 lbxh8 lbxe3 23 'ilt'xe3 i.c5 24 'ilt'g3 l:txh8 is good for Black) 21 ... exf5 22 'ilt'xf5 with a very complex position - Black is very disor­ ganised but has the long-term advantage of that extra bishop. b) 12 ... i.b4 13 e5 i..xc3 14 'ilt'xc3 liJe4 1 5 'ilt'a3 (or 1 5 'ilt'b4 aS! 1 6 'ilt'a3 'ilt'xc4 1 7 l:!.act 'ilt'b4) 15 .. . f5 and Black can continue with . ..'it> f7. 1 2 . . ...tb4 13 tt:\4b3 i.xd2 14 tt:\xd2 1t'c6 1 5 .l:!.fe1 tt:\xe4 1 6 tt:\xe4 1fxe4 1 7 1fxe4 ..txe4 Y:z-Y:z

After 1 8 i..xb6 the position is completely equal.

Game 18 Haii-Hellsten

Umhamn 1998 1 e4 c5 2 tt:\t3 e6 3 d4 cxd4 4 tt:\xd4 a6 5 i.d3 tt:\t6 6 0-0 1fc7 7 'it>h 1

A very flexible move. White eliminates

5 .1t.. d3 lD f 6 6 0 - 0 'fi c 7 : S e v e n th Mo ve A lterna tives

annoying tactics by removing his king from the a7-gl diagonal and prepares £2-f4, while keeping his options open regarding other pieces and pawns. If there is a negative side, then it's that 'iit h 1 is hardly a developing move and Black, if he wishes, can strike immediately in the centre.

7 . ..d5!?

This takes play away from a 'typical' Kan and into independent lines. Normally Black would find it difficult to justify this idealistic freeing advance quite so early in the Kan, but his argument runs that White has effec­ tively 'lost' a tempo with 'iitg t-hl , so why not? Psychologically, though, many Kan players would be reluctant to change the flavour of the position, regardless of whether Black can achieve equality or not. If Black wishes to keep a hedgehog struc­ ture then he can play 7 ... d6 8 f4 4Jbd7 and now: a) 9 4Jc3 transposes to lines considered in Chapter 6. b) 9 c4 with a further branch: bl) 9 ...�e7 10 tDc3 0-0 1 1 'ir'e2 leads back to positions considered in Chapter 1 , where White has already committed his king to h l . b2) 9. . .g6?! is hit once again with 1 0 f5! e5 (or 10 ... gxf5 1 1 exfS e5 12 tt:Je6!) 1 1 tDc2 and White was better in Hi.ibner-Lutz, Bundesliga 1 993. If Black wishes to fi­ anchetto he should play 8 ... g6!, leaving his

e6-pawn defended by the c8-bishop. 8 lDc3

Another major line is 8 exd5 tt:Jxd5

and now: a) 9 �e4 4Jf6 10 �f3 �e7 is equal Hubner. b) 9 c4 4Jf6 10 tt:Jc3 �e7 1 1 'ii'e 2 0-0 1 2 i.g5 lld8 1 3 4Jb3 4Jbd7 1 4 llacl 4Jf8 1 5 llfel �d7 1 6 h 3 �c6 and Black was fine, Andersen-Hellsten, Gistrup 1 997. c) 9 f4 g6! (ECO gives 9 ...�c5 1 0 4Jb3 �d6 1 1 'ii' f3 4Jd7, Pietzsch-Gipslis, Riga 1 959, as equal) 1 0 �e4 4Jf6 1 1 �f3 i.g7 1 2 c 3 0-0 1 3 'Wet 4Jbd7 1 4 �e3 4Jb6 1 5 4Jd2 4Jbd5 and I prefer Black in Herbrechts­ meier-Raupp, Eppingen 1 988 - the kingside is rock-solid. d) 9 lle t �e7 10 c4 4Jf6 1 1 4Jc3 �d7 1 2 �g5 tDc6 1 3 4Jf5!? ( 1 3 �xf6 gxf6! and Black will castle queenside) 13 ...exf5 1 4 4Jd5 tDxd5 1 5 cxd5 0-0 1 6 �xe7 tDxe7 1 7 llxe7 'ir'd6 1 8 llel 'ii'xd5 1 9 .1xa6 .1c6 20 'ii'xd5 was agreed drawn in Zapata-Urday Caceres, Americana 1 997. Another idea for White is 8 tDd2 4Jbd7 (8... dxe4 9 tt:Jxe4 transposes to the text) 9 exd5 tt:Jxd5 10 c4 tD5f6 1 1 tt:Je4!? tDxe4 1 2 �xe4 4Jf6 13 �c2!? 'ir'xc4 ( 1 3. . .�c5 is the safe option) 14 �g5 �c5 1 5 4Jf3 and White has some compensation for the pawn, Sul­ skis-C.Horvath, Budapest 1 994. 8 . dxe4 9 liJxe4 liJbd7 1 0 b3 . .

10 c4!? transposes to Sulskis-C.Horvath 55

Sicilia n Kan

above. 10 ltJxf6+!? ltJxf6 reaches a similar posi­ tion to one from the French Defence (1 e4 e6 2 d4 d5 3 ltJd2 c5 4 exd5 11i'xd5 5 ltJgf3 cxd4 6 �c4 'Wd6 7 0-0 ltJf6 8 ltJb3 ltJc6 9 ltJbxd4 ltJxd4 10 ltJxd4 a6 1 1 �d3 11i'c7), except that White's king is on h1 rather than gl . This adds both positive and negative features to a line which is considered not to be critical. 1 1 l:te 1 and now: a) 1 1 ...�d6? is playable in the French line with the king on g1 but here 1 2 ltJf5! is much stronger than usual. b) 1 1 ...�d7 12 'Wf3!, intending �f4, is uncomfortable for Black. c) 1 1 ...i.c5! 12 ltJf5?! 0-0 13 ltJxg7 �xg7 1 4 'iWd2 is refuted by 1 4...ltJg8! 1 5 'i/Vg5+ �h8 1 6 �f4 �d6, while 1 2 ltJf3 simply leaves f2 hanging - a negative feature of �hl . Instead White should play 1 2 c3 and now 1 2...�d7 1 3 'iWf3 ( 1 3 �g5 0-0-0!?; 1 3 'ife2!?) 1 3...�xd4 1 4 cxd4 �c6 looks equal.

More ambitious is 15 h3!? and now: a) After 15 ... ltJg4? White can safely take the knight: 1 6 hxg4 hxg4+ 17 �g1 'iWh2+ 1 8 �fl 'ifh 1 + 1 9 �e2 'iWxg2 20 l:tg1 and Black doesn't have enough for the piece. b) 1 5 . . . 0-0-0 16 ltJf5!? exf5!? 17 l:txe7 ltJg4! 1 8 'iWg1 ! (18 hxg4 hxg4+ 19 �g1 ir'h2+ 20 �fl ir'd6! wins for Black - 21 l:txd7 l:thl + 22 �e2 ir'e6+) 18 ...ir'c5 1 9 l:te2 f4 20 l:tae 1 !, preparing 21 f3 ltJe3 22 �c l , and Black seems to be running out of tricks. Instead of 16 ... exf5, the line 1 6 ... i.f8! 17 �e5 ir'c5 1 8 ltJd4 �d6 is safe for Black. 1 5 . . . tt'lg4 1 6 g3

16 ltJf3?! is answered by 16 ... �c6!, for example 1 7 ltJe5? f6! 1 8 �g6+ �f8 and Black wins. 1 6 . . . 0-0-0

1 0 . . . �e7 ! ?

1 0. . .ltJxe4! 1 1 i.xe4 ltJf6 1 2 �f3 �d6 looks perfectly acceptable for Black. 1 1 �b2 tt'lxe4 1 2 �xe4 tt'lf6 1 3 i.d3 ..td7 1 4 l:!.e1

A very sharp pos!Uon has arisen, but counterplay on the kingside gives Black a full share of the chances. In the game a tac­ tical sequence soon forces an equal ending. 1 7 h3 �c5! 1 8 c3 e5 1 9 hxg4 hxg4 + 20 'it>g 1 exd4 21

cxd4 l:!.de8 22 'ifc2

l:!.xe 1 + 23 l:!.xe 1 ..td6 24 'ifxc7 + 'tixc7 25 d5 l:!.h5 26 .l:lc1 +

'tid8 27 .txg7

l:!.xd5 28 ..te4 l:!.d2 29 ..txb7 Y2 - Y2 1 4 . . .h 5 ! ?

An idea borrowed from the same French Defence variation quoted above. Black plans ...ltJg4, which is even stronger than usual due to the threat to the f2-pawn. 1 5 'ife2 56

Game 19 C .Koch-Fanghui Feng

Comspondence 2000 1 e4 c5 2 tt'lf3 e6 3 d4 cxd4 4 tt'lxd4 a6

5 JJ.. d3 t:D f6 6 0 - 0 � c 7 : S e v e n th Mo v e A lterna tives 5 JJ.. d 3 t:Df6 6 0-0 'iic 7 7 f4

A very aggressive move - White immedi­ ately goes for a kingside launch, but allows Black to pin the knight on d4 to his king. Here's a summary of other seventh move alternatives for White: a) 7 lLlc3 transposes to Chapter 6. b) 7 b3 d6 is likely to transpose to lines considered in Chapter 1, for example 8 i.b2 t:Llbd7 9 c4 g6 10 lLlc3 i.g7 1 1 'ii'e 2 0-0 and we have reached Short-Sax (Game 8). c) 7 lLld2 with the plan of f2-f4 and lLl2f3 is interesting, but shouldn't be threat­ ening: 7 ... d6 8 a4!? (It's too late to change your mind and fianchetto: 8 b3?? 'ii'c 3! is embarrassing. Also 8 f4 g6! 9 lD2f3 i.g7 1 0 'ii'e 1 lLlbd7 1 1 i.d2 0-0 1 2 'ifh4 eS 1 3 fxeS dxeS 1 4 lLlb3 lLlhS! was fine for Black in Fusthy-Hulak, Berlin 1 988) and now: cl) 8 ...g6 9 b3! i.g7 (9 ...'iWc3 doesn't work now: 10 .l:tb1 'ii'xd4 1 1 i.b2 regains the piece with a large advantage) 10 i.a3 0-0 1 1 lLlc4 l%.d8 1 2 i.b4 lLlc6 1 3 lLlxc6 'ii'xc6 14 i.aS l1e8 1 5 lLlb6 l1b8 was agreed drawn in Kuzmin-Vyzmanavin, Leningrad 1990 but without pointing to anything ob­ vious, I'm suspicious of Black's play here. c2) 8 ... b6 9 f4 i.b7 10 'ii'e 2 lLlbd7 1 1 lLl2f3 lLlcS! 1 2 eS dxeS 1 3 fxeS lLlxd3 1 4 cxd3 t:Lld5 1 5 Wh1 i.cS 1 6 lLlb3 i..e 7 1 7 aS 0-0 and Black was fine in Pablo Marin-Vehi Bach, Tarrassa 1 989. d) i.e3 dS!? (as with 7 Wh1 , Black can if he wishes aim for the French structure; the bishop is rather clumsily placed on e3 in this instance) 8 lLlc3 see following diagram

and now: d1) 8... e5?! (this wins a piece but the price is high) 9 lLlxdS lLlxdS 10 exdS exd4 11 i.xd4 i.e7! (1 1 ...f6 12 'iWhS+ Wd8 1 3 .l:tfe1 i..d 6 1 4 .l:te3 gave White a very strong attack, Hawelko-Mokry, Polanica Zdroj 1986) 12 'iWhS 'ii'f4 1 3 i.xg7 i.g4 14 'ii'e S 'i'xeS 15 i..x eS f6 16 i.. f4 when White's

three pawns are worth slightly more than Black's knight, Kasimdzhanov-Kagirov, Uzbekistan Championship 1 993.

d2) 8 ... dxe4! 9 lLlxe4 lLlbd7 1 0 lLlxf6+ lLlxf6 and again we transpose into a position normally reached via 1 e4 e6 2 d4 dS 3 t:Lld2 cS 4 exdS 'ii'x dS 5 t:Llgf3 cxd4 6 i.c4 'ii'd6 7 0-0 lLlf6 8 t:Llb3 lLlc6 9 lLlbxd4 lLlxd4 10 lLlxd4 a6 11 i.d3 'ii'c 7, except White has the extra move i.c1 -e3. It's debatable whether the bishop is much better here than on c l : 1 1 'ii' f3 i.d6 1 2 h3 0-0 13 .l:tad1 i.d7 14 .l:tfel .l:tae8 1 5 i.gS i.eS 1 6 i.. f1 lLldS 1 7 c 3 bS 1 8 'iWe4 i..h 2+ 1 9 W h 1 i.. f4! and Black had equalised in Ernst-Lau, Dort­ mund 1 992. 7 . . . JJ.. c 5

It seems only natural to pin the knight. 7 ... d6 is also possible when White can choose between 8 c4, with possible transpo­ sitions to Chapter 1 (although White has yet to commit his queen to e2), or 8 lLlc3 with a transposition to Chapter 6. 8 c3 t:Dc6

Inviting White to sacrifice a pawn. 8... d6! is the safest response: 9 �h1 eS (or 9 ...lLlc6 10 lLlb3 i..a7 1 1 c4 [Galliamova Ivanchuk-Ioseliani, Groningen 1 997] and now 1 1 ...0-0 is equal) 1 0 lLlb3 (or 10 fxeS dxeS 1 1 lLlf5 i..x fS 1 2 l:txf5 lLlbd7) 1 0...i..a7 1 1 'ii'f3 lLlbd7 1 2 c4 0-0 1 3 lLlc3 exf4 (Sax­ Dizdarevic, Sarajevo 1 985) and now Diz­ darevic gives 14 i..x f4 lLleS 1 5 i..x eS dxeS 57

Sicilia n Ka n

as being equal. 9 'it>h 1

9 eS!? is more forcing: a) 9 ...tt'lxd4 10 cxd4 i.xd4+ 1 1 Wh 1 tt'ldS 1 2 i.e4 - see the note to White's 1 1 th. b) 9 ... tt'ldS 10 Wh 1 ! (ECO gives 10 i.e4?! d6 1 1 i.xdS exdS 12 exd6 i.xd6 13 �el + tt'le7 1 4 1f f3 0-0 as good for Black, Sax­ Bellon Lopez, Graz 1 984, but I believe 1 0 Wh 1 i s stronger) 1 0...d6 ( 1 0... tt'lxd4 1 1 cxd4 i.xd4 transposes to the note to White's 1 1 th move) 1 1 tt'lxc6 bxc6 12 1We2 0-0 1 3 tt'ld2 fS 1 4 exf6 tt'lxf6 1 S tt'lf3 a S 1 6 i.d2 i.d7 1 7 �ae 1 was a bit better for White in Mokry-Vyzmanavin, Tilburg 1 994. The game continued 1 7 ...�ae8 1 8 i.c2 �e7 1 9 tt'lgS d S 20 tt'lf3! and the knight was ready to jump into the eS-outpost.

�xd1 + 24 l:txd1 was a level ending in Cam­ pora-Costa, Swiss League 1 99S, but I prefer the more ambitious 14 tt'lc3! d4 1 S tt'le4. 1 1 . . .0-0?

Stronger is 1 1 ... i.cS, although following 12 eS tt'ldS 1 3 tt'le4 i.e7 1 4 i.d2 bS 1 S a4 b4 1 6 aS i.b7 1 7 1fhS i.c6 1 8 �fc l , White has more than enough compensation, Lupu-Ionescu, Romanian Ch., 1 992. 1 2 e5 tLld5 1 3 lLlf3 tLle3

13 ... i.a7 allows a 'Greek Gift' sacrifice with 14 i.xh7+! Wxh7 1 S tt'lgS+ Wg6 (or 1 S ...Wg8 1 6 1WhS 1Wc2 1 7 fS!) 1 6 1Wg4 fS 1 7 1ih4 and White has a clear advantage. 1 4 .ixe3 .ixe3 1 5 i.xh7 + ! 'it>xh7 1 6 tLlg5 + 'it>gB 1 7 'ii'd 3 g6 1 8 'ii'x e3

9 . . . tLlxd4

Black still has a chance to play it safe with 9 ... d6, transposing to the note to Black's 8th move. 1 0 cxd4 .i.xd4

White has a strategically winning position due to the weakness of Black's king and the dark squares all over the board. 1 8 . . . 'it>g7 1 9 l:l.ac1 'ii'd B 20 l:l.fd1 b5 2 1 'ii'd 4 l:thB 2 2 'it> g 1 l:tbB 2 3 l:l.c5 l:l. h 5 24 'ii'd 3 'ii'h B 25 l:l.c7 'ii'd B 26 l:l.dc1 aS 27 h3 l:th4 28 'ii'g 3 l:th5 29 'ii'e 3 .l:!.aB 30 1 1 tLld2!

'ii'c 5 .:.hs 31 'ii'x b5 .:.h4 32 g3 l:l.h5 33

I believe this is the strongest move here, but 1 1 eS is also important due to transposi­ tional reasons: 1 1 ...tt'ldS 1 2 i.e4 i.cS (1 2 ... 1VcS!? is possible, but not 12 ...tt'le3? 1 3 1fxd4! tt'lxfl 1 4 1fd1 and the knight is trapped) 1 3 i.xdS exdS 14 1ixdS 0-0 1 S �d1 b S 1 6 tt'lc3 i.b7 1 7 1fxd7 1ib6 1 8 tt'ldS i.xdS 1 9 1ixdS �ad8 20 1f f3 �xd 1 + 2 1 1fxd1 l:td8 22 i.d2 i.b4 23 i.xb4

'ii'a4 .l:!.xg5 34 fxg5 'ii'x g5 35 'ii'f4 'ii'xf4

58

36 gxf4 .ta6 37 l:l.xd7 i.b5 38 l:l.b7 .teB 39 �f2 �h6 40 b3 a4 41 b4 l:l.dB 42 li;e3 a3 43 l:l.a7 li;hS 44 l:l.xa3 g5 45 fxg5 li;xg5 46 'it>e4 f5 + 4 7 exf6 li;xf6 48 l:l.f 1 + li;e7 49 .:.a7 + l:l.d7 50 l:l.xd7 + li;xd7 51 a4 li;c7 52 b5 li;b6 53 l:l.f6 li;a5 54 .l:be6 .i.d7 55 li;d5 li;xa4 56 b6 .tcB 57 l:l.e7 1 -0

5 i. d3 l'i:J f6 6 0 - 0 'ii c 7 : S e v e n th Mo ve A lterna tives Summary

Both 7 ...i.c5 and 7 ... i.d6 (Games 1 4- 1 5) are probably worth the occasional outing but I wouldn't advise employing them as 'stock' defences. 7 c4 has received some stick in the popular press and Black's results have been encouraging, but I believe that Games 1 6- 1 7 show that Black must still play accurately t o achieve a reasonable position. These lines arc less explored than those in Chapter 1 and independent srudy will pay dividends for both Black and White. 1 e4 c5 2 l'i:Jf3 e6 3 d4 cxd4 4 l'i:Jxd4 a6 5 i.d3 l'i:Jf6 6 0-0 'iic 7 7 'iie 2

7 c4 ltJc6 8 i.e3 - Game 1 7 8 ltJxc6 dxc6 (D) - Game 1 6 7 'it>h l - Game 18 7 f4 (D) - Game 19 7 . . . i.c5 - Game 14

7 ... i.d6 (D) - Game 15

8

. . .

dxc6

7 f4

7

. . .

i.d6

CHAPTER THREE

I

5 ..td3 tiJf6 6 0-0 d6

1 e4 c5 2 ltlf3 e6 3 d4 cxd4 4 ltlxd4 a6 5 ..td3 ltlf6 6 0-0 d6

In this chapter we shall be studying Black's main sixth move alternative, which is 6 ... d6. As well as this, we will take a brief look at other sixth moves for both White and Black. Let's go through the opening moves agam. 1 e4 c5 2 ltlf3 e6 3 d4 cxd4 4 ltlxd4 a6 5 ..td3 ltlf6 6 0-0

It's quite easy to skip past other sixth move options for White, concentrating solely on 6 0-0 (even the usually compre­ hensive ECO is guilty of this). There are, however, some tricky alternatives which, by using clever move orders, attempt to steer the game into lines Black may be wishing to avoid. a) For 6 c4, sec 5 c4 tDf6 6 ..td3 in Chap­ ter 9. b) 6 tDc3 is an important transpositional move. 6 .. .'ific7 would then transpose into lines discussed in Chapter 6. If Black wishes instead to steer the game into variations discussed in Chapter 7, he should try 6 .. .'ilib6 although I have to say that there has been very little practical experience with this move order (those playing 5 ..td3 nor­ mally wish to leave the option open of mov60

ing the c-pawn) . The other line to consider is 6 tbc3 d6, which is closely linked to the Sicilian Scheveningen and indeed can be reached by this move order (1 e4 c5 2 tbf3 d6 3 d4 cxd4 4 tbxd4 tiJf6 5 tbc3 e6 6 ..td3!? a6). After 7 0-0 the move 7 . . .b5!? keeps a Kan flavour and is studied in Chap­ ter 8 under the move order 5 tDc3 b5 6 ..td3 d6 7 0-0 lDf6. c) 6 f4!? is another tricky move (I can only find eleven examples of this in my database of over 22,000 Kan games): 6 ... ..tc5!? (6 ... d6 is normal; after 7 0-0 we transpose to the note to White's 7th move) 7 tDb3 ..ta7 8 tDc3 d6 9 'ilif3 tbc6 1 0 g4 tbd7 1 1 g5 b5 1 2 h4 ..tb7 led to something unusual in the game E.Berg-Hellsten, Ron­ neby 1 998. In this type of position Black's bishop is normally on e7 (or perhaps, with ... g7-g6, g7) but not on a7. Nevertheless, with Black probably castling quccnside, this also looks perfectly satisfactory for him. d) 6 'ilie2!? virtually forces 6 ... d6 due to the threat of e4-e5. A fter 7 0-0 we reach normal lines except that the price of forcing Black into an early ... d7-d6 is that White has committed his queen to c2. Independent lines can be reached if White castles on the queenside (this is very rare in the Kan) : 7 f4 g6 8 tbf3 ..tg7 9 tbc3 'i!ic7 10 ..td2!? 0-0 1 1

5 i. d3 CiJ f6 6 0 - 0 d6

0-0-0 lL'lc6 1 2 l:the1 b5 1 3 e5 dxe5 14 fxe5 lL'ld7 15 �f4 �b7 16 '1t>b1 lL'lc5 and Black looks to be in perfectly good shape here, Rowson-Spraggett, Edmonton 2000. 6 . . . d6

Apart from 6 ... 1Vc7 (see Chapters 1 and 2) and 6 ... d6 (the main subject of this chap­ ter) Black doesn't really have too much choice: a) 6 ... d5?! is dubious. After 7 e5 the most narural retreat loses: 7 ... lL'lfd7? 8 lL'lxe6 fxc6 9 1Vh5+ g6 10 i.xg6+!. b) 6...�c5 7 lL'lb3 �a7, trying to trans­ pose into ...�c5 lines where White is com­ mitted to castling kingside, suffers from the fact that it allows White to hit the f6-knight: 8 e5 lL'ld5 and now 9 1Vg4 emphasises Black's bare-looking kingside. It's very rare that Black can get away with allowing White an early knight-hitting e4-e5. c) 6 ... e5?! is a very cheeky move; Black decides to play a Najdorf-srylc position a whole tempo down! However, it has been played a few times, notably by Romanishin. 7 lL'lb3 d6 8 c4 �c7 9 lL'lc3 0-0 (Computer P ConNers-Romanishin, Lippstadt 1 999) and now, instead of the machine's choice of 1 0 lL'ld5, I prefer one of 1 0 f3, 1 0 �e3 o r 1 0 i.e2, all of which give White some edge.

