Colin Crouch - Why We Lose at Chess.pdf

Colin Crouch

we lose at chess

EVERYMAN CHESS Gloucester Publishers pic www.everymanchess.com

First published in 2010 by Gloucester Publi shers pic (formerly Everyman Publi sh ers pic ) , Northburgh House, 10 Northburgh Street, London EC1V OAT Copyright © 2010 Colin Crouch The right of Colin Crouch to be identified as the author of thi s work h as been as­ serted in accordance with the Copyrights, Designs and Patents Act 1988. All rights reserved. No part of thi s publication m ay be reproduced, stored in a retrieval system or transmitted in any form or by any m ean s, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publish er. British Libra ry Cataloguing-in-Publication Data A catalogue record for thi s book is avail able from the British Library. I S B N : 978 1 8 5 744 6 3 6 4 Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 0643 7-0480. All other sales enquiries should be directed to Everym an Chess, Northburgh House, 10 Northburgh Street, London EC1V OAT tel : 020 7 2 5 3 7887 fax: 020 7490 3 708 email : info@everym anchess.com ; website: www.everym anchess.com Everyman is the registered trade m ark of Random House Inc. and is used in this work under licence from Random House Inc.

Everyman Chess Series Chief advisor: Byron J acobs Commissioning editor: John Emms Assi stant editor: Rich ard Palliser Typeset and edited by First Rank Publishing, Brighton . Cover design by Horatio Monteverde. Printed and bound in the US by Versa Press.

Co n t e n t s

Preface

5

Introduction

7

Test One

17

Test Two

25

Test Three

33

Test Four

42

Test Five

53

Test Six

62

Test Seven

72

Test Eight

88

Test Nine

101

Test Ten

115

Test Eleven

124

Test Twelve

136

Test Thirteen

145

Test Fourteen

155

Test Fifteen

165

Your Move?

174

Preface

While writing up reports for various m ag azines, I h ave been thinking about the idea of 'ordin ary chess', of g ames which are not technically perfect, but m ay still be of interest to players, spectators, and hopefully to readers. For this to work well, the writer h as to take the annotation s seriously. The idea i s that while the g ame is interesting anyway, just think of wh at spectacul ar ideas might h ave been thought of if the pl ayer could h ave found the occasional improvem ent. Often in chess, bril­ liancy is just around the corner. I h ave used a similar perspective in this book, but with a different, almost op­ posite perspective. I am writing up 'ordin ary games', my own, with th e thought of systematically going through them, spotting any mistakes of my own (and there are several), and finding better moves. I am aiming to find ways of cutting out mi stakes, thereby improving both my pl ay and th at of the reader. Many g ames h ave been played in local and national league events, and I dedi ­ cate this book to those who continue to keep chess clubs going, in wh at is often quite a difficult time. These days I am cautious about playing in long tourna­ ments, and al so quickpl ay tournaments, sometimes travelling from one end of Britain to the other. It is good to play in my local club, Harrow, where there are often fifty chess pl ayers in a single evening, sometimes close to sixty if there are visiting team s. My thanks to colleagues. Colin Crouch, Harrow Weald, April 2010 5

I n t rod uct i o n

Thi s i s a book of my own g ames. It i s definitely not a compilation of my best wins, attempting to impress the reader. On the contrary, what I am trying to do i s to identify all my serious mistakes over a period of several months. I am not quite sure whether thi s exercise has been tried in public. Attempts h ave certainly been m ade to an alyse the losses of great World Champion s, such as Capabl anca and Fisch er, but there seem s to be very little publi shed autobiographical work of a pl ayer's own losses. We can feel sure th at strong grandm asters will h ave an alysed their games in depth, in order to examine any weaknesses in their