7 . . . i.d7

This move, with the idea of following up with ... lL'lc6, has become quite popular in recent years, having been taken up by Kan specialists Rublcvsky and Milov. Thus four games in this chapter arc devoted to 7...�d7. The fianchetto with 7 ...g6 is srudicd in Game 24, while the classical 7 ... �e7 is stud­ ied in Games 25 and 26. Other seventh move alternatives for Black are studied in the notes to Game 25. 8 CiJc3 CiJc6 9 CiJxc6

By far the most popular move. White's only real alternative, 9 �e3, is discussed in Game 23. 9

. . .

.ixc6

7 c4

For 7 f4 sec Game 27. After 7 lL'lc3, 7 ...b5 transposes to a position reached in Chapter 8 while 7 ... 1Vc7 leads to Chapter 6.

This is the starting position for Games 20-22.

Game 20 Ulibin-Milov

Frankfurt (rapid) 1999 1 e4 c5 2 CiJf3 e6 3 d4 cxd4 4 CiJxd4 a6 5 .id3 CiJf6 6 0-0 d6 7 c4 .id7 8 CiJc3 CiJc6 9 CiJxc6 i.xc6 1 0 b3

Another move order employed to reach the same position at move 12 is 10 1Ve2 �e7 1 1 b3 0-0 1 2 �b2. 1 0 . . . .ie7 1 1 .ib2 0-0 1 2 'iie 2

1 2 f4?! is a bit too early here and allows Black a rypical break: 1 2. . . b5! (exploiting a 61

Sicilia n Ka n

gain in time associated with ... .id7 / . ..Ci'Jc6 variations; Black has managed to play ... b7b5 in one go, having not spent a tempo playing ... b6 earlier): a) 1 3 cxbS?! axbS 1 4 'iff3 (or 14 eS dxeS 15 fxeS 'ifd4+!) 14 ...'ifb6+ 15 �ht 'il'b7 and Black has strong pressure against e4, Konguvel-Kostenko, Calcutta 200 1 . b) 1 3 'ife2 bxc4 1 4 .ixc4 d S 1 5 .id3 dxe4 16 tt::lxe4 tt::lxe4 17 .ixe4 .ixe4 1 8 'il'xe4 'il'd2 was agreed drawn i n Sanchez Romero-Sospedra Sebastian, Aragon 1 996.

1 2 . . . ii'a5!

Revealing another pos10ve feature of forgoing ...b7-b6: the queen has access to the aS-square. I believe it was the Bosnian GM Emir Dizdarevic who first played this move and more recently it has found favour with both Rublevsky and Milov. One of Black's main plans here is to swing the queen across to hS to offer a trade. In gen­ eral, Black's position becomes easier after an exchange of queens as he doesn't have to worry about a direct kingside assault by White. Alternatives include: a) 12 ... l:.e8 13 f4 dS!? (13. ..tt::ld7 is safer) 1 4 cxdS exdS 1 5 eS tt::le4 1 6 tt::lxe4! dxe4 1 7 .ixe4 .ibS 1 8 'ifg4?! 'ifd2! 1 9 .ixh7+ when Black can take a draw with 1 9 ... �xh7 20 e6 'ifxb2 21 'ifhS+ �g8 22 'ifxf7+ �h7 23 'ifhS+ or play on with 19 ...�f8!? as in Ar­ nason-Toshkov, Jurmala 1 987. As irn62

provements over 1 8 'ifg4?! , Nunn gives 1 8 'if f3 while I prefer to flick i n 1 8 'ifhS! g6 1 9 'iff3. Both seem to give White good com­ pensation for the small material deficit. b) 12 ... 'ifb8 13 a4 .:te8 14 .:tact tt::ld7 1 5 f4 .ih4 1 6 .:tdt .if6 was unclear in Nunn­ Bischoff, Dortmund 1 987. Nunn suggests that the more direct 13 f4 is stronger: 1 3 ...b5?! 14 cxbS axbS 1 5 eS is indeed good for White. 1 3 l:tad 1

O r 13 f4 'ifhS and now: a) 14 .:tf3!? with a further split: at) 14 ... d5?! 15 cxdS .icS+ 1 6 �ht exdS 17 tt::lx dS tt::lx dS 18 exdS .:tfe8? 19 'ifc2 .ixdS 20 .:tg3 i.f8? 21 .:tgS was a quick win for White in the game Tairi-Astrom, Swe­ den League 2000, but Black was far too adventurous too soon in the centre. a2) One idea I like here is 14 ... i.d8!, planning to redeploy the bishop more ac­ tively on b6 (yet another hidden advantage of forgoing ...b7-b6 in this line!). However, Black must be careful to avoid tactics against his king: 1 5 �hl ! and now the im­ mediate 1 5 ... .ib6? lands Black in big trou­ ble:

a21) 16 tt::la4? tt::lg4! 1 7 .:th3 tt::l f2 + 1 8 'ifxf2 'ifxh3 1 9 tt::lxb6 'ifxd3 20 tt::lxa8 .:txa8 is fine for Black. a22) 16 tt::ld S! exdS 17 .ixf6 dxe4 18 l:.g3 and the best Black can do is 1 8...'il'g6 1 9 lhg6 hxg6 20 .ixe4 gxf6, which is pretty

5 i.. d3 l?J f6 6 0 - 0 d6

gnm. Going back to Black's 1 5th move, much stronger is the prophylactic 1 5 ...l:te8!, plan­ ning to answer 1 6 l:taft with 16 ...d5!. I be­ lieve Black has excellent counterplay in this line. b) 1 4 'iWxhS ltJxhS 1 5 i.e2 1 5 ...tt"Jf6 1 6 i.f3 l:tfd8 1 7 l:tadl ltJd7 1 8 i.a3 ltJcS 1 9 l:.d2 b 6 20 'iti> £2 'iti>f8 2 1 l:tfdt 'iti>e8 and Black has equalised comfortably. In the game Egger-Milov, Istanbul Olympiad 2000, White now blundered a pawn with 22 eS?! i.xf3 23 gxf3? dxeS 24 fxeS l:txd2+ 25 l:txd2 ltJxb3!. Black always has to watch out for ltJdS ideas but here it doesn't work: 13 ltJdS? exdS 14 exdS ltJxdS 1 5 cxdS 'iWxdS 16 f4 l:tfe8 and Black was just a pawn up in Pe­ runovic-Cabrilo, Herceg Novi 200 1 . 13

. . .

'ii'h 5

17 'iW£2 ltJd7 1 8 g4 'ifh3 19 'ifg3 'i!Vxg3+ 20 hxg3 t2Jb6 21 i.e2 i.e7 22 l:td4 l:ted8 23 l:tfd1 i.e8 24 f4 i.f6 25 l:txd6 l:txd6 26 l:txd6 ltJxc4 27 i.xc4 l:txc4 28 eS i.e7 and Black slowly converted his two bishops advantage in Zakic-Dizdarevic, Pula 1 990. b) 16 'iWd2 and now: b1) 16 ... b5!? 17 cxbS axbS 18 f4?! i.d8 19 l:tf3 i.b6+ 20 'iti>h1 dS 21 exdS exdS 22 i.xbS i.xbS 23 ltJxbS tt"Je4 24 'i!Vd3 ltJ£2+ 25 l:tx£2 i.x£2 worked out well for Black in Dobrovolsky-Toshkov, Primorsko 1 987, but I prefer 18 f3!, blunting Black's light­ squared bishop. Black must be careful not to be simply left with a weakness on bS. b2) 16 ...l:ted8 17 l:tfet i.e8 18 l:te3 'iWgS 1 9 ltJe2 and the players opted for an 'early bath' in the game Psakhis-Rublevsky, Po­ lanica Zdroj 1 997. It's true that not much is going on in the final position.

Breaking on the queenside with 1 3 ... b5?! now runs into 14 ltJdS!. It is possible, how­ ever, to delay the offer of the queen exchange: 13 ...l:tfe8!? 1 4 a3 l:tac8 1 5 b4 and only now t S ...'iWhS!.

1 4 f3

It does seem worthwhile for Black to wait if White is simply going to force the issue with a2-a3 and b3-b4, so maybe White should contemplate something different on move 14 and 1 5. Anyway, after 1 S ...'iWhS White has played: a) 1 6 f3 (White gains nothing by avoiding a queen exchange in this manner) 1 6 ...i.d8!

As I stated before, I don't believe that White gains much from avoiding the queen exchange, although Black has nothing to fear there either: 14 'iWxhS ltJxhS 1 5 i.a3 l:tfd8 and now: a) 1 6 i.bl tt"Jf4! 1 7 g3 ('/z-1/z Almasi­ Milov, Groningen 1 998) 17 ...ltJg6 1 8 f4 bS 19 l:td2 l:.ab8 20 cxbS axbS 21 i.b4 dS and Black has equalised. b) 1 6 eS!? 'iti>f8 (16 ...ltJf4? 17 i.e4! is good for White) 17 f4 (1 7 i.e2?! is met by 17 ...ltJf4!, intending 1 8 exd6 i.f6!) 17 ... dxe5 63

Sicilian Kan

1 8 .ixe7 + c4 46 l:!.f8 b3 47 .l:tb8 Wc3 48 l1c8 + 'it>d3 49 .l:td8 + Wc2 50 .l:tc8 + 'iti>d 1 5 1 l:ld8 + 'it>c1 52 .l:tc8 + l1c2 53 l:lh8 l:lc5 0-1 64

A different approach. White immediately makes his mark on the half-open d-file. 1 1 . . .tt:ld7

Blocking the d-ftle and thus ruling out e4-e5 ideas. Alternatively: a) 1 1 ...0-0?! 1 2 e5! is problematic for Black, as 1 2... dxe5 allows 1 3 .ixh7+ . b) 1 1 ...'ilf'a5 1 2 .if4 0-0 1 3 a 3 and now: b 1) 13 ... .l:.fe8?! 14 h3! (preventing a later ...lt::lg4) 14 ...ifh5 1 5 ife3 ltJd7? (Dolmatov gives 1 5 ...'ilf'c5 as an improvement but White is still clearly better after 1 6 'ilf'd2) 1 6 .ie2 ifc 5 1 7 ifg3 'lib6 1 8 b4 lt::lf6 1 9 .ixd6 .ixe4 20 .ie5 20 ... .ig6 21 c5 'lia7 22 'lie3 and White's pieces dominate the board, Dolmatov-S.Kovacevic, Ubeda 2000. b2) 1 3 ...'ilf'h5 14 'lixh5 lt::lx h5 1 5 .ie3 lt::l f6 16 f3 lt::ld7 17 .ifl .l:.fd8 1 8 b4 b6 1 9 .l:h2 h 1 'ii'b 6

The earlier game Lastin-Landa, Tomsk 2001 continued 1 8... 'ifcS!?, trying to induce White into playing b2-b3 and thus killing the a2-bishop. Instead White initiated tre­ mendous complications with 19 f4!? lt:lxc4!? 20 b3 lt:le3 (20 ... lt:lb6 21 liJdS 'ifxcl 22 lt:lxe7+ f8 23 lt:lxc8 'ifxc8 24 l:.xd6 looks better for White) 21 l:td3 dS 22 b4 'ifxb4 23 'ifxe3 dxe4 24 l:tdd1 'ifa3 with a posicion which is very difficult to assess. White has an extra knight but some coordination problems. Black has three extra pawns but no dangerous pawn majority. One thing that can be said is that Lastin was prepared to face 1 8 ... 'ii'c S again, albeit in a blitz game! 1 9 f4! t:Dc6 20 f5!?

This is a very dangerous plan. White in­ tends to give the bishop on a2 his life back! 20 . . . t:Db4 21 fxe6 fxe6

Given what happens in the game, Black should seriously consider 21 ... lt:lxa2!? here. After 22 exfl + i.xfl 23 lt:lxa2 'ii'b3 Black regains his pawn and is only a little worse. 22 c5!

This pawn sacrifice releases the a2bishop and now both White's bishops angle beautifully across the board. 65

Sicilian K a n ltJc6 9 ltJxc6 �xc6 1 0 b4!?

2 2 . . . dxc5 2 3 a5 'ii'c 6 2 4 ..tc4 �dB 25 �f1 �f7?

2S ... i.d7 (Fedorowicz) is a stronger de­ fence but I would still very much prefer to be White: 26 1i'g4 l:t£8 27 i.eS g6 28 h4!, planning h4-hS, looks good. 26 'ii'g 4 l:UB 27 �e5 g6 28 'ii'f4 g5 29 'ii'g4 h5 30 'ii'h 3 �adS

There are three points to this ambitious move. Firstly, Black now has to be on a constant lookout for b4-bS from White, especially as the knight on f6 often ends up on d7. Secondly, Black is denied access to the aS-square which, as we have seen in previous games, is a handy place for the queen. Thirdly, White prepares to fianchetto with i.b2. So far so good for White, but there is a downside. By moving the b-pawn two squares instead of one, the c4-pawn is denied protection from its partner and can prove to be a liability on the half-open c­ f!.le. Whether Black can become active enough to exploit this is another question. 1 0 . . . b6

31 �xf7! �xf7 32 �xe6 ltJd3 33 �xf7 + �xf7 34 .l:l.f 1 + �eB 35 'ii'x h5 + �d7 36 'ii'h 3 + 'ii'e 6 37 'ii'x d3 + �cB 38 'ii'g 3 'ii'c4 39 'ii'g 4 + .U.d7 40 �d 1 ..td6 4 1 �xd6 1 -0

Game 22 Khalifman-Rublevsky

Kazan 200 1

._______________.

1 e4 c5 2 ltJf3 e6 3 d4 cxd4 4 ltJxd4 a6 5 �d3 ltJf6 6 0-0 d6 7 c4 �d7 8 ltJc3 66

The safest reply; Black gives his bishop a secure retreat to b 7. However, Black can also simply develop with 10 ... i.e7 and now: a) 1 1 bS!? (critical, but White has yet to come up with an advantage after this move) 1 1 ...i.d7 1 2 a4 (12 i.e3 axbS 13 cxbS "iVaS 14 i.d2 1i'a3 1 S l:tb 1 0-0 1 6 l:tb3 "iVa7 1 7 b6 1i'b8 18 i.bS l:tc8 19 a4 i.d8 20 i.f4 i.xb6 21 i.xd6 i.c7 22 i.xd7 l:td8! 23 tiJbS was agreed drawn in Savanovic-Ionescu, Bijeljina Dvorovi 2002; 23 ... i.xd6 24 1i'xd6 l:txd7 is very drawish) 1 2...0-0 1 3 i.e3 1i'c7 14 aS axbS 1 S i.b6 "iVc8 1 6 tDxbS i.c6 1 7 tt:Ja7 1i'd7 1 8 tDxc6 1i'xc6 1 9 1i'a4 1i'c8 20 "iVbS tiJd7 21 l:tfb1 i.f6 22 l:ta3 tt:Jxb6 23 1i'xb6 i.d8 24 1i'xb7 was another draw in

5 i. d3 Ci:J f6 6 0 - 0 d 6

Lobzhanidze-Lastin, Kstovo 1 997. b) I can't find any examples of 1 1 i.b2, but after 1 1 ...0-0 12 1i'e2 it's not clear whether Black can achieve anything by de­ laying ... b7-b6 any further. 1 1 i.b2 i.e7

This is not forced. Black could try to take the sting out of a possible kingside attack from White by fianchettoing, although of course there are then potential problems with the defence of d6. Nevertheless, 1 1 ...g6!? 12 1i'e2 i.g7 13 l:.ad1 1i'e7 (13...0-0 14 eS! dxeS 1 5 i.xg6 is promising for White, as is 13 ...1i'c7? 1 4 lLJdS!) 1 4 bS axbS 1 5 cxbS i.b7 16 1i'e3 1i'd8 17 i.e2 0-0 18 i.f3 .U.e8 19 eS lLJdS 20 lLJxdS exdS 21 i.xdS i.xdS 22 .U.xdS i.xeS saw Black equalising in Cheparinov-S.Kovacevic, Dos Hermanas 2002.

1 3 . . . Ci:Jd7

In later games Black instead played 1 3 ...'iic 7 14 f4 and now: a) 14 ... l:.ad8?! (I believe the rook should instead be on c8, hitting c4) 1 5 �hl i.b7 16 l:tf3! g6 17 l:th3 lLJd7? 18 lLJbS!! and Black resigned in Kindermann-Thesing, Bundesliga 2001 because of 1 8 ... axb5 1 9 'iVhS!. This i s a great illustration of the dan­ gers that await Black in this line. b) 14 ...l:tac8 1 5 .U.f3 .U.fd8 16 l:th3 h6 1 7 .U.g3 i.b7 1 8 a3 a S 1 9 eS dxeS 20 fxe5 lLJe8 21 lLJbS 1i'b8 (Luther-Rublevsky, World Team Ch., Yerevan 2001) and here Rublev­ sky gives 22 i.cl axb4 23 axb4 i.xb4 24 i.xh6 with a clear advantage for White. Rublevsky mentions no improvements for Black in this line so we should assume that White is better after 13 ...'iic 7.

1 2 'i!fe2 0-0 1 3 l:lad 1

14 f4 i.b7

It should be pointed out that this line is very similar to the one studied in Vogt­ Gheorghiu (Game 25), but with two differ­ ences: White has not committed himself to playing �hl , and White's b-pawn is on b4 rather than b3. The first difference is obvi­ ously favourable for White as he may well be able to do without 'it>h1 , using the extra tempo in his kingside attacking ambitions. The second difference, though, favours Black; the c4-pawn can become quite vul­ nerable down the half-open c-ftle, especially if White plays the 'normal' retreat i.b 1 .

Preparing to pressurise the c-pawn. The line 1 4... l:.c8 1 5 .U.f3 i.b7 1 6 l:th3 would transpose to the note to Black's 1 6 th move, while 14 ... i.f6? gives Black problems with d6: 1 5 i.c2 1i'c7 1 6 'iVd3 (Stohl) and Black is faced with threats of 'iix d6 and e4-e5. 1 5 l:lf3

Here comes a typical rook swinger! White is going straight for the throat on the kingside. 15

. . .

g6

A prophylactic measure against l:th3 fol­ lowed by c4-e5. At some point Black will 67

Sicilian Kan

also want to bolster his kingside with ... i.f6g7 or ... l:.eB and ... i.f8. 16 :h3

b) 21 ...i.xe5? 22 i.xe5 l2Jxe5 23 lLlf6+. 22 ll'lf6 + i.xf6 2 3 exf6

Now White has mating threats with 'ii'c 3h6. Black does extremely well to last as long as he does. 23 . . . e5 24 l:r.f3 l::.fd8 25 .l:f.d5!

But not 25 'ii'e 3? on 25 ...'ii'xc4 26 'ii'h6 lLlxf6!.

account

of

25 . . . h5 26 l:.fd3 ll'lf8

26 ... lLlxf6? 27 'ifxe5! 'ifxeS 2B l:.xdB+ 'ii'e B 29 i.xf6 and 26 ... 'ii'xc4? 27 'ii'd2 both win for White. 27 'ikxe5 'ikxe5 28 i.xe5 :xd5 29 .l:f.xd5

1 6 . . . i.f6?

This move allows a tactical breakthrough, after which White converts his initiative into a concrete advantage. The only consistent way to play is with 1 6 ... l:.cB! and now, as Slovakian GM Igor Stohl has shown, no combinations work for White: a) 17 liJdS? (this works against 17 .. .'ific7 but not when the queen is on dB) 17 ... exd5 1 B 'WhS gxhS 19 l:.g3+ i.gS!. b) 1 7 eS?! dxeS 1 B i.e4 'Wc7 and White has no good way to continue. c) Rublevsky gives 1 7 i.b1 with a 'clear advantage to White' but it's not totally ob­ vious how White strengthens his position, while Black can improve his defences with ... l:teB and possibly ...i.f8. 1 7 e5!

Okay, Black has survived into an end­ game, but the extra pawn is always going to be a winner. 29 . . . ll'le6 30 c5 bxc5 3 1 bxc5 l:.c8 32 �d6 g5 33 'it>f2 'ith7 34 �e7 'it>g6 35 'it>e3 l::. b8 36 l:!.d2 'itf5 37 'it>d3 l:!.b4 38 'it>c3 :b1 39 'it>c4 .l:f.c1 + 40 'itd5 h4 41 c6 g4 42 'it>d6 g3 43 hxg3 hxg3 44 'it>d7 l::.c 3 45 �d6 'it>xf6 46 c7 1 -0

Both players should always be looking out for this move. 1 7 . . . dxe5 1 8 i.e4 'ikc7

Game 23 Shirov-Rublevsky

Or 1B ... i.xe4 19 lLlxe4 i.g7 20 l:.hd3 l:.a7 21 i.xeS 'WeB 22 i.xg7 �xg7 23 l:.d6 and White's pressure is becoming unbear­ able.

5 i.d3 ll'lf6 6 0-0 d6 7 c4 �d7 8 ll'lc3

1 9 i.xb7 'ikxb7 20 ll'le4 �g7 21 fxe5

ll'lc6 9 �e3

'ikc7

The only real alternative to 9 lLlxc6. White defends the knight on d4 and devel­ ops another piece.

Or: a) 21 ...lLlxe5? 22 liJd6!. 68

Montecatini Terme 2000 1 e4 c5 2 lt:'lf3 e6 3 d4 cxd4 4 ll'lxd4 a6

5 i. d3 0. t 6 6 0 - 0 d6 1 0 f4

10 nct is less threatening: 10 ...0-0 1 1 "ii'e 2 lDe5 1 2 i..b 1 "ifc7 1 3 b3 lDeg4 1 4 i..g5 "ii'c 5 1 5 l2Jf3 h6 and Black was fine in the game Balashov-Liberzon, Palma de Mal­ lorca 1 989. 1 0 . . . 0-0

9

. . .

i.e7

An important alternative here is 9 ... l2Je5!? 10 i..e 2 and now: a) 10 ... i..e7 is sensible: 1 1 f4 lDg6 1 2 'it>h1 0-0 1 3 "ifd2 e5! 1 4 fxe5 (or 1 4 lDf5 exf4! 1 5 i..x f4 i..xf5 1 6 exf5 l2Jxf4 1 7 1:hf4 d5! 1 8 cxd5 i..d 6 1 9 .I:tffl .te5! followed by ..."ifd6, ...nfe8 and ... nac8 [Fogarasi] when Black has good dark-squared compensation for the pawn) 1 4...dxe5 1 5 ltJfS i..b4! and Black's counterplay against e4 compensates for White's powerful knight on f5, Acs­ Fogarasi, Budapest 1 998. b) 1 0.. J:tc8!? and now: b1) 1 1 f4!? lDxc4 1 2 i..xc4 nxc4 1 3 e5 is an interesting pawn sacrifice. In the game Renet-Adamski, Katowice 1 992, White got a strong attack after 1 3 ...l2Jg8 1 4 "ii'e 2 d5 1 5

Black can also go for an immediate counter-attack with 10 ...l2Jxd4 1 1 i..xd4 e5!? 12 fxe5 dxe5 13 i..xe5 and now: a) 13 ...l2Jg4 (Hawelko-Adamski, Polish Ch. 1 987) and now 14 i..d 4! looks good: 14 ..."ii'c 7? 1 5 e5! i..c 5 16 i.. f5!. b) 13. ..i..c 5+ 1 4 'it>h 1 l2Jg4 15 i..g3 lDe3 1 6 "ii'h 5 lDxfl 1 7 nxfl i..e 6 1 8 l2Jd5 i..d6 19 i..h 4!? (Hawelko) and now after 1 9 ..."ii'b 8 I like the move 20 e5!. White cer­ tainly has a strong attack. 1 1 �h 1 0.xd4 1 2 i.xd4 i.c6 1 3 'ife2 0.d7

fS.

b2) 1 1 b3 with a further split: b21) 1 1 ...b5 1 2 f4 l2Jg6 1 3 lDc2 i..e7 14 a4 with an edge to White, Shorr-Hjartarson, Amsterdam 1 99 1 . b22) 1 1 ..."ii'a5 1 2 nct ( 1 2 i..d2 "ifc5 1 3 i..e3 "ii'a 5 1 4 i..d2 "ii'c 5 1 5 lD£3 i.. e7 1 6 nc1 0-0 17 i..e3 "ii'a5 was equal in Koch­ Cramling, Haifa 1 989) 1 2... -te7 13 f4 lDc6 14 i.. f3 0-0 1 5 "ii'd2 lDxd4 1 6 i..xd4 i..c 6 1 7 nfd1 'it>h8 1 8 "ii' £2 b 5 1 9 e 5 dxe5 20 fxe5 l2Jd7 21 cxb5 axb5 and Black had equalised in Tondivar-Adamski, Giessen 1 994, even though it seems a little strange to carry out the manoeuvre ...lDc6-e5-c6!

Black has reached his typical defensive set-up. Compared to previous games in this chapter, White has gained a bit of time be­ cause the exchange of knights occurred on d4 instead of c6. On the other hand, the bishop on d4 is not ideally placed; it can be hit by a timely ... e6-e5. 1 4 l:!.ad 1 e5!

Black gives up the d5-square but secures e5. This is a reasonable trade, especially since the bishop on c6 is ready to chop on d5 as soon as the white knight arrives. 1 5 .ie3 exf4 1 6 .ixf4 0.e5 69

Sicilia n Kan

1 6 ...'ifa5 1 7 lDd5 .i.xd5 1 8 exd5 l:Iae8 19 'ifc2 h6 20 ..tf5 lDf6 21 a3?! lDh5 22 b4 'ifd8! 23 .tel i.g5 24 .i.b2 .i.f6 equalised for Black in Hernandez-Christiansen, Phila­ delphia 1 998, but I like White after the more ambitious 21 c5!. 1 7 .tc2

Lining up an enticing sacrificial idea. Also worth attention is Ribli's suggestion of 17 l2Jd5!? .i.xd5 1 8 .i.xe5! (18 exd5 .i.f6 is very solid for Black) and now: a) 1 8 ... .i.e6 19 .i.c3 is favourable for White, who can attack with e4-e5. b) 18 ... dxe5 19 exd5 looks pleasant for White, although after 1 9 ...'ifd6 (19... ..td6 20 'ir'c2 g6 21 c5 l:tc8 22 b4 is tremendous for White) 20 l:tdel f6 21 ..tf5 g6 22 i.e6+ �g7 it's also difficult for White to improve his position.

lDxa8 lha8 24 l:Id 1 t2Jd3! 25 .i.xd3 .i.xf4 26 .i.xb5 - Shirov. 20 . . . .tb5 2 1 'ifh5 f6 22 l:tf4

22 lDxf6+?! l:txf6 23 .i.xe5 llxfl + 24 l:txfl 'ir'e6 25 i.b3 .i.c4 is a bit better for Black; White's attack has run out of steam and he is left with more pawn islands. 22 . . . g6 23 'ifh6

1 7 . . . 'ifc7 1 8 c5!? 23 . . . l:tad8!

Imaginative play by Rublevsky, who cal­ culates that he will obtain enough compen­ sation for the queen. 24 lLlxf6 + l:txf6 25 l:txd6 l:tfxd6 26 l:tf1 ! i.f8?

Sacrificing a pawn in order to undermine the support for the e5-knight and open lines for the light-squared bishop. 1 8 .i.xe5 dxe5 19 t2Jd5 .i.xd5 20 exd5 g6 gives White an inferior version of note 'b' to White's 1 7th move.

Undoing the previous good work. Shirov suggests 26 ...g5! and now: a) 27 'ir'h5? .i.xfl 28 .i.xe5 l:tg6 29 i.c3 b5 30 e5 .i.d3 31 i.xd3 llxd3 and the ex­ change of bishops leaves the position in Black's favour. b) 27 'ir'h3 .i.xfl 28 .i.xe5, which Shirov assesses as unclear. 27 'ifc1 .ixf1 28 .ixe5 l:te6?

28 ... ..tc4 limits the damage. 29 'ifxf1 J:txe5 30 .ib3 + c4 3 1 'ifxc4 + 'it>g7 32 'iic 7 + 1 -0

1 8 . . . dxc5 1 9 lLld5 'iid 6

19 ... .i.xd5? loses after 20 exd5 .i.d6 (or 20... f6 21 'ir'e4 g6 22 d6 .i.xd6 23 'ir'd5+) 21 i.xe5 .i.xe5 22 d6! .i.xd6 23 'ir'd3. 20 .ig3

20 t2Jb6 leads to a drawing ending after 20... ..tb5 21 'ir'xb5 axb5 22 l:hd6 ..txd6 23 70

Game 24 Timman-Kengis

Yerevan O(ympiad 1996 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4 a6 5 .id3 lLlf6 6 0-0 d6 7 c4 g6

5 Ji.. d3 rt:J f6 6 0 - 0 d 6

After this move we reach positions simi­ lar to those in Games 1 -8 with two subtle differences: White's queen is on d1 instead of e2, and Black's queen is also at home (instead of c7). I believe that this difference works in White's favour. White's most ob­ vious plan of action is pressure down the half-open d-ftle against Black's vulnerable d6-pawn so the white queen is clearly well placed on either d1 or d2, allowing a dou­ bling on the ftle with _:.dl . Casting our minds back to Chapter 1 , White would of­ ten eventually play the move 'ife2-d2 any­ way, thus effectively losing a move. In the line here, White generally manages to put earlier pressure on d6 and it's more difficult for Black to reach an ideal defensive set-up. 8 rt:Jc3 Ji.. g 7

9 JigS

Not the only move by any means. White has two dangerous alternatives: a) 9 tt:J£3 aims for a similar set-up to the one considered in Games 1 -5: a1) 9 ...tt:Jc6 1 0 h3! (10 �f4 tLlg4 1 1 'ifd2 0-0 1 2 �e2 tt:Jge5 1 3 l:tfd1 tLlx£3+ 1 4 �x£3 tLld4 1 5 �e2 'ifc7 1 6 l:r.acl �d7 1 7 �fl i.c6 and Black had managed to equalise in Tsuboi-Hernandez, Havana 1 998) 10 ...tt:Jd7 (or 1 0 ... 0-0 1 1 � f4) 1 1 �g5 'ifc7 1 2 :c1 0-0 13 'ifd2 tLlce5 1 4 tLlxe5 tLlxe5 1 5 :fd1 (1 5 b3!?) 1 5...b6 ( 1 5 ...tt:Jxc4 1 6 �xc4 'ifxc4 1 7 'ifxd6 is awkward for Black) 1 6 �fl (16 b3 avoids any complications; I like White

after 1 6...�b7 1 7 �fl) 1 6 ... f6 (16...tt:Jxc4!?) 1 7 �e3 tLlf7 1 8 tLla4 :b8 1 9 'ifb4 and Black was in some trouble in Nyysti­ Cabrilo, Belgrade 2002. a2) 9 ... 0-0 10 i. f4 and now: a21) 10 ...tLlc6 1 1 h3!, followed by 'ii'd 2. a22) 10 ... tt:Jg4 1 1 i.e2 tLle5 12 tLlxe5! dxe5 13 'ifxd8 :xd8 14 �g5 f6 (14 ...:e8?! 15 tLla4 tLlc6 16 tLlb6 l:r.b8 17 c5 left Black struggling to complete development in Yu­ dasin-Movsesian, Pula 1 997) 15 :fd 1 �d7 16 �e3 l:tc8 17 :acl is a bit better for White; Black will find it hard to utilise the d4-outpost. a23) 10 ... tLlh5 1 1 �g5 'ifc7 12 'ifd2 tLld7 1 3 :acl with a typical edge for White, Aseev-Kochyev, St. Petersburg 1 995. Hav­ ing played 'ifd1 -d2 in one go, White is ef­ fectively a tempo up on similar lines consid­ ered in Chapter 1 . b) 9 tLlb3!? is recommended i n the popu­ lar Beating the Sicilian series by John Nunn and Joe Gallagher. 9 ... 0-0 10 i.e2 and: b1) 10 ...'ife7 1 1 � f4 l:.d8 12 'ifd2 tLlc6 13 l:.fd1 tLle8 14 i.e3 l:tb8 (or 14 ...tLlf6 1 5 �b6 l:td7 1 6 'ife3 and Black is in a tangle, Gallagher-Gokhale, Sangli 2000) 1 5 i.b6 tLlc7 1 6 .l:tacl and White has a pleasant ad­ vantage, Lutz-Cvitan, Passau 1 997. b2) 1 O ...tLlc6 1 1 �f4 tLle8 (1 1 ...tLle5 12 c5 tLle8 transposes to note 'b23') 12 c5! (12 'ifd2 b6 13 .l:tfd1 tLle5 14 :acl 'ifc7 was equal in Nunn-Gheorghiu, Vienna 1 986)

71

Sicilian Ka n

and now: b21) 1 2 ... e5 13 i..e 3 i.. e6 1 4 ll'ld5 i..xd5 15 exd5 ll'le7 (FeBland-Micklethwaite, cor­ respondence 1 994) and here I like the sim­ ple 16 .l:tc t , which looks pleasant for White. b22) 12 ... dxc5 13 'iix dB ll'lxdB 14 ll'la4! e5 (Nevanlinna-Seeman, Jyvaskyla 2001) and now I like the look of 15 i..e 3! ll'le6 (or 15 ... c4 1 6 ll'lb6 l:tbB 17 ll'lc5) 1 6 ll'lb6 .l:tbB 1 7 ll'lxc5. b23) 12 ... ll'le5 13 i..e 3 f5 14 f4 ll'lf7 and here I think White should play 1 5 exf5! (15 i.. O fxe4 1 6 i..xe4 d5 was a quick draw in Gallagher-Cvitan, Cannes 1 996; signifi­ cantly, though, Cvitan moved onto 10 ...'ii'e7 for his next encounter) 1 5 ... exf5 16 i.. O and Black has problems moving his pieces from the back rank.

White's position is much the easier to play. He could try to attack the weak dark squares on the kingside, but much the safer bet is to concentrate on the queenside, spe­ cifically the weaknesses on d6 and b6. 1 6 . . . 'ii'c 5 1 7 f3 J:.fe8 1 8 .l:tfd1 lt'le5 1 9 lt'la4 'ii'a 5?!

Obviously, this is not really what Black wants to do. 1 9 ... 1Wc7 20 iVb4 looks trou­ blesome for Black, but after 20... .l:tb8! I can't find anything too devastating for White. 20 'ii'x a5 bxa5 21

c5 dxc5 22 lt:\xc5

i.c8 23 �g 1 !

Bringing the king into the game. 23 . . . g5 24 �f2 �g8 25 �e3

9 . . . lt'lbd7 1 0 J:.c1

1 0 iVd2 0-0 11 .l:tad1 1Wc7 1 2 b3 b6 13 i..b 1 .U.e8 14 ll'lde2 i.. f8 15 'it>h1 i..b7 1 6 f4 was perhaps a little better for White in Tompa-Bellon Lopez, Metz 1 985, but I prefer Tirnman's handling of the position. 1 0 . . . 'ii'c 7 1 1 b3 0-0 1 2 'ii'd 2

White is in control. Only by a mixture of heroic defence by Black and inaccurate play by White does the game end in a draw. 25 . . . .l:td6 26 lt:\e2 llxd 1 27 l.txd 1 lt'lfd7 28 llc 1 lLlb6 29 ..td3 l:!.d8 30 l.tc3 f5 3 1 g3 fxe4 3 2 .be4 �f7 3 3 lt'ld4 �f6 34 a4 h5 35 l.tc2 h4 36 gxh4 gxh4 37 J:.g2 lLld5 + 38 ..txd5 exd5 39 f4 lt:\g4 + 40 �f3 lt'lh6 41 lt:\c6 l:!.e8 42 lt:\e5 ..tt5 43 1 2 . . . b6?!

J:.d2 .l:td8 44 �e3 h3 45 �f3 ..te4 + 46

I would be tempted to play 12 .. J�e8 here, planning to meet 13 .ih6 with 13 ... i..h 8!. It seems that with the exchange of the dark­ squared bishops, Black loses much of the potential counterplay in his position.

�f2 �f5 4 7 lt'lxe4 �xf4 48 lt'lc6 Y2 - Y2

1 3 ..th6! ..tb7 1 4 ..txg7 �xg7 1 5 i.b1 J:.ad8 16 �h 1 72

Game 25 Vogt-Gheorghiu

Hambur;g 1984 1 e4 c5 2 lt:\f3 e6 3 d4 cxd4 4 lt:\xd4 a6

5 � d 3 l1J f6 6 0 - 0 d6 5 .td3 l1Jt6 6 0-0 d6 7 c4 �e7

Again we eventually arrive at this classical development from Black. Variations here should be compared closely to those in Games 9- 1 1 , and of course there are many transpositional possibilities. Other possibilities for Black include: a) 7...lbbd7 8 lbc3 b6 is possible, but then White can exploit the fact that Black has committed his knight to d7 so early: 9 a4! ..tb7 10 aS and White breaks up Black's structure: 1 0... ..te7 (lO... bxaS 1 1 lbb3!) 1 1 ..te3! (1 1 axb6?! allows 1 1 ...'ifxb6 1 2 lbb3 .l:tb8!, planning to meet lbaS with ... ..taB, Dolmatov-Loginov, USSR 1 983) 1 1 ...0-0 1 2 axb6 lbxb6 1 3 lbb3 and White is better. b) An important alternative is for Black to choose first to fianchetto on the queen­ side. Although this sometimes simply trans­ poses, there are some independent points to remember: 7 ... b6 8 lbc3 ..tb7

though: 11 fS eS 1 2 lbc2 lbcS isn't so bad for Black on this occasion as the bishop is not particularly well placed on e3) 1 1 lbxc6 (the more restrained 1 1 'iff3 looks stronger) 1 1 .....txc6 1 2 eS!? dxeS 1 3 fxeS lbd7 14 'ii'g4 (14 .l:txf?? loses to 1 4...'it>xf7 1S 'ifhS+ 'it>g8 16 .l:tf1 'ife8!) 14 ... g6 1 S 'ii'g3 (or 1 S .l:txf7 �xf7 1 6 .l:tfl + 'it>g7 1 7 'ifxe6 .l:tf8! ­ Gheorghiu) 1 S ...'ifc7 1 6 i.d4 lbcS 1 7 ..tc2 .l:td8 1 8 l:tad 1 0-0 1 9 'ife3 .l:txd4! 20 'ifxd4 lbd7! 21 'iff4 lbxeS and Black has excellent compensation for the exchange, Kinder­ mann-Gheorghiu, Zurich 1 984. c) 9 f4 and now: cl) 9 ...lbc6 10 lbxc6 i.xc6 1 1 'ife2 ..te7 1 2 b3 0-0 13 i.b2 leads to note 'b' to White's 1 1 th move. A plus for White is that he has not committed his king to h 1 . c2) 9 ...l2Jbd7 allows White to force the issue with 10 fS! . As we have seen before, the type of position reached after 10 ... eS 1 1 lbc2 is generally better for White. c3) 9 ... ..te7 1 0 fS!' (10 'it>h 1 0-0 1 1 'ife2, reaching the main game, is of course possi­ ble) 10 ... lbc6 1 1 ..te3 lbxd4 12 ..txd4 0-0 13 fxe6 fxe6 14 'ife2 �h8 1 5 l:tad1 l2Jd7 1 6 .l:txf8+ 'ifxf8 1 7 lba4 ..tf6 1 8 ..txf6 'ifxf6 19 'ife3 .l:td8! and White was just a tiny bit bet­ ter in Adams-Ivanchuk, Linares 2002. c4) 9 ... g6!? is obviously risky but totally thematic; Black answers the £2-f4 lunge with a fianchetto. This is a favourite of the Ro­ manian GM Constantin Ionescu.

and now: a) 9 a4!? lbc6! (9 ... lbbd7?! 10 aS! trans­ poses to note 'a' above, but 9 ... ..te7, keep­ ing the option of ...lbc6, also looks okay) 1 0 lbxc6 ..txc6 1 1 ..te3 ..te7 1 2 f3 0-0 1 3 'ifd2 'ii'b 8 14 .l:tfd1 l:c8 1 S ..tfl lbd7 and Black was okay in Ivanchuk-Rublevsky, Monte­ catini Terme 2000. b) 9 ..te3 (trying to reach the set-up stud­ ied in Game 26) 9 ... ..te7 10 f4 lbc6!? (10 ... 0-0?! transposes into note 'a' to Black's 9th move in Game 26; 10 ...lbbd7 is possible 73

Sicilia n Kan

White can play: c41) 10 f5 gxfS! 1 1 exfS .l:tg8 12 tLl£3 eS 13 .igS tLlbd7 14 .ie4 bS! 1 5 .ixb7 'ii'b 6+ 16 'it>h1 'ifxb7 17 b3 .ie7 18 'ii'e 2 l::tc 8 and Black had reasonable counterplay in Magem Badals-Ionescu, Manresa 1 996. c42) 1 0 'it>h1 and now: c421) 10 ... .ig7 1 1 fS 'ii'e 7 12 fxe6 fxe6 13 'ii'a 4+! (Zeller) is problematic for Black: both 1 3. .. tLlbd7 14 tLlc6 and 13. ..tiJfd7 1 4 .ie3 leave Black struggling t o complete his development. c422) 10 ...tLlbd7! 1 1 fS 'ii'e7 (an advan­ tage of not playing an early .. .'ifc7) 1 2 fxe6 fxe6 13 'ii'B .ig7 14 'ifh3 tLlcS 15 i.gS 0-0 (Zeller) and Black has an ideal set-up. Zeller goes on to give 1 6 .ic2 'ifd7 with the idea of 1 7 .l:tae 1 ? lLlfxe4!. If this analysis holds up then 9 ... g6 looks like an excellent answer to 9 f4.

able. Conversely, a queenside fianchetto for White works well when Black plays ...tLlc6. So with moves like 'ife2, f4 and 'it>h1 White is playing logically while waiting to see how Black reacts. That said, there are also advantages with ploughing right ahead with a queenside fi­ anchetto, for example 9 b3 b6 10 i.b2 i.b7 1 1 'ife2 and now: a) Given what I have said before, 1 1 ...tLlbd7 looks most logical. After 1 2 f4 'ii'c 7 we transpose to note 'a' to White's 1 1 th move. b) 1 1 ...tLlc6 12 tLlxc6 .ixc6 13 .U.ae t !? (13 f4 transposes to note 'b' to White's 1 1 th move) 1 3 ...ttJd7?! (I believe Black should play 13. .. .l:te8! 14 f4 tLld7!)

8 lt:\c3 0-0

9 'i!fe2

At some point White will have to decide what to do with the dark-squared bishop (b3 and .ib2, or .id2). As John Nunn points out in Beating the Sicilian, White would quite like to know how Black develops his b8-knight before committing himself. Plans involving .id2, 'ife2 and .l:tae1 are most effective when Black has played ... tLlbd7. The point is that if White manages to achieve the advance e4-e5, then the f6knight does not have the d7-square avail74

14 tiJdS! .ih4 1 5 g3 .if6 16 tLlxf6+ lLlxf6 17 .l:td1 and the absence of Black's dark-squared bishop promises White a pleasant edge, Jansa-Polugacvsky, Sochi 1 976 For 9 i.e3 and other alternatives, sec Game 26. 9 . . . b6

9 ... tLlc6!? is not really a typical Kan idea but it is playable nevertheless. 10 tLlxc6 bxc6 1 1 f4 (1 1 b3 and .ib2 is also sensible) 1 1 ...e5 1 2 'it>h1 tLld7 (Bonsch recommends 12 ... exf4!? 13 .ixf4 tLld7) 13 f5! aS 14 .ic3 was a bit better for White, Adams­ Christiansen, Reykjavik 1 990. 1 0 f4 ..tb7 1 1 �h 1

5 i. d3 Ci:J f6 6 0 - 0 d 6

Again White may try to do without 'it>h 1 . After 1 1 b 3 we have: a) 1 t ...lLlbd7 (the most logical, given the reasoning above) 1 2 i..b 2 'Wc7 (12 .. J:te8!?) 13 l::ta d1 l::t fe8 14 i..b 1 transposes to Game 10. b) 1 t ...lLlc6 1 2 lLlxc6 i..x c6 1 3 i..b 2 lLld7 14 l::ta d1 and now probably best for Black is 14 ...g6!, asking the question whether White has anything better than to transpose into the main game with 1 5 'it>h 1 . At some point White has to do something about tricks on the a7-g1 diagonal but note that the imme­ diate 1 4... b5? doesn't work for Black: 1 5 cxb5 axb5 1 6 i.xb5 'Wb6+ 1 7 l::t f2! i..x b5 18 'ir'xb5 'iVxb5 1 9 lLlxb5 l::txa2 20 i..xg7 .l:!.fa8 21 l::txa2 l::txa2 22 i..d4 1eft White with a clear extra pawn in Hellers-Adamski, Eeklo 1 985. Black should answer 1 1 i..d 2 with 1 t ...lLlc6! 1 2 lLlxc6 i..xc6 and now White's bishop is misplaced on d2.

lLlxe8+ l::txe8 2 5 'ii £3 and White was better in Hellers-Servat, Gausdal 1 986.

1 6 . . .i.f8?

1 t ...lLlbd7! 1 2 i.d2 'Wc7 13 l::ta e1 trans­ poses to Game 9. Given the problems Black experiences in the text, I believe this is Black's best way to continue.

Black has two better defences: . a) 1 6 ... 'iVc7 1 7 'iVd3 i.. f8 1 8 lLlb5!? i..x b5 19 cxb5 axb5 20 'ir'xb5 'iVb7 (Nunn­ Gheorghiu, Hamburg 1 984) and now Nunn recommends 21 a4, relieving the b 1 -bishop of its defensive duty on a2 and pinning down the b6-pawn. This is enough to keep an edge. Against 1 6 ...'iib 8 Nunn likes 17 'iid3 b5 18 cxb5 axb5 19 lLle2 b4 20 lLld4 i..b7 21 'ir'h3 'with a dangerous kingside attack'.

1 2 Ci:Jxc6 .i.xc6 1 3 b3 Ci:Jd7

1 7 e5!

In earlier games Gheorghiu played 13. .. 'Wc7 here: 14 i..b 2 l:t.ad8 1 5 l::ta e1 i..b7 16 i..b 1 lLld7? (16 ...g6! 1 7 'ir'd3 is only a slight plus for White according to Nunn) 1 7 'ir'h5 l::t fe8 1 8 l::te 3 lLlf6 1 9 'ir'h3 g6 20 f5 gave White a winning attack in Nunn­ Gheorghiu, Biel 1 983: 20 ... i..c 8 21 l::tg3 'it>g7 22 'iVh4 l::t f8 23 i..c t l::td e8 24 e5 dxe5 25 'ir'h6+ 'it>h8 26 l::th3 l::tg8 27 i..g5 l::tg7 28 i..xf6 i..x f6 29 lLle4 'Wd8 30 fxg6 i..e 7 31 'ir'xh7+ l::txh7 32 l::tx h7+ 'it>g8 33 gxf7+ 'it>xh7 34 fxe8'iV 1 -0.

Now this is even stronger than in the previous note. White obtains a storming attack on the kingside.

1 1 . . . Ci:Jc6

1 4 .i.b2 g6 1 5 �ad 1 �eB 1 6 .i.b1

The direct 1 6 e5!? is also enticing: 1 6 ... dxe5 1 7 fxe5 'iic 7 1 8 i..e4 lLlxe5 (or 1 8 ... i.xe4 1 9 lLlxe4 lLlxe5? 20 'iif2 f5 21 'iVf4) 1 9 i..xc6 lLlxc6 20 l::tx f7! 'it>xf7 21 l::t fl + i.. f6 22 lLle4 e5 23 lLlxf6 'it>g7 24

1 7 . . . dxe5 1 8 .i.e4 'ikc7

1 8 ...i..x e4 1 9 lLlxe4 i..g7 20 i..xe5 i..x e5 21 fxe5 'it>g7 22 'iVf2 leaves Black with no good defence: 22...'iie 7 23 'ii f6+ ! 'it>g8 24 'ir'xe7 l::txe7 25 l::txd7 or 22 ... l::te 7 23 lLlg5 'ir'g8 24 lLlxf7 l::t f8 25 l::tx d7. 1 9 'ikf3 .i.xe4 20 Ci:Jxe4 f5

Or 20...i.g7 which is met by 21 fxe5 lLlxe5 22 lLlf6+ 'it>h8 23 'iVg3! and the knight is lost. 21 �xd7! fxe4 22 'ikd 1 'ikc6 23 .i.xe5 �acB 24 'ikd4 i.e7

24...i..c 5 runs into the attractive conclu­ sion 25 l::tg 7+ 'it>f8 26 i..d 6+! i..x d6 27 'iVf6 mate. 75

Sicilia n Kan

25 f5!

Excellent attacking play. The immediate 25 i.h8 ..i£8 26 .:.g7+?? 'it>xh8 27 :xg6+ fails to 27 ... e5! so White attempts to deflect the e6-pawn. 25 . . .llcd8

Now 25 ...exf5 26 ..ih8 wins after 26 ... -i£8 27 l:tg7+ Wxh8 28 .l:f.xg6+ . Black's best chance is 25 ...gxf5! 26 ..ih8 eS! (26 ... -i£8 27 .l:f.xfS! exfS 28 .l:f.g7+ Wxh8 29 .l:f.g6+ mates) 27 'il'xeS 'ii'g6 28 l:txe7 .l:f.xe7 29 'ii'xe7 Wxh8 30 'ii'e S+ 'ii'g7 31 .l:%.xf5 l:td8 32 h3 'ii'x eS 33 l:txeS l:td2 34 l:txe4 l:txa2 35 .l:.e6 bS 36 cxbS axbS 37 .l:f.b6 and White wins the ending. 26 �xe7 exf5

Or 26 ...l:txd4 27 l:.g7+ 'it>£8 28 fxg6 mate.

12 ..ie3 tLlg4 1 3 'ir'xg4 ..ixd4 14 'ir'e2 with an edge to White. c) 9 ...tLlbd7 and now: c1) 10 �e3 transposes into the main text. c2) 10 'ife2 'ifc7 transposes into note 'c' to White's l Oth move in Game 9, while Black can also consider 10 ... g6!?. c3) 10 'it>h 1 with a further split: c31) 10 ... 'ii'c 7?! 1 1 g4! is similar to the main game, but I think that Black has played ... 'ifc7 too early. He is not well pre­ pared to meet a kingside storm. c32) I believe that 10 ....:.e8! is best; a good, non-committal move, preparing ... ..i£8. Black should generally wait for White to 'waste' a move with 'ii'e2 before reacting with ...'ii'c 7. Play can continue 1 1 'ii'e2 'ii'c 7 and now: c321) 1 2 lLlf3 lLlcS 1 3 ..ic2 eS 1 4 fS bS! 15 ..igS b4 16 tLld 1 ..ib7 17 tLl£2 aS 1 8 ..ixf6 ..ixf6 1 9 tLlg4 a4 20 lLlxf6+ gxf6 2 1 lLld2 W £8 22 'ir'e3 We7 gave a n unclear po­ sition in Morovic Fernandez-Zapata, Cien­ fuegos 1 996. c322) 12 ..id2 b6 is a position discussed in Game 9, while also playable is 12 ... g6 1 3 tLl f3 ( 1 3 .:.ae 1 eS! as now there is n o tLlfS) 1 3 ... b6 1 4 eS tLlhS 1 5 .:.ae1 (Gallagher­ Erdelyi, Ticino 1 994) and now the sensible 1 5 ... -ib7 looks fine for Black. 9 . . . llJbd7

9 ... b6?! is very risky. White plays 10 f4! and now: Game 26 a) 1 0.....ib7?! 1 1 eS! dxeS 1 2 fxeS lLlfd7 Beliavsky-Gheorghiu 1 3 .:.xf7!! l:.xf7 1 4 lLlxe6 'il'c8 1 5 'il'hS tLl£8 1 6 tLlx£8 g6 1 7 .ixg6 hxg6 1 8 tLlxg6 .:.h7 1 9 Moscow 198 1 '-------• tLlxe7+ l:txe7 20 lLldS! ..ixdS 21 cxdS 'il'e8 22 'il'g4+ .l:.g7 23 'ii'e4 tLld7 24 e6 lLlcS 25 1 e4 c5 2 llJf3 e6 3 d4 cxd4 4 llJxd4 a6 i.xcS bxcS 26 d6 :d8 27 'ir'dS Wh8 28 :e1 5 .i.d3 liJf6 6 0-0 d6 7 c4 .i.e7 8 llJc3 :g6 29 :c3 :d7 30 :h3+ :h7 31 d7 1 -0 0-0 9 .i.e3 was very convincing from White's point of 9 f4!? is a very tricky transpositional view in Smagin-S.Salov, correspondence move: 1984. a) 9 ... b6 10 ..ie3 transposes to note 'a' to b) 10 ... 'ii'c 7 1 1 :c1 tLlbd7 (or 1 1 .....ib7 Black's 9th move. 12 fS 'il'd7 13 g4! h6 14 h4! tLlh7 15 gS! b) 9 ... d5!? is an untried suggestion from hxgS 16 hxgS ..ixgS 17 ..ixgS tLlxgS 18 'il'hS Zeiler. He gives 1 0 eS dxc4 1 1 ..ixc4 ..icS 27 �xeS + 'i'xeB 28 'i'xb6 1 -0

�------�

76

5 i.. d3 liJ f6 6 0 - 0 d6

ike7 1 9 f6 gxf6 20 l:.c2 with a winning at­ tack, Psakhis-Sideif Sade, Moscow 1 983) 1 2 g4! l2Jc5 ( 1 2. . .g6 1 3 g5 l2Je8 1 4 i.e2 i.d8 1 5 ike1 e 5 1 6 l2Jd5 ikb7 1 7 fxe5 dxe5 1 8 l2Jf3 ikc6 19 ikg3 was very good for White in Geller-Panno, Lone Pine 1 980) 13 i.b1 (but not 13 g5?! l2Jg4! 14 ikxg4 l2Jxd3 1 5 l:.c2 l2Jb4 16 l:.ccl l2Jd3, as in Marjanovic­ Rajkovic, Yugoslavia 1980) 1 3 . .. i.b7 1 4 ii f3

12 e5!? d4 13 exf6 ..txf6 1 4 i.xh7+ 'ifi>xh7 15 ikh5+ 'iti>g8 1 6 l:.ad1 was played in Murey-Zapata, La Valetta 1 980. Now 16 ... g6? 17 ikc5 d3 1 8 l:.£2 l:.b8 19 l:.fd2 ikc7 20 l2Je4 rurned out very well for White, but I prefer 1 6 ...l:.a7! 17 l2Je4 l:.d7, which looks less clear. 1 0 f4

The most active. One possible plan in­ volves a kingside pawn storm with g2-g4-g5. For the 'anti-Hedgehog' plan 10 f3 ikc7 1 1 l:.c l b6 1 2 ikd2 i.b7 13 l:.fd1 l:.ac8 14 i.fl see note 'a' to Black's 7th move in Game 75. 1o

and now: b1) 14 ... .l:.fd8 1 5 g5 l2Je8 16 f5 e5 1 7 l2Je6 l2Jxe6 1 8 fxe6 f6 1 9 l2Jd5 i.xd5 20 exd5 gives White positional domination, Stypka­ Przewoznik, Gdynia 1 982. b2) 14 ... l2Jfxe4? 1 5 i.xe4 l2Jxe4 16 l2Jxe4 f5 Qassim-Wians, Sharjah 1 985) 1 7 gxf5 exf5 1 8 ikg3 and the threat of l2Je6 nets White a piece. b3) 14 ...e5 1 5 l2Jf5 exf4 1 6 i.xf4 l:tfe8 1 7 b4 l2Je6 1 8 l2Jd5! and White wins material, Stypka-Grycel, Krynica 1 998. b4) 14 ...g6 (the best of a bad bunch) 1 5 g5 l2Je8 1 6 b4 l2Jd7 1 7 ikh3 and White has a strong attack, Lopez Rodriguez-S.Kova­ cevic, Mislata 1 995. 9 ... l2Jc6!? is not played very much, probably through habit (many Kan players are reluctant to play like this) . However, it has certain positive fearures: 10 l2Jxc6 (for 1 0 l:.cl i.d7 see the note to White's l Oth move in Game 23) 1 0 ... bxc6 and the bishop on e3 isn't necessarily well placed. 1 1 f4 d5!?

. . .

.:.es

Preparing to bolster the kingside de­ fences with ... i.f8 and ... g7-g6. 10 ...ikc7?! looks narural but the queen may well be better off at home if White is insisting on a g2-g4-g5 lunge. White contin­ ues with 1 1 g4! and now: a) 1 1 ...l2Jc5 12 i.c2 (but not 1 2 g5? l2Jg4!) 12 ... d5?! (12...g6 13 g5 l2Jh5 1 4 f5 and 12 ...e5 13 l2Jf5 i.xf5 14 exfS exf4 1 5 l:.xf4 are both good for White, but still preferable to what happens next) 1 3 exd5 exd5 14 g5 l2Jg4 1 5 l2Jxd5 l2Jxe3? (15 ...ikd8 is the only move) 1 6 i.xh7+! and White wins, Macieja­ Gratka, Koszalin 1 997. b) 1 1 ...g6 12 g5 l2Jh5 13 i.e2 l2Jxf4? (13 ...l2Jg7 1 4 f5 i.d8 is more resilient, al­ though still clearly better for White) 14 l:.xf4 e5 1 5 l2Jd5 ikd8

77

Sicilia n Ka n

(Marulovic-Pikula, Arandjelovac 1 997) 1 6 .l:tx£7!! and Black is lost i n all lines: b1) 16 ... l:Ix£7 17 lDe6 'tWaS 1 8 i..d2 traps the queen. b2) 1 6...'iitx f7 17 lDe6!! 'ifi>xe6 1 8 lDc7+! '3;£7 (18 ...'ii'x c7 19 'iWdS mate) 19 'iWdS+ 'iitg7 20 lDe6+. b3) 1 6...exd4 1 7 l:txe7 dxe3 1 8 'ii'd4 l:I£7 19 l:tx£7 'iitx £7 (or 1 9 ...\WxgS+ 20 'iith 1 'iitx f7 21 'iWhB!) 20 l:tft + 'iti>g8 21 'ii'xe3 and White is winning (analysis by Matulovic).

(planning to break up Black's queenside with a2-a4-aS) 1 S ... l:tac8 16 a4 'iWbB 17 aS and now in Pritchett-Gheorghiu, London 1 980, the Romanian GM played the the­ matic break 17 ... dS but after 1 8 cxdS i.. xb4 1 9 lDa2! l:txc1 20 i..x cl bxaS 21 lDxb4 axb4 22 dxe6?! fxe6 23 eS lDdS 24 'iWhS lDf8 2S fS White's attack was too strong.

1 1 �h 1

Naturally White can still attack like a caveman with 1 1 g4 but, with his rook on e8 and queen on dB, I believe Black is far better placed to combat it: 1 1 ...g6 1 2 gS lDhS, with the idea of undermining the sup­ port of gS with ... e6-eS, for example 1 3 i..e2? eS! 1 4 lDfS lDxf4 1 S lDh6+ 'iitg7 1 6 ..txf4 exf4 1 7 h 4 i..xgS! 1 8 hxgS 'ii'xgS+ 1 9 lDg4 lDeS and White's position falls apart. Instead White should take prophylactic measures against ...e6-eS, for example with 1 2 lDb3, but obviously the attack will then be a lot slower. White has a more methodical attacking plan at his disposal: 1 1 'ii'f3 !? i.. f8 1 2 l:tad1 'ii'c 7 13 'ii'h3 g6 (I prefer to leave well alone on the kingside and play 13 ... b6, for exam­ ple 1 4 lDf3 i..b 7 and now 1 S eS can be met with 1 S ...dxeS 16 fxeS ..tcS!) 14 lDf3 b6 1 S 'ii'h4 i..b7 1 6 lDgS! h6?! 1 7 lDxf7! 'iitx f7 1 8 e S dxeS 1 9 fxeS 'ii'x eS 20 i..e4! 'iitg8 21 i..d 4 i..c S 22 i..xb7 gS! 23 'ii' f2 lDg4 24 ..txeS i..x f2+ 2S l:txf2 lDdxeS 26 l:te2 and White had a better endgame, Ljubojevic­ Bellon Lopez, Linares 1 98 1 . However, 16 ...i..g7! doesn't look bad for Black.

1 2 . . . g6! 1 3 g5 lDh5

Now Black plans to undermine the sup­ port of the gS-pawn with ... e6-eS so White is forced into complications which are not unfavourable for Black. 1 4 f5 exf5 1 5 lDxf5 ltJc5! 1 6 ltJd5 lDxe4 1 7 i.d4 i.xf5!?

Also playable is 17 ...'iWxgS 1 8 i..e3 (18 i..xe4? l:txe4 1 9 'ii'x hS gxhS 20 l:tg1 J:tg4 21 lDf6+ 'ii'xf6 and Black wins) 18 ...'ii'd8 19 i..b 6, when Black can either repeat with 19 ... \WgS 20 i..e3 or try for more with 1 9 ...'ii'd 7!?. 1 8 �xf5 i.g7

18 ... gxfS? 19 'ii'x hS gives White a win­ ning attack: 1 9 ...i..g7 20 i..xg7 'iit xg7 21 i..xe4 fxe4 22 'ii'h 6+ 'iith 8 23 lDf6.

1 1 . . . i.f8 1 2 g4?!

19

Repeating what I said in the previous note, Black is now in a good position to defuse this idea. In this instance a more positional plan proves to be stronger: 1 2 'ii'e 2 'ii'c 7 (I think 1 2... g6!?, planning a possible ...e6-eS, is stronger) 1 3 l:tac1 b6 14 b4! ..tb7 1 S lDb3!

�xf7 + �xf7 22 'i'f3 + �gB!

78

i.xg7

�xg7

20

.i.xe4 l:!.xe4

21

The right square. 22...'iitg7?! 23 'ii'xe4 is good for White as 23 ... \WxgS? loses to 24 lDc7!. 23 'i'xe4 'i'xg5 24 'i'e6 + �hB 25 �f1

White has just enough compensation to keep the balance, but no more.

5 .1J.. d3 li:J f6 6 0 - 0 d 6

25 . . Jtd8 26 b3 b5 27 'ife4 bxc4 28 bxc4 'ife5 29 'ifxe5 + dxe5 30 l:!f7 li:Jf4 31 li:Jxf4 Y2 -Y,

Game 27 Hellers-Gheorghiu

Haifa 1989 1

e4 c5 2 li:Jf3 e6 3 d4 cxd4 4 li:Jxd4 a6

5 .1J.. d 3 li:Jf6 6 0-0 d6 7 f4! ?

move i s another question. Other possibilities for Black include: a) I don't like 7 ... 'ii'b 6?!, which can be met simply by 8 �h 1 ! . b) 7. . .1t...e 7 8 �h1 lLlbd7 (8. . .'ii'c 7 9 lt:Jc3 transposes to a line considered in Chapter 6, but there is no point committing the queen so soon) 9 'it' £3!? lLlc5 1 0 lt:Jc3 0-0 1 1 1t...e 3 'ii'c 7 12 a4 (12 g4!? b5 13 g5 lt:Jfd7 14 a3 1t...b7 is unclear) 12 ... b6 13 f5 e5 14 lt:Jb3 lt:Jxd3 (14... 1t...b7 1 5 lLlxc5 dxc5 1 6 b3! 1t...c6 17 g4 gave White a strong kingside attack in Tukmakov-Shamkovich, USSR 1 972) 1 5 cxd3 1t...b7 1 6 l:Iacl 'ii'd8 1 7 d 4 exd4 1 8 lt:Jxd4 l:te8! gave Black good counterplay in Romito-Carrettoni, correspondence 1 999. c) 7 ...lt:Jbd7!? keeps options open regard­ ing the fianchetto: c1) 8 lt:Jc3 transposes to a very complex line of the Scheveningen: 8 ...'ii'b 6! 9 1t...e3 'ii'xb2 10 lLldbS! axbS 1 1 lLlxbS l:taS 12 l:Ib 1 l:Ixb5 13 l:.xb2 l:.xb2 14 'it'a 1 l:Ib6 1 5 i.xb6 lt:Jxb6 1 6 'ii'c 3 i.e7 1 7 l:Ib1 lt:Jfd7 1 8 'ii'xg7 1t... f6 1 9 'ii'h6 (Anand-Kasparov, Tilburg 1 991) and now 1 9 ... l:.g8! 20 1t...b5 �e7 21 l:Id1 l:Ig4 is unclear according to Kasparov. c2) 8 �h1 g6 9 fS lLleS! reveals another advantage of having the queen on d8 over c7; the f6-knight is protected. Black can follow up with ... 1t...g7 and ... 0-0. 8 li:Jc3

A s with the line 6. . .'ii'c 7 7 'ii'e2 d6 8 f4!? (see Game 1 3), White doesn't need to commit himself to playing c2-c4 and can instead take immediate action on the king­ side. 7 . . . g6!?

As we have seen many times before, the kingside fianchetto is always the idealistic choice against an £2-f4 plan. Whether in this particular instance it's actually the best

Another idea is 8 lt:J£3 1t...g7 9 'it'c l . In a way this plan makes more sense than in the line 6 ...'ii'c 7 7 'ii'e 2 d6 8 f4 as White has obviously not wasted a tempo with the queen. However, Black has also not 'wasted' a move with ... 'ii'c7 and the queen is just as well placed on d8 in this line. The variation 9 ...0-0 10 �h 1 lt:Jc6 1 1 lLlc3 lt:Jb4! 1 2 eS lLlxd3 1 3 cxd3 lt:Je8 1 4 1t...e3 1t...d7 1 5 'ii' £2 1t...c6 1 6 1t...b 6 'ii'd7 led to an equal position in Kosten-Cvitan, Haifa 1 989. 8 . . . .1J.. g 7 9

cJi>h 1

0-0 1 0 f5!

I believe that this is the most critical line. White can also try the lt:JO/'ii'e 1 plan but, as we saw in the note to White's 8th move, 79

Sic ilia n Kan

this is not particularly dangerous for Black.

lent defensive moves. 1 8 h3 lbf1 +

1 9 .bf1 'ilff2 20 'ilfxf2

lLlxf2 + 21 �g1 ..txc3! 22 bxc3 lLle4 23 ..th6 ..ta6 24 ..txa6 �xa6 25 c4 �aS 26 cxd5 cxd5 27 llg7 + �hB 28 �d7 �gB

1 O . lLlc6 1 1 lLlxc6 bxc6 1 2 ..tg5! exf5 . .

Following 12 ...l::tb8? White has the the­ matic sacrifice 13 fxc6 fxe6 1 4 eS! dxeS 1 5 lDe4. 1 3 exf5 d5 1 4 'ili'd2 a 5

29 �xd5?

In order t o exchange one of White's at­ tacking weapons with ... i.a6.

29 a3! preserves some small wmmng chances.

1 5 .l:!.ae 1 'ili'b6

29 . . . lLlc3! 30 �d7 lLlxa2 31 c4 lLlb4 32

1 5 ...i.a6? 1 6 fxg6 fxg6 17 l::te 6! (Gheor­ ghiu) is very strong for White.

�g7 + �hB 33 llb7 lLld3 34 ..te3 �cB

1 6 fxg6 fxg6 1 7 .l:!.e 7 lLlg4!

�fB 38 .l:!.xh7 .l:!.c6 39 .l:!.a7 lLlb3 40

This just about keeps the balance, al­ though Black has to play a series of excel-

..tg7 +

80

35 .l:!.d7 lLlc5 36 ..td4 + �gB 37 .l:!.g7 +

Y, - Y,

�eB 4 1

..tc3 l::tx c4 42 ..txa5

5 i. d3 lb f6 6 0 - 0 d 6 Summary

Development with 7 ... ..id7 (Games 20-23) is certainly a viable alternative to main lines and there is the added advantage that white players are less likely to have studied these less fash­ ionable variations. As far as I can see, White's best chance of keeping some sort of advan­ tage lies with 1 1 l:td1 (Game 21) and 9 ..ie3 (Game 23). If Black is going to play ... g6 variations, then I believe that it's best to play those dis­ cussed in Chapter 1 as 7 ... g6 only seems to give White extra options. Lines with classical development shown in Games 25-26 are very similar to those in Chapter 1 (Games 9- 1 1 ), with the different queen placements proving to be a minor point. Despite the bad press, these lines are playable for Black as long as he knows exactly what he's doing regarding king safety. 1 e4 c5 2 lbf3 e6 3 d4 cxd4 4 /bxd4 a6 5 i.d3 /bf6 6 0-0 d6 7 c4

7 f4 - Game 27 7 . . . i.d7

(D)

7 ...g6 - Game 24 7 ... ..ie 7 8 lLlc3 0-0 9 'ife2 - Game 25; 9 .ie3 - Game 26 8 /bc3 /bc6 9 /bxc6

9 ..ie3 (D) 9

. . .

-

Game 23

..txc6 1 0 �e2

10 b4 - Game 22 10

. . .

..te7 1 1 b3

1 1 l:1d1 - Game 21 1 1 . . 0-0 .

(D) - Game 20

9 i.e3

1 1 . 0-0 . .

81

CHAPTER FOUR

I

5 i.d3 i.c5

1 e 4 c5 2 lUf3 e 6 3 d 4 cxd4 4 lDxd4 a6 5 .td3 .tc5

In this chapter we srudy the variation 5 .td3 .tc5. Black takes advantage of White's slightly cluttered d-file and attacks the knight on d4. The idea is to force the knight to an inferior square before deciding the furure of the bishop. Generally the dark­ squared bishop ends up outside the pawn chain along the a7-g1 diagonal, but Black does also have the option of a more defen­ sive retreat with ... .te7. Let's take a look at the opening moves: 1 e4 c5 2 lDf3 e6 3 d4 cxd4 4 lDxd4 a6 5 i.d3 .tc5 6 lDb3

tempo by attacking the bishop. 6 c3 and other sixth move alternatives are discussed in Game 4 1 . 6 . . . .ta7

Black's other main retreat here is 6 ... .te7, heading for a Hedgehog set-up where White's knight has been pushed from d4 to b3. For the standard white treatment with 7 0-0 d6 8 c4, see Games 37-38, while 8 f4 'ii'c 7 9 lLlc3 lLlf6 will be discussed under the move order 5 lLlc3 'ii'c 7 6 .td3 lLlf6 7 0-0 .ic5 8 lLlb3 .ic7 9 f4 d6 in Chapter Six. 7 'ii'g4 is perhaps the most critical variation. Black can then allow White to caprure on g7 with 7 ...lLlf6 (see Game 39) or prevent it with 7 ...g6 (see Game 40). 6 ... .ib6 is also possible and often trans­ poses to 6 ... -ta7 lines, although this is seen as a slightly inferior version as it prevents the possibility of ... b7-b5. White may con­ sider exploiting this by forgoing 'ii'e 2 and .te3, playing lines with an early 0-0 and either 'ii'g4 or c4 and �hl (as he does against ... .ia7 in Game 36). 7 'iife 2

By far the most popular response. White moves the knight again but regains the 82

White prepares to trade off Black's active bishop with .ie3. Traditionally this plan has been extremely popular. The main alternative is 7 0-0 and now: a) 7 ... lbe7!? is an interesting option,

5 i.. d3 i.. c 5

avoiding 'Wg4 lines. Of course, the flipside is that Black is then committed to playing ...ltJge7 lines. b) 7 ...l2Jc6 and now: b1) 8 'Wg4 is an extremely important idea which highlights a minus side to Black's plan - the g-pawn is lacking cover. Black can either sacrifice it with 8...l2Jf6 (see Game 34) or protect with 8 ...'Wf6 (see Game 35). b2) The plan starting with 8 c4, erecting the Maroczy Bind, is considered in Game 36. I will also mention 8 �h 1 , planning £2f4, as another possible plan. Other moves tend to transpose: 7 c4 l2Jc6 8 0-0 reaches note 'b2' above, while 7 l2Jc3 l2Jc6 8 'We2 d6 9 ..i.e3 transposes to the main line. 7 'Wg4 can be found in the notes to Game 34. 7 . . tt'lc6 8 i..e 3 .

8 l2Jc3 d6 9 ..i.e3 transposes. 8 . . . d6

For many years the main line consisted of Black simply capturing on e3 with 8 .....i.xe3 9 'Wxe3 and now: a) 9 ...l2Jf6!? allows White an extra oppor­ tunity to instigate complications: a 1) 10 l2Jc3 0-0?! (1 o. . d6! transposes to Game 33) 1 1 e5! l2Jg4 1 2 ..i.xh7+! �xh7 1 3 'ii'e 4+ �g8 1 4 'Wxg4 l2Jxe5 1 5 'ii'e 2! gave White a clear advantage in Klinger­ Meinsohn, Bad Lauterberg 1 984. b) 10 e5!? and now: b1) 1 0...l2Jg4 1 1 'Wg3 l2Jcxe5 (or 1 1 ...h5 12 h3 l2Jgxe5 13 'Wxg7 with a clear plus Klinger) 1 2 ..i.e2 f5! (12... h5 1 3 h3 h4 1 4 'Wc3 win s a piece) 13 h 3 f4! 1 4 'Wc3 'Wh4 1 5 0-0 h 5 1 6 ltJ1 d2! and, despite the extra pawn, Black's position is a bit of a mess. b2) 10 ... l2Jd5! 1 1 'Wg3 0-0 looks okay for Black, who will attack e5 with ... f7-f6. c) 9 ...d6! is the most exact move. A fter 10 l2Jc3 we reach Game 33. .

9 tt'lc3

With this move White keeps the option open of castling on either wing. White can

also play more positionally with 9 0-0, in­ tending c2-c4 (see Games 31 -32) .

9 . . . tt'lf6

By no means the only choice: a) I t's also possible to develop the knight with 9 ... l2Jge7!? (see Game 30). b) The move 9 ... b5!? is interesting; Black waits to see what White does before com­ mitting his g8-knight. 1 0 0-0-0 l2Jge7! trans­ poses to Game 30, while 10 0-0 l2Jf6! reaches Game 29. White can, however, also keep his options open with 10 f4. A fter 10 ...l2Jf6 1 1 0-0-0 we reach the main line. 1 0 0-0-0

Or: a) 10 f4!? b5 (10 ... e5!?) 1 1 i.xa7 .l:txa7 1 2 0-0-0 reaches the main line. b) 10 ..i.xa7 .l:txa7 1 1 f4 b5 (1 1 ...e5 12 f5 b5 1 3 0-0-0 b4 gives White the added pos­ sibility of 14 l2Jd5!?) 12 0-0-0 and again we reach the main line. c) White can opt for a quieter life with 1 0 0-0 (see Game 29), although it's less flexible here than on the 9th move as White no longer has the option of c2-c4. d) 10 g4!? is an interesting sideline: 10 ...b5 1 1 g5 l2Jd7 (Wolff-Christiansen, San Francisco 1 991) and now Christiansen rec­ ommends 1 2 f4, assessing the position as slightly better for White. Black can also play 10 .....i.xe3 1 1 'ii'x e3, transposing to the note to White's 1 1 th move in Game 33. 1 0 . b5 . .

83

Sicilia n Ka n

I believe this to be the most accurate move order. For 10 ...0-0 1 1 f4 eS 1 2 i.xa7 �xa7 1 3 fS bS 14 g4 b4 1 S tt:'la4, sec the note to Black's 1 4th move in Game 28.

... i.d7 and the c6-knight moving. I can't find any examples of 1 3 tt:'lb1, which does little to stem Black's counter­ play on the quccnside: 13 ...'iVc7 14 tt:'l1d2 aS! and Black will continue with ... aS-a4. 1 3 . . . e5

Preventing White from playing e4-eS himself, although 1 3. .. i.d7 14 cS dxcS 1 S fxeS tt:'ldS and 1 3. . .0-0 1 4 c S tt:'ldS arc plau­ sible alternatives. 14 f5

1 1 ..txa7

It makes sense to exchange bishops now as it forces Black to recapture with the rook. 1 1 f4?! is answered strongly by 1 1 ...b4! and now: a) 12 tt:'lb 1 does nothing to hold up Black's quccnside counterplay: 12 ... eS 13 ttJ 1 d2 0-0 1 4 �hfl i.xc3 1 S 'iVxc3 tt:'lg4 1 6 'iVg3 a S 1 7 �b1 a4 1 8 tt:'lc1 'iVc7 19 tt:'lc4 cxf4 20 'iVxf4 tt:'lceS and Black was clearly better in D.Gross-Bezold, Wi.irzburg 1 99S. b) 12 tt:'la4 i.d7! and now White has some problems with his knight on a4, as 1 3 i.xa7? loses to 1 3 . . .tt:'lxa7!.

Clamping down on the kingsidc. White has a simple plan of a kingside attack with g2-g4-gS. The ideas and tactics from this position are discussed in Game 28.

Game 28 Niggemann-Pfrommer

1 1 .. Jba7

Correspondence 1998

1 1 ...tt:'lxa7? not only looks silly, but 12 cS! is a very strong reply.

1 e4 c5 2 tt:lt3 e6 3 d4 cxd4 4 tt:lxd4 a6

1 2 f4

5 ..td3 ..tc5 6 tt:lb3 ..ta7 7 'ife2 tt:lc6 8

As we have already seen, move orders are very important throughout this variation. Here White could delay f2-f4 in favour of the immediate 12 g4!? b4 13 tt:'la4. Now 13 ...0-0 1 4 f4 eS 1S fS transposes to the note to Black's 1 4th move in Game 28, while 1 3. ..i.d7 14 gS forces 14 ... tt:'lg8.

..te3 d6 9 tt:lc3 b5 1 0 0-0-0 tt:lt6 1 1

1 2 . . . b4 1 3 tt:la4

This blocks Black's pawn advance on the queenside, but White must be careful as in some lines the knight can be snared by 84

..txa7 :xa7 1 2 f4 b4 1 3 tt:la4 e5 1 4 f5 ..td7!

Some exciting correspondence games have breathed new life into this variation as far as Black is concerned. Earlier Black had been playing 14 ... 0-0, but this move seems to run into a strong attack on the kingsidc. Nevertheless, there remain some unanswered questions (sec variation 'bS2' below). As far as I can sec,

5 i.. d3 i.. c 5

the critical line runs 14 ...0-0 1 5 g4 i.d7 (1 5...ltJd4? 1 6 ltJxd4 exd4 1 7 'ii'£2 and the d4-pawn drops) 1 6 g5 ltJe8 17 'ii'e3 (so that the problem knight can escape via b6) 17 ...l:tb7! (it's worth a pawn to keep the knight imprisoned) 18 i.xa6

and now: a) Nunn points out the natural 1 8 ...l:ta7, which had been previously neglected by players and annotators: 1 9 i.b5 ltJd4 and here Nunn gives 20 i.xd7 'ii'xd7 21 ltJb6 'ii'c 6 22 l:txd4! exd4 23 'ii'xd4 l:tc7 24 ltJa1 , which is slightly better for White, while I also like 20 ltJxd4 exd4 21 'ii'xd4 l:txa4 22 .ixa4 i.xa4 23 f6! . b) 1 8. . .l:tb8 i s answered strongly b y 1 9 .ic4!. I was o n the receiving end o f this over-the-board novelty in a game against the Russian GM Yuri Yakovich. In some lines Black does indeed win the a4-knight, but by that stage White is usually crashing through on the kingside. Earlier 1 9 'it>b 1 had been tried, but 1 9 ...'ii'c7 20 l:td5 (Speelman-Lobron, Munich 1 992) 20...ltJd4 (Speelman) 21 ltJxd4 exd4 22 l:txd4 i.xa4 sees White struggling to justify himself. After 1 9 i.c4 we have the following pos­ sibilities: b1) 19 ... g6? 20 fxg6 hxg6 21 h4 ltJd4 22 h5 i.xa4 23 hxg6 (Yakovich) and the threats of l:th8+ and 'ii'h3 are too much for Black to handle. b2) 19 ... ltJd4!? 20 g6 ltJf6 21 gxf7+ 'it>h8

2 2 ltJxd4 exd4 23 'ii'g5 (Yakovich gives 23 l:txd4 i.xa4 24 e5 as winning, but I'm not sure how White should continue after 24...'ii'e7!) 23 ... d5 (or 23 ... ltJxe4 24 'ii'xd8 l:tbxd8 25 l:the1 ltJ£2 26 l:td2 i.xa4 27 l:tx£2 d5 28 i.d3 l:txf7 29 l:tf4 and the pawn on d4 drops, while 23 ...i.xa4 loses to 24 l:thg1) 24 i.xd5 h6 (24... ltJxd5 25 'ii'x d8 l:tbxd8 26 ltJc5 ltJe3 27 l:txd4 l:txf7 28 l:te 1 ltJg4 29 l:ted 1 ltJe5 30 l:td5 l:te7 3 1 ltJe6 wins for White) 25 'ii'g6 ltJxd5 26 l:thg1 'ii' f6 27 ltJc5 ltJb6 28 'ii'x f6 gxf6 29 l:tg6 l:txf7 30 l:txh6+ 'it>g7 31 l:tg6+ 'it>f8 32 l:txd4 and White, who is picking up a fifth pawn for the knight, should be winning. b3) 19 ...'ii'c7 20 'ii'd3 (threatening to solve the knight problem with ltJac5!) 20...ltJa5 21 ltJxa5 'ii'xa5 22 b3 'it>h8 23 f6 l:td8 24 fxg7 + ltJxg7 25 l:thfl i.e6 26 'it>b 1 and White has a clear extra pawn with a good position, Chopin-Chorf, correspon­ dence 1 993. b4) 19 ...ltJa5 20 ltJxa5 'ii'xa5 and now: b41) 21 ltJc5 'ii'a7 22 g6 hxg6? 23 fxg6 dxc5 24 gxf7 + l:txf7 25 l:thfl ltJf6 26 l:txf6 gxf6 27 'ii'h6 1 -0 was the game Wojnar­ Stepanov, correspondence 1 994, but I'm not so convinced after the stronger defence 22 ... dxc5! 23 gxf7+ l:txf7. b42) 21 b3! i.xa4 22 bxa4 'ii'xa4 23 l:thg1 (threatening g5-g6) 23 ...'it>h8 24 l:tg3 l:tc8 25 .ib3 'ii'a 3+ 26 'it>b1 ltJc7 27 11h3 ltJb5

85

Sicilia n Ka n

28 .:.xh7+ ! (but not 28 g6?? lLlc3+ 29 �a1 'ii'b 2+ !!) 28...�xh7 29 'ii'h 3+ �g8 30 g6 .:.fd8 31 f6! and White soon mates. bS) 1 9 .. .'ith8 (my choice in the game) 20 'ii'd2 20...l2JaS 21 lLlxaS 'ii'xaS 22 b3 .ixa4 (22 ... .:.c8 23 'ii'd S! 'ii'a7 24 �b1 'ii'e 3 2S .:.he1 'ii'xgS 26 lLlb6 .ic6 27 'ii'aS 'ii'd 8 28 .idS .ixdS 29 .:.xdS .:.c6 30 l2Jc4 and White eventually converted his advantage, Yakovich-Emms, Cappelle Ia Grande 1 993) 23 bxa4 'ii'xa4 24 :hg1 lLlc7 (24 ....:.b6 2S .:tg3 lLlc7 26 l:th3 [Yakovich] seems very strong for White, for example 26 ...'ii'c 6 27 g6 'ii'xc4 28 .:.xh7+ �g8 29 f6! lLle6 30 .:.xg7+ lLlxg7 31 'ii'h 6 and White mates) 2S .:.g3 liJbS 26 .ixbS and now Black has two possibilities: bS1) Nunn gives 26 ... .:.xbS 27 'ii'xd6 .:tg8 28 .:,h3 'ii'xa2 29 g6 which wins for White on account of 29 ... fxg6 30 .:txh7+ ! or 29 ... h6 30 l:txh6+ gxh6 31 'ii' f6+ .:tg7 32 .l:td8. bS2) I prefer 26 ... 'ii'x bS! and believe that Black is still well in the game, for example 27 .:th3 'ii'c4 28 g6 'ii'xa2 29 lhh7+ �g8 30 'ii'gS 'ifa 1 + 31 'itd2 'ii'd 4+ is a draw by perpetual check as 32 �e2 'ii'x e4+ 33 �f2 'iix c2+ 34 .:.d2 loses to the surprising re­ source 34 ...'ii'x fS+!! 3S 'ii'x fS fxg6. Plenty of mind-boggling variations and in most of them White is winning, but line 'bS2' seems to make a good case for Black's defence.

1 6 . . . a5

If Black is unsure about the game con­ tinuation, then 1 6....ie8 also looks reason­ able: 17 liJdS aS 18 l2Jd2 l:tb7 1 9 lLlxf6+ 'ii'xf6 20 g4 l2Jd4 was unclear in Ruch­ Chorfi, correspondence 1998. 1 7 i.b5 �g4 1 8 ..g 1

1 8 . . . �d4!

This leads to an amazing sequence after which Black is a rook down but has a very strong attack against the white king. 1 9 �xd7?!

1 9 lLlxd4! leads to an unclear endgame after 1 9 ... 'ii'xb6 20 l2Jc6 'ili'xg1 21 .:thxg1 .ixc6 22 .ixc6 .l:td8. 1 9 . . . �xb5 20 �xf8 a4! 21 �d2

Perhaps 21 lLla1 , although Black's attack is still dangerous after 21 . . .a3 22 b3 'ifc7 23 'iie l l2Jc3. 21 . . . �d4 22 b1 -.c7 23 .l:!.c1 a3 24 b3

1 5 -.e3

-.c3 25 �c4 l:l.c7! 26 -.e 1 .l:.xc4 27

White must take measures to rescue the knight on a4. 1 S 'ii f2 0-0 leads to similar positions, while 1 S g4? l2Jb8! is good for Black.

-.xc3 .l:.xc3 28 �d7 �e3 29 �b6 �e2

1 5 . . . 0-0 1 6 �b6

16 h3 is too slow: 16 ... .:.b7 (threatening ...ltJaS or ...l2Jd4) 17 .ixa6 .:ta7 1 8 .:.xd6 .:txa6 19 lLlacS .:.xa2 20 lLlxd7 lLlxd7 21 .:.xc6 'ii'a8 22 .:.c7 .:.a 1+ 23 lLlxa 1 'ii'xa 1 + 24 Wd2 'ii'x h1 2S .:.xd7 'ili'xg2+ 26 'iie 2 'ii'x h3 and Black's extra pawn told in Marques-Chorfi, correspondence 1 997. 86

30 �a4 l:l.c6 31 l:l.he 1 �xc 1 32 .l:!.xe3 �xa2! 33 g4

Or 33 'it>xa2 l:txc2+ 34 �b1 .:.xg2 3S .:td3 .:txh2 36 l:hd6 hS! and Black's kingside pawns will win . 33 . . . �c3 + ! 34 �xc3 .l:!.xc3 35 .l:!.xc3 bxc3 36 �a2 d5 37 �xa3 dxe4 38 lt>b4 e3 39 �xc3 e4! ! 0-1

An incredible finish. White is in zugzwang, for example 40 gS �f8 41 h4 �e7 42 hS g6 43 fxg6 fxg6 44 hxg6 hxg6 4S

5 .1L d3 .1L c 5

b4 'it>d6 46 b S 'it>cS and White finally runs out of pawn moves.

Bezold, Bad Homburg 1 996. This idea of exchanging dark-squared bishops and then playing ... e6-eS is a typical equalising plan for Black. b) 1 1 a4!? b4 12 ltJd1 0-0 13 i.xa7 lha7 14 ltJe3 ltJaS 1 S ltJxaS 'ii'x aS 1 6 ltJc4 'i'cS 17 ltJxd6!? 'ii'xd6 1 8 eS 'ii'd4 19 exf6 'ii'x f6 20 'ii'e4 g6 21 'ii'xb4 :i.b7 22 'ii'c 3 'ii'xc3 23 bxc3 gave White an endgame edge in Acs­ Sax, Lillafured 1 999. However, Ribli's sug­ gested improvement of 1 4... dS looks okay for Black, for example 1 S exdS exdS 1 6 aS :.es 1 7 'ii' f3 (or 1 7 'ii'd2 �ae7) 1 7...�ae7, when the pawn on aS is just as vulnerable as the one on a6. 1 1 . Jba7 1 2 f4 b4 1 3 lLld1 .

Game 29 Hjartarson-Bezold

Bermuda 1997 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4 a6 5 .1Ld3 .1Lc5 6 lLlb3 .1La7 7 lLlc3 lLlc6 8 'i'e2 d6 9 .1Le3 lbt6 1 0 0-0 b5

1 3 . . . e5!

Despite the fact that White has opted for a quieter life by castling kingside, it's natural for Black to continue with queenside operations. 1 1 .1Lxa7

Or: a) 1 1 a3 0-0 12 �ad1 eS 13 i.xa7 (13 i.gS?! h6 1 4 i.h4? gS 1S i.g3 i.g4 wins material) 1 3. .. �xa7 1 4 'ii'e 3 i.e6 with a comfortable position for Black, Unzicker-

Model strategy from Black. After the bishop exchange, Black puts his central pawns on dark squares, thus complement­ ing his light-squared bishop and preventing White from playing e4-eS. It's true that White has a potential outpost on dS but he is in no real position to exploit this. 1 4 lLle3 'iib6 1 5 Wh 1 exf4!

Correctly relieving the tension in the cen­ tre and obtaining a very useful outpost on eS. 1 S ... O-O? would have allowed 16 fS! fol­ lowed by a very straightforward plan of g2g4-gS. 1 6 .l:!.xf4 i.e6 1 7 .l:!.af 1

17 ltJdS!? looks more threatening, al­ though after 1 7 ... i.xdS 1 8 exdS+ ltJeS 1 9 87

Sicilia n Kan

'ii'd2 aS 20 a3 bxa3 21 l:txa3 0-0 everything seems to hang together for Black. 1 7 . . . tt'le5!

With this dominating knight on eS, Black is guaranteed a comfortable equality. 1 8 tt'lt5 .bf5 1 9 exf5 0-0 20 'i'd2 �b8 21 .te2 .l:tc7 22 �d4 d5 23 'i't4 a5! 24 l:.d2 l:.e8 25 'i'd4 'i'xd4

Yz -Yz

Game 30 Z .Aimasi-Stangl

Altensteig 1994 1 e4 c5 2 tt'lt3 e6 3 d4 cxd4 4 tt'lxd4 a6 5 i.d3 .tc5 6 tt'lb3 .ta7 7 tt'lc3 tt'lc6 8 'i'e2 d6 9 .te3 tt'lge7

When black players were experiencing difficulties in the ... ltJf6 lines, many turned their attention to developing the knight on e7 instead. One advantage this development has over the classical ...4Jf6 is that White's pawn storm on the kingside is generally less effective as g2-g4-gS does not gain a possi­ bly crucial tempo on the knight.

Votava, Reykjavik 2002 - Black has some problems defending aS and d6. b) I can find only two examples of 10 ... 0-0!?, but I can see nothing wrong with it. Black is going to castle kingside after all, so why delay the inevitable? Play continues with 1 1 l:tad 1 bS! and now: b 1) 12 a4 (this works better when the rook is on at) 12 ...b4 13 ltJb1 ltJg6 14 ltJ1d2 'ii'f6 1S .ixa7 l:txa7 1 6 ltJc4 dS 17 ltJb6 ltJf4 1 8 'ii'e3 d4 1 9 'ii'g3 (Bashkov­ Tunik, Minsk 1 993) and now 19 ...eS! (Tunik) leaves Black standing well. b2) 12 a3 eS! 13 .ixa7 l:txa7 1 4 1i'e3 .ie6 1S ltJdS l:tb7 16 4::lx e7+ 'ii'xe7 17 4Jd2 aS was slightly better for Black in G.Ginsburg­ Tunik, Minsk 1 994. b3) 12 f4!?, leading to play similar to that in the note to 10 f4 below, is probably best. c) 10 ... .ixe3!? 1 1 'ii'xe3 eS comes into consideration given that the rook is not so effective on a7 when White castles kingside. Black has to be careful not to play the ... b7bS lunge too early (White may hit back with a2-a4) so the rook often shuffles back to a8 to re-enter the game. d) l O ... eS is the most common move. Black takes prophylactic measures against f2-f4, which will now be met by a pawn exchange.

1 0 0-0-0

Given the reasoning above, castling queenside is not such an attractive option for White as it is against ...ltJf6 lines, so many choose to castle short instead. After 1 0 0-0 Black has the following possibilities: a) 10 ... bS 1 1 a4! b4 12 4::lb 1 0-0 13 4::\ 1 d2 eS 14 4::\c4 4::lg6 1 S l:tfd1 .ie6 1 6 g3 aS 1 7 :td2 was a bit better for White, E.Berg88

Now 1 1 f4 exf4 1 2 :txf4 4::lg6 1 3 :tf2 4::\geS 14 h3 0-0 lS l:taft .ie6 was equal in Otero-Vilela, Havana 1999, as was 1 1 :tad 1

5 �d3 �c5

0-0 1 2 tiJdS i.e6 1 3 i.xa7 lha7 14 'ii'e3 bS 15 c3 l:r.b7 16 f4 exf4 17 tLlxf4 i..xb3 1 8 axb3 tLleS, Kotronias-Milov, FIDE World Ch., Groningen 1 997. White's main response is 1 1 i.xa7 l:txa7: d1) 12 f4 exf4 (Black should caprure: 12 ... 0-0?! 1 3 fS! is strong) 1 3 l:txf4 tLlg6 1 4 l:tffl 0-0 1 5 tLldS i..e 6 1 6 'iV £2 l:ta8 was equal in Kaminski-Milov, Moscow Olym­ piad 1 994. d2) 1 2 a4 i.e6 13 i.c4 0-0 14 l:tfd1 tLlg6 1 5 g3 l:a8 1 6 tiJdS l:c8 1 7 c3 tLlce7 and again Black had no problems, Naumann­ Computer P ConNers, Lippstadt 1 999. 1 0 f4!? makes Black think a bit more, as now 1 0 ... e5?! can be met effectively by 1 1 fS, when suddenly the knight is not so well placed on e7. Play can continue with 1 0 ... b5 1 1 0-0 tLlg6 12 l:tad1 (12 i.xa7 l:txa7 13 'ii'e3 0-0 1 4 l:tad1 l:td7 1 5 'i¥g3 eS!? 1 6 fS tLlf4 1 7 tiJdS tiJhS 1 8 'i¥e3 tLle7! 1 9 c4 bxc4 20 i.xc4 tLlxdS 21 i.xdS tLlf6 was at least equal for Black, Acs-Oral, Varadero 2000 White's knight on b3 is very poor) 1 2...'ii'e 7 13 l:td2 0-0 14 h1 i.xe3 1 5 'i¥xe3 i.b7 1 6 tLle2 l:r.fe8 1 7 c 3 l:r.ad8 with equality in Holrnsten-Oral, Istanbul Olympiad 2000. 1 0 . . . b5

l:tb8 and Black's attack is the stronger, Ernst-Sunye Neto, Manila Olympiad 1 992. b) 1 2 tLla4 i.d7! leaves White's knight on a4 in some trouble: 13 eS (13 'ii'd2? i.xe3 14 'ii'xe3 tLlc8! 1 5 eS tLlaS 1 6 i.e4 l:tb8 0-1 Saldana-Zapf, correspondence 2000; and 1 3 i.xa7? tLlxa7 Adams-Hjartarson, Paris [rapid] 1 994 show the precariousness of White's position) 13 ...tLld5! 1 4 i.xa7 (or 1 4 .id2 dxeS! 1 5 fxeS 'i¥c7! 1 6 l:the1 tLlce7 0-1 Coenen-Chuchelov, Fourmies 1 998) 14 ...tLlxa7 1 5 exd6 i.xa4 1 6 tLlcS 1 6 ...i.c6 17 tLlxe6 fxe6 18 'ii'x e6+ f8 19 i.c4 'i¥f6 20 i.xdS 'i¥xe6 21 i.xe6 i.xg2 22 l:thg1 i.c6 23 l:td4 l:td8 leaves White with insuffi­ cient compensation for the piece, Collins­ Sulava, Istanbul 2000. 1 1 . . . l:!.xa 7 1 2 f4

Or 12 'i¥e3 and now: a) 12 ... 0-0 13 eS!? dS (De Vreugt-Moo, Zug 2001) and now instead of 13 ... d5, I prefer 1 3 ... l:td7 as in the main game. b) 1 2 ... l:td7!? (preventing any tricks with e4-e5) 1 3 f4 "iic 7 14 g4 i.b7 1 5 h4 0-0 1 6 gS tLlaS 1 7 �bl tLlxb3 1 8 cxb3 (Sanchez Martin-Goczo, Oropesa del Mar 2000) and here I like the look of 1 8 ... d5!. 12 . . .0-0 13 e5!?

Perhaps deciding that Black will be at least equal in any pawn-storming race, White opens the d3-h7 diagonal and rums his attentions to the centre. 1 3 . . . l:!.d7 !?

1 1 �xa7

1 1 f4?! b4! looks promising for Black: a) 1 2 tLlb1 aS 13 tiJ3d2 'i¥c7 14 4Jf3 a4 1 5 tLlbd2 dS 1 6 i.xa7 l:xa7 17 eS i.d7 1 8 'iV £2 b3! 1 9 a3 bxc2 20 i.xc2 0-0 2 1 b 1 89

Sicilian Kan

Again we see how the rook can be useful along the second rank. Black can also block the centre and then organise counterplay on the queenside: 13 ...d5 14 'ii'h 5 lLlg6 1 5 g3 b4 t 6 lLle2 aS 1 7 lLlbd4 lLlxd4 1 8 lLlxd4 'ii'b 6 1 9 i..xg6 fxg6 20 'ii'e 2 i..d7 21 'iitb t l:tc7 22 l:td2 was agreed drawn in !.Rogers-Johansen, Mel­ bourne 1 998. White's positional advantages include a nice knight on d4, but Black's queenside counterplay mustn't be underes­ timated.

cxd3 lt:\xd3 24 'ii'd 2 lt:\xe 1 25 'ii'x e1 "ilfc4 26 �f3 i.b7 27 tt:lf6 + �g7 28 lt:\a5 'ii'c 7 29 �c3 'ii'xf4!

30 g3 'ii'd4 3 1

lt:\h5 + gxh5 3 2 lt:\xb 7 l:.bS 3 3 lt:\a5 �dB 34 Wc1

h6 35 tt:lb3 'ii'e4 +

36 'ii'c 2

'ii'e 1 + 37 lt:\c 1 .:td2 0-1

Game 3 1 Parma-Capelan

Solingen 1968 1 e4 c5 2 tt:lf3 e6 3 d4 cxd4 4 lt:\xd4 a6

1 4 'ii'e4

5 i.d3 i.c5 6 tt:lb3 i.a7 7 0-0 lt:\c6 8

Or 14 exd6 l:txd6 and now: a) 1 5 lLle4 l:td5 16 lLlc3 l:td6 1 7 lLle4 l:td5 is a repetition of moves; if Black wants to play on he could try 1 6 ...l:td7!?. b) 15 'ii'h S!? g6 1 6 'ii'c S 'ii'c 7 17 lLle4 looked a bit better for White in Jens­ Haldemann, Arco 2000, but I prefer 15 ... f5!, killing the bishop on d3 and preventing lLle4. Black then has time to pursue queen­ side operations.

'ii'e 2 d6 9 ..i.. e 3 i.xe3

14 . . . g6 1 5 exd6 l:bd6 1 6 'ii'e 3 'ii'c 7 1 7 lt:\e4 lt:\d 5!

Black can resist exchanging on e3 for the time being, but normally he will have to acquiesce in the long run. After 9 ...lLlf6 1 0 c 4 Black can play: a) 1 O... eS 1 1 lLlc3 i..g4 12 f3 i..e6 1 3 l:tfd 1 0-0 1 4 i.xa7 l:txa7 1 5 iVe 3 l:ta8 1 6 l:td2 and White's pressure down the d-ftle ensures an edge, Malakhov-Landa, Elista 1997. b) 10 ... 0-0 1 1 lLlc3 'ii'c7 12 l:tact i..xe3 (Black needs to do this if he wishes to play ... b7-b6) 13 iVxe3 lLle5 1 4 l:tfd l b6 1 5 i..e2 i.b7 16 f4 lLlg6 and we have transposed to the note to Black's 1 4th move. For 9 ...lLlge7 see Game 32. 1 0 'ii'x e3 tt:lf6

If anything I prefer Black's position. His pieces coordinate well and his king is slightly safer than White's. White's next move compounds his problems by allowing Black to gain a tempo later on. 1 8 'ii'f2 !? (Ribli) looks best. 1 8 'ii'g 3? l:.ddS 1 9 �hf1 lt:\ce7! 20 'ii'f2 lt:\f5 21 �de 1 tt:lb4 22 'iPb1 �xd3! 23 90

1 1 c4

Setting up the Maroczy Bind is the main point to White's decision to castle short. In

5 i. d3 i. c 5

contrast, 1 1 tbc3 promises nothing: 1 1 ... 0-0 12 .l:.ad1 'ii'c 7 13 'ito>h1 b5 14 f4 i.b7 1 5 'ii'h3 b4 1 6 tDe2 .l:.fdB 1 7 .l:.del h6! (taking the sting out of a possible e4-e5) and if any­ thing Black was a bit better in Padevsky­ Smyslov, Monte Carlo 1 968.

I believe Black should probably play more quietly with 1 4... b6, for example 1 5 f4 lbg6 16 g3 i.b7 1 7 .l:.acl l:tfdB 1 8 a3 .l:.acB 19 tbd4 i.aB and White was only a little bit better in Karpov-Hiibner, Graz 1 972. 1 5 i.xc4 'iixc4 1 6 l:!.xd6

1 1 . . 0 0 1 2 ltJc3 ltJe5 .

-

The alternative way to play for Black is to accept the 'Kalashnikov' pawn strucrure (reached after 1 e4 c5 2 tt:Jf3 tbc6 3 d4 cxd4 4 lDxd4 e5 5 lbb5 d6 6 c4) with 12 ... e5!? 1 3 .l:.fd1 i.e6 1 4 i.e2 and now 1 4. . .'ii'b 8!.

1 6 . . . e5?

With this move Black keeps defending d6, prepares to play ... .l:.fdB, and presents an option of opposing queens with ...'ii'a7. Play can then continue 1 5 :d2 l:td8 16 .l:.ad1 (16 .l:.cl !? prevents Black's next move) 1 6 ... b5!? 17 cxb5 axb5. Now 18 lbxb5?! .l:.xa2 1 9 .l:.xd6 l1xd6 20 .l:.xd6 tbd4! 21 lD3xd4 .l:.a1 + 22 i.fl exd4 was good for Black in the game Nicevski-Gheorghiu, Athens 1 969, but I prefer 18 i.xb5! lbb4 19 tbc1 tbxa2 20 tD1 xa2 ..txa2 when White's passed b­ pawn coupled with the pressure on the d­ ftle promise an edge. 1 3 i.e2 'iic 7 1 4 l:!.fd 1

After 14 l:tacl Black can grab on c4 as 14 ...tbxc4 1 5 ..txc4 'ii'xc4 16 lDd5? can be answered by 1 6 ... tbxd5!. Instead 1 6 .l:.fd1 'ii'b4 17 .l:.d4 'ii'h 6 1 8 ltcd 1 d5 leads to a level ending after 1 9 exd5 exd5 20 tbxd5 tbxd5 21 .l:.xd5 'ii'xe3 22 fxe3 i.e6. 1 4 . . . ltJxc4?!

After this move Black is lost, but his po­ sition is already difficult in any case: a) 16 ... b5 17 e5?! lbg4 18 'ii'f3 .l:.bB 1 9 .l:.d4 lDxe5 20 'ii'g3 'ii'c 7 21 .l:.cl f6 2 2 lDxb5 'ii'b6 was unclear in Klovans-Tunik, Par­ dubice 1 994. The move 17 l:tcl ! , however, looks a lot more threatening to me: 1 7 .....tb7 (or 1 7 ...lbg4 1 8 'ii'g3) 1 8 lba5 'ii'c 7 19 e5! lbg4 20 'ii'b 6 'ii'xb6 21 .l:.xb6 i.cB 22 .l:.e1 and this ending looks very suspicious for Black. b) 1 6 ...'ii'c 7! 1 7 .l:.ad1 is also better for White, although after 1 7 ... b5 1 8 e5 lDeB 1 9 .l:.dB i.b7 Black is hanging i n there. 1 7 l:lxf6!

This exchange sacrifice is simply devas­ tating here. 1 7 . . . gxf6 1 8 ltJd5 .l:td8 1 9 ltJxf6 + �f8 20 .l:l.c1 i.e6 21 h4 'iid 3 22 'iih 6 + �e7 23 'iig 5 �d6 24 ltJc5 'iib 5 25 ltJxe6 fxe6

Or 25 .. .'iti>xe6 26 lbg4 l:t£8 27 'iff6+ 'ito>d7 28 tbxe5+ 'it>eB 29 .l:.c7. 26 'iid 2 + 1 -0

White mates after 26 ...'ito>e7 27 .l:.c7+ 'ito>xf6 28 'ii'g5. 91

Sicilian K a n

Game 32 Herschei-Fiensch

Comspondence 1979 1 e4 c5 2 ltJf3 e6 3 d4 cxd4 4 ltJxd4 a6 5 .i.d3 .i.c5 6 ltJb3 .i.a7 7 'ir'e2 ltJc6 8 ..te3 d6

Another possibility for Black is to delay the advance of the d-pawn: 8 .. .'!2Jge7 9 0-0 0-0 10 c4 eS 1 1 tt:\c3 tt:ld4!? (1 1 ...d6 trans­ poses to the note to Black's 1 1 th move) 1 2 �xd4 exd4 1 3 tt:ldS d 6 1 4 l:tadl tt:lc6?! (14 ... tt:\xd5 15 cxdS looks only a little better for White) 15 'it>h 1 i.e6 16 f4 .l:.e8 17 'ifhS g6 1 8 'ifh6 .i.xdS (Leko-Kramnik, Wijk aan Zee 2000) and now Leko's suggestion of 1 9 cxdS! looks very good for White, for exam­ ple 19 ...tt:lb4 20 f5 'iff6 21 i.bl d3 22 a3 tt:\c2 23 .l:.xd3. 9 0-0 ltJge7

Uncovering the d-ftle. White has a straightforward plan of ganging up on d6pawn. 1 4 . . . l:!.c8

14 ...tt:\c8 is very passive: 1 5 tt:ldS aS 1 6 .l:.acl b 6 1 7 tt:la 1 ! (this knight is heading somewhere nice!) 1 7 ... f6 1 8 tt:lc2 .l:.f7 1 9 tt:la3 l:td7 20 tt:lbS .l:.b8 (Unzicker-Petrosian, Palma de Mallorca 1 969) and now either 21 f4 or 21 g3 (preparing f2-f4) would have kept a significant advantage. 1 5 .l:.d2 ltJa5

Originally this position was assessed as equal by ECO, and this seems to have caught out quite a few players who were religiously following the 'bible'! 1 6 ltJxa5 'ir'xa5 1 7 .l:.xd6 ..txc4 1 8 l:!.d7!

Naturally! The double attack nets White a pawn. 1 8 . . . ..txf 1 1 9 .l:.xf1

19 .U.xe7 also looks good.

Experience has shown that this devel­ opment is not very effective when White can play c2-c4. 1 0 c4!

10 tt:\c3 transposes to the note to White's 1 Oth move in Game 30, but 10 c4 looks stronger. 1 0 . . . 0-0 1 1 ltJc3 .i.xe3

1 1 ...e5 12 .ixa7 .l:.xa7 13 .l:.fd l i.e6 1 4 'ife3 .Ua8 reaches the same position, albeit with an extra move played by both sides. 1 2 'ir'xe3 e5 1 3 .l:.fd1 .i.e6 1 4 .i.f 1 1 9 . . . 'ir'b4??

Given that this was a postal game, there's really no excuse for this incredible blunder. I suspect, however, that Black was simply following what had gone on before. 19 ...tt:\c6! is best: 20 .l:txb7 .U.b8 21 l:txb8 .l:.xb8 22 b3 tt:ld4 23 'it>h 1 .l:.c8 24 .l:tcl h6 and it won't be easy for White to convert his extra pawn, Leroy-Davies, correspon­ dence 1 982. 20 a3!

The 'stem game' had continued 20 'ifgS?? 92

5 i. d 3 i. c 5

and was eventually drawn in Lewandowitz­ Herschel (!), correspondence 1 973. A dou­ ble piece of luck for Herschel! 20 .. .'ii' x b2 21 lL:Jd1

And White wins a piece ... 21 . . .'ii'a 1 22 l:be7 b5 23 'ii'b 3 llc 1 24 lL:Je3 �xf1 + 25 lL:Jxf1 'ii'd 4 26 'ii'd 5 'ii'c3 27 'ii'd 6 a5 28 �xeS b4 29 �xa5 b3 30 llb5 1 -0

Game 33 Kengis-B.Nevednichy

USSR 1979 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lL:Jxd4 a6 5 i.d3 i.c5 6 lL:Jb3 i.a7 7 'ii'e 2 lL:Jc6 8 lL:Jc3 d6 9 i.e3 i.xe3

For many years this was the main line be­ fore black players began allowing White to exchange on a7 (see Game 28).

available as in Game 30, but 1 3 ...d 5 fol­ lowed by pushing the 9ueenside pawns looks okay (compare with the note to Black's 13th move in Game 30). 1 1 0-0-0

White can also go for it on the kingside immediately with 1 1 g4!? and now: a) 1 1 ...b5 12 0-0-0 0-0 13 gS lLle8 (13 ...lLld7 14 i..e2 leaves the d6-pawn vul­ nerable) 14 f4 b4 1 5 lLle2 aS 16 ltJbd4 lLlxd4 17 ltJxd4 and White was a bit better in Anand-Ninov, Baguio City 1 987. b) 1 1 ...lLlxg4! 12 'ii'g3 lLlf6 13 'ir'xg7 l:.g8 1 4 'ii'h6 i..d7 1 5 0-0-0 'ii'e 7 1 6 lLla4 0-0-0 17 ltJb6+ ..ti>c7 1 8 lLlxd7 l:.xd7 19 l:.hg1 l:.dd8 was e9ual in Ebeling-Haapasalo, Tampere 1991 . 1 1 . . . 0-0 1 2 f4

1 0 'ii'x e3 lL:Jf6

Also possible is 10 ...lLlge7!?, developing in a similar fashion as in Game 30. Now 1 1 0-0 0-0 leads to similar play to that in the note about 10 0-0 in Game 30, the only difference being that the bishops have al­ ready been exchanged. White can also play 1 1 0-0-0 and now: a) 1 1 ...e5 12 f4! exf4 13 'ii'x f4 0-0 1 4 i.e2! was good for White i n Renet-Farago, Budapest 1 987. The game continued 14 ...ltJg6 1 5 'ii'g3 ltJgeS 16 l:.d2 f5? 17 exfS i.xfS 1 8 l:thd 1 l%f6 1 9 ltJd4 lLlxd4 20 .l:txd4 'ii'c 7 21 i.c4+!! lLlxc4 22 ltJdS 'ii'c S 23 lLlxf6+ ..ti>h8 24 lLle4 'ii'c 6 25 'ii'c 3 lLleS 26 lLlxd6 and Black resigned. b) 1 1 ...0-0 12 f4 'ii'c 7?! 13 'ii'h 3! lLlb4 14 .l:thf1 eS 1 5 fS dS 16 exdS lLlexdS 1 7 lLlxdS lLlxdS 1 8 'ii'h 4! lLlf6 19 g4 e4 20 g5! exd3 21 l:.xd3 gave White a strong attack in Ko­ rolev-Batakovs, correspondence 1 984, for example 21 ...lLlh5 22 'ii'x hS i.xfS 23 l:.xfS g6 24 'ii'h4 gxfS 25 l:.h3 f6 26 g6!. Instead of 1 2 ... 'ii'c 7?!, I believe Black should start immediate 9ueenside operations with 12 ... b5. After 13 eS Black has no ...l:.d7

1 2 . . . 'ii'c 7

The line 12 ... b5? 13 eS! is good for White, as 13 ... b4 fails to 14 exf6 bxc3 1 5 'ii'e4. The other way to prevent e4-e5 is with 1 2 ... e5 1 3 fS bS and now: a) 14 g4!? lLlxg4 1 5 'ii'g3 and now instead of 1 5 ...ltJf6 16 l:thg1 lLle8 17 lLldS which gave White a strong attack in S.Hansen­ Fries Nielsen, Aarhus 1 999, Black should play 1 5 ...'ii'g 5+ 16 l:.d2 'ii' f4 17 lLldS 'ii'xg3 1 8 hxg3 when White certainly has compen­ sation, but there's no obvious way to con­ tinue. b) 14 i..e2! b4 15 lLla4 'ii'c 7 1 6 g4 i.d7 93

Sicilian Kan

1 7 gS liJd4 1 8 liJxd4 exd4 1 9 l:hd4 ltJxe4 20 lhe4 .txa4 21 .td3 l:.ae8 (or 2 1 ....tc6 22 l:tc4 .U.fe8 23 'ifh3 'ifas 24 .l:td 1 .tbs 2S l::th4!) 22 l:.e1 and White's advanced pawns on the kingside promise a clear edge, Chris­ toph-Hollis, Hastings 1 96S.

ltJa4 is certainly worthy of attention) 16 ... aS with a further split: b1) 17 'ifh3 eS? 1 8 f5 a4 19 liJd2 liJcS 20 i.c4 b3 21 g6 hxg6 22 .l:txg6 ltJe7 (or 22 ... bxa2 23 .l:txg7+! 'ifi>xg7 24 .l:tg1 + 'ifi>f6 2S 'ifh4 mate) 23 .l:tg4 bxa2

1 3 g4!?

This is the most ambitious way to play. If Black declines the pawn offer then White's attack is quicker than normal as he hasn't been forced to play a preparatory l:.hgl . The main alternative is 1 3 .l:thg1 bS Qogi­ cally Black gets on with his queenside op­ era tions) when the assessment of the posi­ tion depends upon the relative speeds and stn:ngths of the two attacks. Play continues 1 4 t;4 b4 1 S gS

and now: a) 1 S...liJe8 1 6 liJe2 aS 17 liJbd4 liJxd4 18 'ifxd4 .ta6 19 i.xa6! (19 'ifi>b 1 eS 20 'ife3 a4 21 .l:tg3 was unclear in Arnason­ Suetin, Sochi 1 980) 19 ... l:txa6 20 f5 .l:tc6 21 'ifd3 exfS 22 exfS dS 23 f6! gxf6 24 liJd4 'iff4+ 2S �b1 l::tc4 26 liJfS 'ifi>h8 27 'ifxdS 'ife4 28 liJh6 'ifxc2+ 29 �a1 .l:tc7 30 gxf6 a4 31 'ii'gS 'ir'g6 32 'ii'e3 and Black resigned in Spitz-Larrouy, correspondence 1 993 on account of 32 ... 'ifxf6 33 .l:tdfl 'ifc6 34 'ir'gS 'ii'g6 3S 'ii'e S+ f6 36 lhg6 fxeS 37 .l:txf8 mate. b) 1 S ...liJd7! (many annotators have not mentioned this despite the fact that it looks more narural than 1 S ...liJe8) 1 6 liJe2 (16 94

24 .l:tdg1!! (go on - have another queen!) 24...al'ii' + 2S liJb1 and White has a winning attack, for example 2S ... liJxfS 26 exfS liJe6 27 .l:th4 g6 28 f6 and White mates. I couldn't resist putting in this line, but I should point out that Black has a strong improvement with 1 7 ... .ta6! 1 8 eS .txd3 1 9 exd6 'ii'a7 20 l:txd3 a4 2 1 liJbd4 liJxd4 22 liJxd4 'ireS, which is very unclear. b2) 17 liJbd4 liJxd4 1 8 liJxd4 liJcS 1 9 liJbS 'ii'b 6 20 i.c4 and now: b21) 20... .l:td8? 21 fS .ta6 22 g6 fxg6 23 .:txg6! hxg6 24 fxg6 was winning for White in Virtual Chess X-P ConNerS, Paderborn (computer tournament) 1 999, as there is no good defence to 'ifh3. b22) I wouldn't want to question a com­ puter, but 20 ... dS! looks like an improve­ ment: 21 exdS .ta6 22 g6! .txbS (22... hxg6? 23 dxe6 .txbS 24 exf7+ .l:txf7 2S .txf7+ 'ifi>xf7 26 l:.dS wins) 23 gxh7+ (23 gxf7+ .l:txf7 24 dxe6 .txc4 2S exf7+ .txf7 26 'ii'e S g6 looks good for Black) 23 ... 'ifi>h8 24 .txbS 'ii'x bS 2S 'it'eS f6 26 'it'c7 gS 27 fxgS 'it'c4 28 'ifi>b1 .l:tac8 29 'it'a7 with massive compli­ cations. b3) 17 fS a4 1 8 g6!? fxg6 (1 8 ... axb3? 1 9

5 i.. d3 i.. c 5

gxh7+ �h8 20 �gS g6 21 fxg6 fxg6 22 �xg6 tt'lf6 23 eS! [threatening �xf6+] 23 ... tt'lxh7 24 exd6 �f7 25 �h6 and White wins) 19 tt'lbd4 tt'lxd4 20 �xd4 (20 tt'lxd4 tt'leS) 20 ... b3!? 21 fxg6 tt'leS 22 gxh7+ �h8 and if I had to choose, I would take Black, Atanasov-Rodin, correspondence 1 984. Finally, the line 13 �h3 bS 1 4 g4 tt'lb4 1 5 gS tt'ld7! leads t o the note t o White's 1 5th move.

Halle 1978, I like 1 8 ...l:tb8! intending ... b5b4 with good counterplay. 1 5 . . . b4 1 6 lLle2 a5 1 7 Wh3

1 7 . . . exf5?

1 3 . . . b5

13. ..tt'lxg4!? 14 'iWg3 tt'lf6 15 l:thgl tt'le8 obviously promises a strong initiative to White, who has scored well in practice. One recent example is 1 6 �h4!? �e7 17 l:tgS f6 1 8 l:tg3 g6?! (18 .. .f5 1 9 l::.g S tt'lb4 20 tt'ld4 tt'lf6 was unclear in Ivanovic-Velikov, Saint John 1 988) 19 f5 tt'leS 20 tt'ld4 i.d7 21 fxg6 hxg6 22 i.c4! and now: a) 22 ... tt'lxc4 23 l:txg6+ tt'lg7 24 l:tdgl tt'leS (24...l:tf7 25 l:th6 wins) 25 l:t6g3! and there is no good answer to l:r.h3. b) 22...tt'lg7 23 l:r.dgl �f7 24 tt'lfS! tt'lxfS 25 exfS l:th8 26 �e4 bS 27 l:txg6 1 -0 Freiknecht-Scheidt, correspondence 1 998.

The rather obvious-looking 17 ... tt'lde5! is suggested as an improvement on MegaCorr2 (a database of over 350,000 correspondence games) without any further analysis. Now the same plan doesn't work for White: 1 8 tt'lf4 a4! and there is no tt'ldS so White is forced to retreat with the b3knight. Therefore the most logical way for­ ward for White looks to be 1 8 f6, which led me to some incredible complications: 1 8... a4 19 fxg7 l:td8! (19 ...�xg7? 20 tt'lg3! tt'lxd3+ 21 lhd3 tt'leS 22 tt'lhS+! �g6 23 tt'lf6 tt'lxd3+ 24 'ifi>d2! and mate follows) 20 tt'lg3 (after 20 tt'lbd4 tt'lxd4 21 tt'lxd4 a3! it's Black's attack which is stronger) 20... axb3 21 tt'lhS (threatening mate with tt'lf6+)

1 4 g 5 lZ:ld7 1 5 f5!?

Another way to continue the attack is 1 5 �h3 tt'lb4 1 6 a3?! ( 1 6 �bl ! looks stronger) 1 6 ...tt'lxd3+ 1 7 �xd3 dS! 1 8 eS and here, instead of 1 8 ... i.b7?! 19 l:r.hgl .U.ac8 20 h4 tt'lcS?! 21 tt'lxcS �xeS 22 �d4 aS? 23 �xeS l:txcS 24 tt'le2 which gave White a big posi­ tional advantage in T.Giorgadze-Bohlig, 95

Sicilia n Kan

Black has two possibilities: a) 21 ...lbxd3+? 22 l:txd3 'ike7 23 'ikh4!! (23 lDf6+ 'ilf'xf6 24 gxf6 bxa2 25 Wd2 a1'ilf' 26 l:ha1 l:ha1 27 'ikh4 e5! wins for Black) 23 ... bxa2 24 Wd2 a1'ik 25 l:ha1 l:.xa1 26 l:lh3! h6 27 lDf6+ 'ilf'xf6 28 gxf6 and White wms. b) 21 ...lbg4! 22 'il'xg4 (or 22 e5 lbcxe5 23 �xh7+ Wxh7 24 g6+ Wxg6 and Black wins) 22 ...lbe5 23 'ilf'h3 lbxd3+ 24 'ii'xd3 bxa2 25 Wd2 'il'c5 looks winning for Black. It may be that this is just the tip of the iceberg and that there are several other pos­ sibilities for both sides, but it does show you how a whole line can be (sometimes incorrectly) written off just as a result of one high-profile game. 1 8 exf5 �de5 1 9 �f4 a4 20 �d5 'jj'd 8 21 .:l.hg 1 !

2 1 o o o�xd3 + ?

Another mistake, although by now Black is in a difficult position. Gipslis gives the line 21...axb3 22 lDf6+ gxf6 23 'ilf'h6! lbg6! (23 ... Wh8 24 gxf6 l::tg8 25 l::tg7 lbxd3+ 26 :xd3 �xfS 27 l::th 3! and White wins) 24 fxg6 fxg6 25 �xg6! l::ta7 26 gxf6 hxg6! and then Nunn continues the analysis with 27 .l:l.xg6+ Wf7 28 l::te 1 ! lbe5 29 l::tg 7+ We6 30 1:.xa7 bxa2 31 'ii'h 3+ Wd5 32 'ii'b 3+ Wc6 33 'il'xa2 'with a slight material plus for White together with a raging attack'. Indeed, it's very doubtful that Black can survive in this final position. 96

22 .:.xd3 �e5 23 �f6 + ! gxf6 24 Wh6! ! �xd3 + 25 �b 1 ! fxg5

Or 25 .. oWh8 26 g6! fxg6 27 fxg6. 26 f6 Wxf6 27 Wxf6

The smoke has cleared, leaving White with a winning material advantage. 27 og4 28 Wg5 + �h8 29 Wt6 + �g8 0 0

30 �d4 �e5 31 h3 h5 32 Wg5 + �g6 33 Wxh5 gxh3 34 Wd5 i.e6 35 �xe6 h2 36 l:.xg6 + 1 -0

Game34 J . Horvath-Farago

Budapest 1987 1 e4 c5 2 �f3 e6 3 d4 cxd4 4 �xd4 a6 5 i.d3 i.c5 6 �b3 i.a7 7 0-0

The immediate 7 'iig4 is not so effective: 7 ...lbf6 8 'iig3 (the point is that 8 'ilf'xg7? runs into 8 ...l1g8 9 'ilf'h6 �x£2+!) 8 .. od6 9 lbc3 lbc6 10 0-0 h5!? (10... b5, 10 ...lbe5 and 10 ... 0-0 are also playable, but I like this move) and now: a) 1 1 Wh 1 lbe5 12 f4 h4 1 3 'ilf'e1 (13 'ikxg7?? lbg6!) 13. ..lbxd3 1 4 cxd3 h3 15 g3 b5 and Black's two bishops will point im­ pressively towards the white king, Mainka­ Kengis, Germany 1 997. b) 1 1 'ii'xg7 (1 can't find any examples of this move) 1 1 ...l::tg8 12 'ikh6 lbg4 13 'ilf'xh5 (13 'ii'h7 lDf6 repeats moves) 1 3. ..'ilf'f6 1 4 lbd 1 'ikg7! 1 5 'il'h3 l:lh8 1 6 'ii'g3 l::tx h2 and ...'ii'h 8! is corning. c) 1 1 i.e3 h4 12 'ifxg7 l::th7 13 'ikg5 l::th 5 1 4 'il'g7 l::th7 is a draw by repetition. d) 1 1 i.g5 lDe5 12 i.h4 Wf8 13 l::ta d1 lbg6 14 �e2 lbxh4 1 5 'ilf'xh4 'ilf'e7 16 'ii'g3 e5 1 7 'iid3 i.e6 1 8 'ii'xd6 ('/2-1!2 Boey­ Arnlind, correspondence 1 975) 18 ...'il'xd6 19 l:txd6 �xb3 20 axb3 �d4 21 lDd5 i.c5 22 l1xf6 gxf6 23 lDxf6 is unclear. 7 o o�c6 8 Wg4 �f6 o

For the safer 8 ...'ikf6, see Game 35. 9 Wxg7

For 9 'ikg3?! d6 10 lbc3, see the note to White's 7th move.

5 1J.. d3 1J.. c 5 9

. . .

J:tg8 1 0 'ifh6

White is a pawn up but Black certainly has some compensation in the form of open lines and active pieces. 10

. . .

l:tg6!?

The main alternative is 1 0 ...ltJe5 and now: a) 1 1 'ir'h3 ltJfg4! 12 ltJc3? (better is 1 2 'ir'xh7 ltJf6 1 3 'ir'h3 ltJfg4 and probably White should now repeat with 1 4 'ir'h7 as after 14 i.e2 I like 14 ... 'ir'f6 1 5 'ii'g3 'ii'h8!) 12 ...'ir'f6 13 'ir'g3. Now 13 ... ltJ£3+ 14 gx£3 itJh6 1 5 i.xh6 l:hg3+ 16 hxg3 'ii'x h6 17 �g2 was better for White in Korneev­ Gonzalez Gil, La Coruna 1 999, but 13 ...l:tg6! looks very strong to me: 14 'ir'f4 ltJ£3+ 1 5 Wh1 'ii'h4 and Black is winning. b) 1 1 i.f4? ltJeg4! 1 2 'ir'h4 (or 12 'ii'h3 e5! followed by ... d7-d5!) 12 ...ltJx£2 13 .l:tx£2 l:.g4 14 'ii'h6 lOgS 15 'ii'x h7 i.x£2+ 1 6 �x£2 l:txf4+ 1 7 'it>g1 'ii'g5 was winning for Black in Erdelyi-Szeberenyi, Budapest 2001 . c) 1 1 i.e2!, covering the £3- and g4squares, looks to be White's strongest re­ sponse: c1) After 1 1 ...b5, Ljubojevic's recom­ mendation of 1 2 i.e3! l:tg6 (or 12 ...ltJeg4 1 3 i.xg4 l0xg4 1 4 'ir'xh7) 1 3 'ir'h3 leaves Black with insufficient compensation for the pawn. c2) 1 1 ...d5 and now: c21) 12 i.g5 ltJeg4! 13 i.xf6 ltJxf6 1 4 exd5 'ir'xd5 1 5 g3 'ir'f5 1 6 ltJ1d2 ltJg4 1 7

i.xg4 l:txg4 was unclear in Van der Wiel­ Ehlvest, Groningen 1 997. c22) 12 ltJc3!

with a further split: c221) 12 ... dxe4 13 i.g5 l:tg6 (13 ... ltJfg4 1 4 i.xd8 ltJxh6 1 5 i.c7! ltJf3+ 1 6 �h1 ltJh4 17 g3 ltJ4f5 18 ltJxe4 [Ribli] leaves White just a pawn up for nothing) 14 'ir'h4 i.d7 (or 1 4...ltJf3+ 1 5 i.x£3 ex£3 1 6 ltJe4 l:txg5 1 7 ltJxg5 fxg2 1 8 l:tfd 1) 1 5 ltJxe4 i.c6 1 6 i.xf6 'ii'd 5 1 7 i.xe5 'ir'xe5 1 8 i.f3 1 -0 was a pretty convincing argument for White in the game Hellers-Sjoberg, Malmo 1 994. c222) Ribli gives the line 1 2 ... d4!? 13 l:td 1 ltJeg4 14 i.xg4 ltJxg4 1 5 'ii'x h7 ltJf6 1 6 'ii'h4 .:.g4 1 7 'ir'h6 l:tg6 1 8 'ir'f4 and con­ cludes that White has a clear advantage. I agree with this, but I would point out that Black keeps some practical chances after 1 8 ... 'ii'e 7! 1 9 ltJe2 e5 20 'ii' f3 i.g4 21 'ii'd3 0-0-0. 1 1 'ifh3

1 1 'ii'd2?! ltJe5 12 �h 1 ? ltJfg4! 13 'ir'e2 ltJxh2 was winning for Black in Davie­ Kaplan, Oerebro 1 966. Possible, however, is 1 1 'ir'h4!? and now: a) 1 1 ...l:tg4 12 'ir'h3 (a little nuance so that the rook blocks the c8-h3 diagonal) 1 2...ltJe5 13 ltJ1d2 d5 (13 ... b5!? may be stronger) 1 4 ltJ£3! ltJx£3+ 1 5 'ir'xf3 dxe4 1 6 i.xe4 i.b8 ( 1 6. . ..:.xe4 1 7 i.g5!) 1 7 h3 .:.gs 1 8 i.xb7 i.xb7 1 9 'ii'xb7 with a clear plus for White, Santo Roman-Koch, French Ch. 97

Sicilian K a n

1 996. b) 1 1 ...lLle5 12 .ie2 b5 is interesting as White can no longer play .ie3. Play can continue 13 .ig5 .ib7 14 lLl1d2 h6 1 5 .ixf6 nxf6 (Ljubojevic-Lobron, Plovdiv 1 983) and now 1 6 c3 (Ljubojevic) keeps a slight advantage. Instead, the game contin­ ued 16 :ad 1 ? tt:lg6 17 'ii'x h6 lLlf4 18 'ii'h 8+ We7 19 'ii'x d8+ nxd8 20 .if3 nh8?? 21 e5 and now Black could have won with the pleasing 21...nfh6 22 g3 lLle2+! 23 .ixe2 nxh2 24 .if3 nh 1 + 25 .ixh 1 nxh 1 mate.

pawn on h5 looks a litde silly. 1 0 i.g5 'ii'g 6 1 1 'ii'h 4

1 1 . . . e5

1 1 . . . f6

In Popovic-Schlosser, Brno 1 992, Black played the 'trappy' 1 1 ...lLle5?? (intending 1 2 .ixe7?? tt:lf3+) but had to resign after 12 .ie2! as there is no good way to deal with the twin threats of 13 .ixe7 and 13 .ih5!. 1 2 i.d2

12 .if4 lLlb4 13 .id6 lLlxd3 14 cxd3 lLlc6 1 5 l:.fe 1 tt:le5 1 6 'ii'h3 lLlf7 1 7 .ig3 0-0 was fine for Black in Remus-Figueiras, corres­ pondence 1 997. Planning ...d7-d5. 1 2 'ii'h 4

White offers a draw by repetition, which is accepted. Playing on involves some risk: 1 2 g3 (Alben-Varela, Buenos Aires 1 999) and now I like 1 2... h5! 1 3 tt:lc3 d6 14 'ii'g2 h4, with real counterplay against the white king. 1 2 . . . l:.g4 1 3 'ii'h 6 J:.g6 1 4 'ii'h 4 l:.g4 1 5 'ii'h 6 l:tg6 % - Yz

Game 35 Berthelot-Velikov

Trignac 200 1 1 e4 c5 2 tt:lf3 e6 3 d4 cxd4 4 tt:lxd4 a6 5 i.d3 i.c5 6 tt:lb3 i.a7 7 0-0 tt:lc6 8 'ii'g 4 'ti'f6 9 tt:lc3 tt:lge7

After 9 ... h5?! White simply retreats with 1 0 'ii'e 2 and will continue with .ie3. The 98

1 2 . . . tt:le5!

An improvement over the earlier 1 2...'ii' f7 1 3 Wh 1 g5 14 'ii'h6 'ii'g6 1 5 'ii'h3 lLle5 16 f4 gxf4 1 7 nxf4, which gave White an edge in Popovic-Martinovic, Novi Sad 2000. 1 3 ..te2 'ii'f7 1 4 Wh1 g5! 1 5 'ii'h 6 tt:\7g6 16 f4

Blasting open the f-ftle, but this does give away the e5-square forever. I know that the move 1 6 g3, intending to recapture on f4 with the g-pawn, is more thematic but there is an obvious risk to the white king along those long diagonals: 1 6 ... b5! 17 f4 gxf4 1 8 gxf4 tt:lc4! 1 9 f5 exf5 20 ltJdS! (20 exf5? tt:lge5 21 .ih5? loses to 21....ib7+) 20 ... d6! and I would have to favour Black in this very messy position. 1 6 . . . gxf4 1 7 i.xf4 tt:lxf4 1 8 1:1xf4 h1

White wins the f-pawn but Black's con­ trol of the dark squares and powerful knight situated on e5 provide sufficient compensacion. 20 'ilixf6 + 'ilixf6 21 .l:txf6 b5 22 h4?

White's plan to expand on the kingside is doomed to fail. 22 l"Lld1 i.g5 23 .:t6f2, with the plan of opposing Black's knight with l"Llc5-d3 (or l"Lld4-f3), is much stronger. 22 . . . .l:tg8 23 'it>h2 l:.g7 24 'iti>h3 .ib7 25 tt'la5 l:.ag8! 26 tt'lxb7

Or 26 g4 i.a8 and White's position col­ lapses. Black threatens both ... b5-b4 and ...l"Llxg4. 26 . . . l:f.g3 + 27 'iti>h2 lbg2 + 28 'iti>h 1

Signalling a different plan to the one in­ volving 'ife2 and i.e3. White intends to attack on the kingside with f2-f4 and/ or i.g5, hoping to exploit the absence of a dark-squared defender on the kingside. 10 'ife2 0-0 1 1 i.e3 would transpose to a line considered in Game 3 1 . 1 0 . . . h5!?

28 . . . .ig 1 ! 0-1

Dissuading White from playing an Im­ mediate f2-f4 on account of ...l"Llg4. Black has two other important alternatives: a) 1 0...e5 (again preparing ...l"Llg4) 1 1 i.e2! i.e6 (1 1 ...h6 1 2 f4 exf4 1 3 i.xf4 l"Lle5 14 c5 i.b8 1 5 cxd6 'ifxd6 1 6 'ifxd6 i.xd6 17 .:tad1 and 1 1 ...0-0 12 f4 exf4 13 i.xf4 l"Lle5 14 i.g5 both look good for White) 1 2 99

Sicilia n Kan

f4 h6 1 3 'it'd3 'tlfe7? 1 4 ltJdS �xdS 15 cxdS l2Jd4 16 fxeS dxeS 17 l2Jxd4 .ixd4 18 'it'g3! left Black struggling in Bellia-Haldemann, Arco 1 999. b) 10 ... 0-0 and now: b1) 1 1 'ife2?! eS 1 2 �gS h6 1 3 �h4 gS 1 4 .ig3 �e6 was fine for Black in Sandvik­ Kauko, Finnish Team Ch. 1 998. b2) 11 f4 'ifc7 12 .id2 l2Jd4 13 �cl l2Jxb3 (there is no need to rush this; 1 3. ..�d7 is preferable) 14 axb3 �d7 15 g4 �c6 16 gS l2Je8?! (Black should play the more natural 16 ... l2Jd7, not fearing 17 b4 �d4 1 8 bS? on account of 1 8 ... l2Jc5!) 1 7 b4 'i*'e7 1 8 bS axbS 19 cxbS �d7 20 eS! gave White a strong attack in Ramesh-Emms, British Ch. 2002. 1 1 .tg5 'ili'c7 1 2 'ili'd2?! A mistake. Shirov suggests 12 f4! h4 1 3 'iff3 with an edge t o White. 1 2 . . . ..td7 1 3 �ad 1 lt'le5 14 .te2 .tc6? Black misses a great chance for counter­ play: 14 ...l2Jfg4! 1 5 �h4 f6! 16 'ifxd6 'ifxd6 1 7 �xd6 gS 1 8 �g3 h4 1 9 �xeS ltJxeS (Shi­ rov) and Black's control of the dark squares and well-placed pieces provide enough compensation for the pawn deficit. 1 5 f3! Preventing any ...l2Jg4 tricks and securing a clear plus. 1 5 . . . 0-0-0

Given that White was threatening the d6pawn, this looks natural. However, now 1 00

White is able to unleash a strong attack on the queenside. Perhaps 1 5 ... �d8 is a try. 1 6 lt'lb5! axb5 1 7 cxb5 .tea 1 8 �c 1 ..tc5 1 9 lt'ld4 'it>b8 20 b4 b6 2 1 a4 h4 22 aS lt'lh5 23 .txh4 g5 24 'ili'xg5 f6 25 'i'e3 'ili'g7 26 g4 lt'lf4 27 bxc5 lt'lxe2 28 lL:lxe2 �xh4 29 c6 �c8 30 'i'xb6 + 'it>aS 31 c7 1 -0

Game37 Shirov-Agrest

European Team Ch., Leon 200 1 1 e4 c5 2 lt'lf3 e6 3 d4 cxd4 4 lt'lxd4 a6 5 .td3 .tc5 6 lt'lb3 i.. e7 The main alternative to 6 ...�a7. Black prefers to keep his bishop as a defensive piece on the kingside. 7 0-0 d6 8 c4 b6 9 lt'lc3 ..tb7 1 0 f4 lt'ld7 Although Kan players tend to prefer sys­ tems with ... l2Jd7, Black can also develop with ... l2Jc6, for example 10 ... l2Jc6 1 1 �e3 l2Jf6 12 'iff3 l2Jd7 1 3 �ad1 'ifc7 14 'ifh3 ltJcS 1 5 ltJxcS bxcS 1 6 fS with an edge to White, Rytshagov-Agrest, Stockholm 1 998. Significantly, perhaps, Agrest has since switched to 10 ... l2Jd7. 1 1 .te3 With this move White auns for 'third rank' set-up, which will be completed with 'iff3. For the main alternative 1 1 'ife2, see Game 38. 1 1 . . .lt'lgf6 1 2 'ili'f3

5 �d3 �c 5

From here the queen has options to go to both g3 and h3. On g3 it attacks the g7pawn while on h3, combined with f4-f5, it helps to pressurise the e6-pawn. White will also seriously consider the space-gaining g2g4-g5, especially since Black's f6-knight doesn't have its natural retreat square at d7. We have reached an important moment; Black now needs to make a difficult deciSIOn. 1 2 . . . g6! ? By no means the only choice. Black can also play: a) 12 ... h5!? (Agrest likes to play this move - see also Game 38) 1 3 .l:tad1 "fic7 14 h3 g6 (Agrest assesses the position after 1 4... h4 1 5 'i'f2 ltJcS 1 6 ltJxcS dxcS 1 7 e S ltJd7 1 8 fS as good for White, but I think it's quite play­ able for Black after 18 ...0-0-0) 1 5 "fi f2 ltJcS 16 ltJxcS dxcS 1 7 eS ltJd7 18 ..ie4 0-0-0 was agreed drawn in Rytshagov-Agrest, Istanbul Olympiad 2000. After the bishops are ex­ changed Black can aim for counterplay with ...ltJb8-c6. Nevertheless, I still prefer White and, given that he chose not to repeat this against Shirov, perhaps Agrest agrees. b) 1 2...ltJc5 1 3 ltJxcS bxcS 1 4 "fih3 ltJd7 15 eS! ( 1 5 f5 ltJeS 16 ..ic2 0-0 17 fxe6 ..ic8! [Ribli] is fine for Black as 18 ex£7+? .l::t x£7 19 "fig3 ..ih4! traps the white queen) 1 5 ... dxe5 1 6 f5 1 6 ...ltJf6 1 7 .l:tad 1 "fib6 1 8 fxe6 "fixe6 1 9 ..ifS gave White reasonable compensation for the pawn, Shirov­ Kveinys, Istanbul Olympiad 2000. c) 1 2... 0-0 1 3 .l:tad1 .l:te8 14 "fih3 "fic8!? (covering the e6-square although 14 ... "fic7 15 g4! g6 1 6 gS ltJhS 17 f5 exfS 1 8 exfS ltJeS doesn't look bad for Black) 1 5 g4 g6 1 6 gS ltJhS 1 7 flo ..i£8 1 8 "fif2 "fic7 1 9 ..ie2 ltJg7 20 ltJd4 ltJcS 21 ..if3 eS! 22 ltJde2 .l:tab8 (22 ... exf4?! 23 ltJxf4 ltJxe4? 24 ltJxe4 ..ixe4 loses to 25 ltJdS) 23 fS and now, instead of 23 ... a5 24 ..ig2 which was good for White in Vallejo Pons-Korneev, Elgoibar 1 997, I prefer 23 ... gxf5 24 exfS e4! followed by ...ltJd3.

1 3 l:!.ad 1 'i!i'c7 1 4 'ili'h3 h5!? Again we see this move. Black doesn't want to commit his king and plays actively on the kingside. 14 ... 0-0 1 5 fS! is pleasant for White. 1 5 f5 gxf5! A big point of 12 ... g6. 1 6 exf5 e5 Black has blocked the centre and now seeks counterplay down the g-file and along the long h 1 -a8 diagonal. 1 7 �e2! White prepares to oppose Black's power­ ful bishop. 1 7 . . .l:!.g8 18 �f3

1 8 . . . �xf3? Giving up the fight too easily. After this move White manages to obtain crucial light­ square domination in the centre. Instead Black should aim for the complications of 1 8 ... e4 1 9 ..ie2 ltJeS 20 ltJd2, which Shirov assesses as unclear. Now 20 ... 0-0-0?! 21 liJdS! is good for White, so I prefer 20 ... ltJeg4! with ideas of ... d6-d5 and ... ltJxe3. 1 9 'ili'xf3 .l:!.b8 20 tt'ld2! b5 21 cxb5 axb5 22 tt'lde4 b4 23 tt'ld5 tt'lxd5 24 l:!.xd5 tt'lf6 25 :lc 1 'ili'b7 26 tt'lxf6 + �xf6 27 l:!.cd1 h4 28 �f2 'it>f8 29 'ili'h3 :ld8 30 �c5 :lg5 31 �e3 l:!.g8 32 �c5 :lg5 33 'i!i'd3! :d7 34 �f2 l:!.g8 35 h3! 'i!i'c6? Dropping a pawn, although by now Black's position is extremely difficult. 101

Sicilian Kan

36 �e4 .:.c7 37 �xb4 i.e7 38 i.xh4 f6 39 .:.1 d2 �c1 + 40 h2 .:.c4 41 �b8 + f7 42 �b3 �f1 43 .:.t2 �c1 44 .:.d1 1 -0

Game 38 Moberg-Agrest

1 2 . . . h5!? The most ambitious of Black's options, though not necessarily the best. Black gives the knight a square on g4 and can also con­ sider ... h5-h4 and ... lLlh5. Black's alternatives include: a) 12 ... g6?! 13 e5! is strong, as 1 3 ... dxe5 14 fxe5 lLlxe5? fails to 1 5 'i'xe5 'i'xd3 1 6 .l:.xf6. b) 1 2... 0-0 1 3 .l:.ae1 and now: b1) 13 ....l:.e8!? 14 .l:.f3 .tfB 1 5 .l:.h3 e5 1 6 f5 b5!? (thematic but. . .) 1 7 cxb5 axb5 1 8 .txb5 d5 1 9 exd5 lLlxd5 20 a4 doesn't give Black enough for the pawn. b2) 1 3 ... .l:.c8 1 4 .l:.f3! (but not 14 e5? dxe5 1 5 fxe5 lLlxe5!) 1 4... .l:.e8 1 5 l:h3 lLlfB 1 6 e5 ltJ6d7 17 'i'h5 g6 18 'i'h6 was better for White in Kutuzovic-Gallagher, Pula 2000. c) 12 ...'i'c7 (this looks the safest way) 1 3 .l:.ae1

Swedish Team Championship 200 1 1 e4 c5 2 lt'lf3 e6 3 d4 cxd4 4 lt'lxd4 a6 5 i.d3 i.c5 6 lt'lb3 i.e7 7 0-0 d6 8 c4 b6 9 lt'lc3 i.b7 1 0 f4 lt'ld7 1 1 �e2 As we have seen before, the plan involv­ ing 'i'e2, .td2 and .l:.ae1 is one of the most dangerous available to White. These posi­ tions should be compared to the ones where the white knight is on d4. The knight is obviously less active on b3, but it is also less vulnerable to counter-attack. 1 1 . . .lt'lgf6 1 2 i.d2 with a further split: cl) 1 3 ...0-0 14 e5! looks better for White. c2) 1 3 ... h5!? 1 4 lLld4! (White needs to get this knight back into the action) 1 4 ... g6 ( 1 4... 'i'c5?! can be simply met by 1 5 lLla4!) 1 5 lLlf3 and now Black's kingside looks a bit shaky after 1 5 ... 0-0?! 1 6 f5. I believe that Black should play a waiting game with 1 5 ... .l:.d8, although the question remains about the final destination of Black's king. c3) 1 3 ...g6? 14 e5! dxe5 1 5 fxe5 lLlh5 1 6 ltJd5!! exd5 1 7 e6 fxe6 1 8 'i'xe6 ltJdf6 1 9 �b4 lLlg8 20 �xe7 lLlxe7 21 .l:.f? was win1 02

5 i. d3 .i. c 5

ning for White in Langheinrich-Glek, Bad Zwesten 1 999. This game ill u strates how careful Black has to be against White's 'ii'e2, .id2, .nae1 plan. c4) 1 3 ... e5 (this is certainly more playable here than if the white knight was on d4, but it still makes a rather strange impression) 1 4 ltJdS (14 fS should b e enough for a n edge) 14 ...'ii'd 8 1 5 i.c3 (or 1 5 fS 0-0 and Black can aim for ... b6-b5!) 1 5 ...0-0 1 6 lbd2 exf4 1 7 nxf4 lbes 1 8 ltJ£3 l2Jfd7 1 9 b3 g6 and Black was doing okay in Dunis-Kveinys, Bogny sur Meuse 2002. cS) 13 ... nd8 (this should certainly be compared to note 'b' to Black's 9th move in Game 1 2, where White's knight is on d4) 1 4 no!? ( 1 4 ltJdS? exdS 1 5 exdS 0-0! and there is no lbfS here; 1 4 i.c2!? is also playable) 14 ...g6 (14... 0-0 1 5 nh3 g6 1 6 eS dxeS 1 7 fxeS l2Je8 1 8 Wh 1 ltJg7 1 9 i.f4 ltJcS 20 lbxcS bxcS 21 i.e4, Dobrowolski­ Ostrowski, Krynica 1 997 and 1 4... l2Jc5 1 5 lbxcS dxcS 1 6 nh3 l2Jd7 1 7 eS, Westerinen­ Zagorskis, Yerevan 1 996 were both better for White) 1 5 lbdS!? (obviously this is not forced) 1 5 ... exd5 16 exdS 0-0! (16 ... lbc5 1 7 lbxcS 'ii'x cS+ 1 8 i.e3 'ii'c 7 1 9 i.d4 W £8 20 fS! gS 21 'ii'd2 .ng8 22 l:.xe7! 'ifxe7 23 .ne3 and White is winning) 17 ne3 ltJxdS 1 8 cxdS i.f6 and Black was doing more than okay in J urkovic-Sax, Bizovac 2002. 1 3lDd4 Bringing the knight back into play. 1 3 eS?! dxeS 1 4 fxeS lbxeS! exploits the unde­ fended bishop on d3. 1 3 . . .'ii'c7 1 4 'iti>h 1 I prefer 1 4 lb£3!, which looks a good manoeuvre against ... h7-h5 - sometimes the knight can utilise the gS-square. Palac­ Fercec, Zadar 1 999 continued 1 4... h4?! (14... g6!? looks stronger) 1 5 b4 h3 1 6 g3 lbg4 1 7 lbd4 lbdf6 1 8 lbb3 0-0 1 9 l:tacl l:.ac8 20 bS aS 21 lba4 and Black's slightly awkward knights on f6 and g4 gave White an edge. 1 4 . . . g6 1 5 :ae1 h4!

With some ideas of ...ltJhS, castling queenside and, in an ideal world, even a possible ...lbg3 + . White's king can actually be misplaced on h 1 ! 1 6lDd5? This sacrifice simply doesn't work. 16 fS? gxfS 1 7 exfS h3! and 1 6 eS? h3! show the potential in Black's set-up. Best is 1 6 b4!, preventing anything coming to cS and pre­ paring a pawn assault should Black dare to castle queenside. 1 6 . . . exd5 1 7 exd5 lDc5 1 8 i.c2 h3! 1 9 b4 lDxd5 20 cxd5 ..ltxd5 2 1 lDf3 hxg2 + 22 'it>xg2lDe6

Black is a clear pawn up and it's White's king that is more exposed. This adds up to a decisive advantage for Black. 23 i.e4 'ii' b7 24 i.xd5 'ii' x d5 25 'ii' e4 'ii' xe4 26 lbe4 .l:l.c8 27 :c 1 :xc1 28 i.xc 1 'it>d7 29 .l:l.c4 b5 30 :c2 i.f6 3 1 1 03

Sicilian Kan

®g3 ll'lg7 32 lld2 .llc 8 33 i.b2 .ixb2 34 .llx b2 .llc4 35 .lld 2 ll'lh5 + 36 ®f2 .llxb4 37 f5 gxf5 38 ®e3 .lle4 + 39 ®f2 ll'lf6 40 ®g3 l:tg4 + 41 ®h3 ll'le4 42 .llc 2 f6 43 a4 ll'lg5 + 44 ll'lxg5 fxg5 45 axb5 .llh4 + 46 ®g3 axb5 0-1 ,...-----.

Game39 lutz-Milov

European Team Ch., Leon 200 1 1 e4 c5 2 ll'lf3 e6 3 d4 cxd4 4 ll'lxd4 a6 5 .id3 .ic5 6 ll'lb3 .ie7 7 'ii' g4 Perhaps slightly more accurate for White is the move order 7 0-0 d6 8 'i*'g4 and now: a) 8...g6 9 'i*'e2, transposes to Game 40. b) 8... tLlf6 9 'i*'xg7 ngB 1 0 'ii'h 6 looks better for White here than the similar varia­ tion in the note to White's 1 0th move, as Black has expended a tempo on the not-souseful ... d7 -d6. c) 8 ... i. f6 9 'i*'g3! tLlc6 10 tLl1 d2! (targeting d6; tiJ1 d2-c4 looks like a good way of exploiting ... .if6) 10 ... tt:Jge7 (10 ... bS?! 1 1 a4 .U.b8 1 2 axbS axbS 1 3 f4 eS 1 4 tt:Jf3 lUge7 1 S fxeS dxeS 1 6 i.e3 gave White the advantage in Rytshagov-Vehi Bach, Groningen 1 997) 1 1 tLlc4 lUeS 1 2 lUxeS dxeS (12 ... i.xeS?? loses a piece to 13 f4 .if6 14 eS) 1 3 i.e3 (Rytshagov) with a comfortable structural edge for White.

7 . . . ll'lf6 For 7 ...g6, see Game 40. 1 04

7 ...i.f6!? looks slightly better against the immediate 7 'ii'g4 when compared to the note above. 8 0-0 tLlc6 9 f4 d6 and now: a) 1 0 tLlc3 hS 1 1 'i*'f3 tt:Jge7 12 i.e3 'ii'c7 1 3 nad1 nbB! (1 3. .. bS? and 1 3 . .. tLlb4?! are both answered by 1 4 eS!) 14 a4 b6 1 S nd2 (Am.Rodriguez-Vehi Bach, Manresa 1 997) and now Rodriguez's suggestion of 1 S ... i.xc3 1 6 bxc3 i.b7 doesn't look bad for Black. b) 10 tiJ 1 d2!? tt:Jge7 1 1 'i1i>h1 tt:Jg6 1 2 a4 b6 1 3 tLlc4 0-0 14 i.e3 nbB 1 S .U.ad1 bS (1S ... dS 1 6 eS i.e7 17 tLlcd2 tLlb4 18 nf3 was better for White in Van Hassterd­ Spassov, Badalona 1997) 1 6 axbS axbS 17 tLla3 (17 tt:Jxd6? 'ifxd6 1 8 eS tLlcxeS 1 9 fxeS 'ifxeS leaves White a pawn down) 1 7 ... i.xb2 1 8 tLlxbS tLlb4 1 9 'i*'e2 and I prefer White. 8 'ii' x g7 l:.g8 9 'ii' h 6 ll'lc6 9 ... .U.xg2? 10 eS! is strong, as 10 ... tt:Jg4? loses to 1 1 'i*'h3. 1 0 ll'lc3!? Recently white players have been es­ chewing the chance to keep the extra pawn with 1 0 0-0. One older example is 1 0...tLleS 1 1 tLlc3 bS 12 f3 b4 1 3 tLld1 dS 1 4 'i*'e3 i.b7 1 S tt:Jf2 and Black has reasonable compensation in Psakhis-Kurajica, Cap d'Agde 1 994. 1 0 . . Jbg2

1 1 'ii' h 3 An important alternative is 1 1 i.d2 and

5 i.. d 3 i.. c 5

now: a) 1 1 ....:lxf2 1 2 'iVh3 .:lxd2 1 3 lL\xd2 doesn't give Black enough compensation for the exchange. b) In my opinion, the sequence 1 1 ...d6 1 2 0-0-0 .:lx£2! (12....:lg6 1 3 'ii'e3 lL\g4 1 4 'iVe2 lL\ge5 1 5 f4 lL\xd3+ 1 6 'iVxd3 b5 1 7 f5 was very unpleasant for Black in Hernandez­ Kurajica, Las Palrnas 1 995) 1 3 l:thg1 lL\e5 14 �e2 �f8 1 5 'iVh4 has been unfairly con­ demned as Black has the vital resource 15 ... l:txe2! 16 lL\xe2 lL\f3 17 'iVf4 lL\xg1 1 8 l:txg1 ..id7 after which White i s struggling to show anything for the pawn. 1 1 . . ..l:!.g8 Kurajica's latest word on this line is 1 1 ....:lg4!? although 12 f4 e5 1 3 lL\d5 1l2-1l2 Milu-Kurajica, Pula 2001 was hardly a test! 1 2 i.d2 12 ..ie3!? looks more aggressive: a) 1 2 ... d5 13 exd5 with a further split: a 1) 1 3 ... lL\b4 14 0-0-0 lL\xd3+ 1 5 l:txd3 exd5 (15 ...e5 16 'iVfl ..if5 1 7 l:td1) 16 1Vf3 ..ig4 17 'iVf4 l:tc8 18 l:te 1 was obviously good for White in Van der Wiel-Kurajica, European Team Ch., Pula 1 997. a2) 13 ...exd5!? 14 ..tf5 (14 'iVfl ? d4!) 14 ... ..ib4 1 5 0-0-0 ..ixc3 1 6 bxc3 lL\e7 17 ..ixc8 'iVxc8 1 8 'ii'h4! and Black's dark squares look vulnerable. b) Van der Wiel suggests 12 ... d6 is just slightly better for White. I put forward 1 3 0-0-0 e 5 1 4 'ii'fl ..ie6 1 5 ..ic4 to back up this view. 1 2 . . . d5 1 3 exd5 13 0-0-0?! d4 14 lL\e2 e5, as in Ara.Minasian-Milov, Batumi 2002, is not a real test of Black's opening. This does prove, though, that Milov was prepared to give this line another go after his game with Lutz. 1 3 . . . exd5 After 13 ... lL\b4?! 14 0-0-0 lL\xd3+ 1 5 'iVxd3 exd5 1 6 l:thg1 Black's lack o f devel­ opment is starting to show. 1 4 �f 1

White avoids the trap 1 4 ..if5? d4! 1 5 ..ixc8 (1 5 lL\e2?? 'iVd5!) 1 5...'ii'xc8 1 6 'iVxc8+ .U.xc8 1 7 lL\e2 lL\b4, when Black is doing very well. 1 4 . . . i.e6 1 5 0-0-0 �c7 1 6 tt:le2 0-0-0 1 7 ttJed4 ttJxd4 1 8 ttJxd4

1 8 ... i.c5 ! Giving up the bishop pau In order to strengthen the pawn structure. I believe that Black has equalised here . 1 9 tt:lxe6 fxe6 20 f3 'it>b8 2 1 �h3 e5 22 .!:!.he 1 .l:!.de8 23 �h4 �b6 24 �h6 e4 25 fxe4 dxe4 26 i.f4 + 'it>a7 27 i.c4 .l:!.g2 28 l:te2 .l:.xe2 29 i.xe2 e3 30 c3 .l:.e6 3 1 �g5 h 6 3 2 �f5 tt:le4 3 3 .l:!.d3 lLld2 34 .l:.d7 .l:!.e4? Allowing a stunning combination. 35 .l:!.xb7 + ! 'it>xb7? 35 ...1Vxb7 36 'iVxc5+ is good for White but Black is obviously still in the game. 36 �d7 + 'it>a8 37 �c8 + 'it>a7 38 i.b8 + !! 1 -0

Game 40 Torres-Smirin

New York 1998 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 ttJxd4 a6 5 i.d3 i.c5 6 tt:lb3 i.e7 7 �g4 g6 The most common reaction. The ques­ tion is whether the 'extra' move ... g7 -g6 will be a strength or a weakness. 8 �e2 105

Sicilia n Kan

The queen has done its 'job' of provok­ ing ... g7-g6 and now returns to a more fa­ miliar square. White can also wait a bit be­ fore doing this, for example, 8 0-0 d6 9l"Llc3 and now: a) 9 ...l"Llf6?! is too rushed: 10 'ii'e 2l"Llbd7 1 1 ..i.h6! and it's White who has benefited from the ... g7-g6 advance. The rule of thumb is that Black should wait for White to play £2-f4 before he plays ...l"Llf6. b) 9 ...l"Lld7! and now White should trans­ pose to the text with 10 'ii'e 2 as after 10 f4 l"Llgf6 1 1 'ii'e2 White no longer has the op­ tion of ..i.h6. 8 . d6 9 0-0 White normally castles kingside quite early on. One exception was 9 l"Llc3 l"Lld7 1 0 f4 "ilic7 1 1 ..i.d2l"Llgf6 1 2 g4 b 5 1 3 a 3 ..i.b7 14 g5 l"Llh5 15 :n l"Llc5 16 l"Llxc5 dxc5 and Black had no problems in Gofshtein-Glek, Senden 1 997. 9 . . .lLld7 Black keeps developing in typical Kan fashion. Notice, though, that 9 ...l"Llf6?! is still answered by 10 ..i.h6!.

C. Hansen-Akesson, Reykjavik 1 998. b) 12 ... b6 with a further split: b 1) 1 2 i.e3 ..i.b7 13 .:.acl l"Llgf6 (Black could delay this knight move with 1 3 . .. .:.c8!?) 14 ..i.h6 l"Lle5 1 5 h3 l"Llfd7 (15 ... .:.g8!?) 1 6 l"Lld2 g5 1 7 f4 gxf4 1 8 ..i.xf4 h5 1 9 l"Llf3 h4 20 'it>h 1 i.f6 21 ..i.b 1 'it>e7 with an unclear position in Agopov­ Kveinys, Jyvaskyla 2001 . b2) 1 2 f4 and now:

. .

b21) 1 2... i.b7 (this allows White a quick assault on the e6-pawn) 13 f5! l"Llgf6 (or 1 3. ..l"Lle5 1 4 fxe6 fxe6 1 5l"Lld4!) 1 4 fxe6 fxe6 15 l"Lld4 'ii'c 5 16 ..i.e3 "ilie5 17 l"Llf3 'ii'h 5 1 8 h 3 with some advantage to White, Holzke­ Kveinys, Wattenschied 2000 b22) 12 ...l"Llgf6! (I think this is more ac­ curate): b221) 1 3 i.d2 ..i.b7 14 f5 (or 1 4 .:.ael h5! and now the extra tempo for ...g7-g6 looks useful) 14 ...l"Llc5 1 5 fxe6 fxe6 1 6l"Lld4 'ii'd7 and the strong knight on e5 once again holds together Black's position. b222) 1 3 f5!? looks critical: 1 3 ...l"Lle5 1 4 i.h6l"Llfg4 ( 1 4...l"Llcg4 1 5 i.g7 .:.g8 1 6 ..i.xf6 l"Llxf6 1 7 fxe6 fxe6 1 8 e5! dxe5 1 9 l"Lle4 and 14 ... .:.g8 1 5 fxe6 fxe6 1 6 ..i.g5 are both good for White) 1 5 ..i.g7 ng8 and now: b2221) 16 ..i.xe5l"Llxe5 17 f6 ..i.f8 with a bizarre position which I think is fine for Black. He has a monster of a knight on e5, useful dark-squared control, can possibly castle quecnside and can push the g-pawn ·

An important juncture. White has to de­ cide whether to play c2-c4 or not. 1 0 lLlc3 The main alternative is 10 c4 'ii'c7 1 1 l"Llc3 and now: a) 1 1 ...l"Ll e5 12 f4l"Llxd3 1 3 'ii'x d3l"Llf6 1 4 l"Lld4 ..i.d7 1 5 b 3 l:%.c8 1 6 ..i.b2 0-0 1 7 'it>h 1 .:.fd8 1 8 .I:.ael gave White an edge in

106

5 i. d3 i. c 5 up the board! b2222) 1 6 f6 and now Black can choose between the incredible 1 6 ...lt:lxh2!? 17 �xh2 dS 18 �h1 (18 cxdS lt:lf3+ 1 9 �h3 'ii'h2+ 20 �g4 eS+ 21 �xf3 'ii'f4 is a nice mate) 1 8 ... dxc4 1 9 i.xc4 'ii'xc4 20 it'e3 i.b4 or the simple 1 6 ... i.b7, with both positions looking very murky. 1 o.. 'ii'c7 1 1 i.d2 White can also play 1 1 a4 but in view of the next note this seems an unnecessary precaution. 1 1 ...b6 11 ... bS?! is premarure as it allows White to strike on the queenside with 1 2 a4 b4 1 3 lt:la2 aS 1 4 c 3 bxc3 1 S lt:lxc3 followed by lL\bS. But when ... b7-bS is too risky, Black can often rely on a more 'restrained' fi­ anchetto with ...b7-b6. 1 2 l:tae1 i.b7 .

1 3 f4 Finally White succumbs to this narural move. He could wait one further move with 1 3 �h1 !? and now 1 3 ...lL\eS 1 4 f4lt:lxd3 1 5 cxd3 lLlf6 1 6 fS looks promising for White; as does 1 3 ... bS!? 14 a4 bxa4 1 S lLlxa4lt:lgf6 16 lLlaS. After 1 3 ...lt:lf6 1 4 i.h6 play may continue 14 ... lL\eS 1 S f4 (1 S h3!?) 1 5 ...lt:leg4 16 i.g7 .U.g8 1 7 i.xf6lt:lxf6 1 8 eS ltJhS and I would probably just about side with White here. 1 3 . . . lLlgf6 1 4 e5 lLld5! 1 4...lLlhS 1S exd6 i.xd6 1 6 lt:le4 0-0 17

'ii'f2 i.xe4 1 8 i.xe4 .U.ad8 was equal in Bo­ risek-Korneev, Nova Gorica 2002. I prefer 17 lt:lxd6! 'ii'x d6 18 .U.d 1 , which looks a bit better for White. 1 5 lLlxd5 i.xd5 1 6 exd6 1 6 i.xa6 0-0 gives Black good counter­ play, while 16 ... 'ii'xc2 is also not bad. 16 i.c3 dxeS 17 i.xa6? 0-0 1 8 fxeS .U.xa6! 19 'ii'xa6 i.c4 20 'ii'a4 bS 21 'tWaS 'ii'c 6 22 i.d2 b4 left White struggling to find squares for his queen in Wahls-Smirin, Dresden 1 998. 1 6 . . .i.xd6 1 7 c4 i.b7 18 i.c3 0-0

Black has no problems at all in this posi­ tion. He is fully developed and has no weaknesses. In fact, one could already speak of a slight edge due to White's typically 'weak' king (a constant problem for White after f2-f4 in the Sicilian!). 1 9 'ii'g4 lLlc5 20 lLlxc5 i.xc5 + 21 '1t>h 1 l:tadB 22 l:.d 1 l:.d7 23 'ii'h 3 l:tfdB 24 i.xg6 This tactic fails to some ingenious de­ fence. 24 i.f6 i.e7 25 i.xe7 .U.xe7 leaves Black with a slight edge. 24 . . . i.xg2 + ! 24... fxg6? fails to 2S 'ii'x e6+ .U.f7 26 'ii'e 5!. 25 'ii'x g2 Black is also better after 25 �xg2 'ii'c 6+ 26 l:.f3 fxg6!. 25 ... hxg6 26 .ie5 Or 26 'ii'h 3!? 'ii'c 6+ 27 .U.d5 eS!! (a bril-

10 7

Sicilian Ka n

liant defence) 28 .ixe5 f6 and Black will wind up with the better pawn structure. 26 . . . �d6 27 .l:[xd6 .l:[xd6 28 �h3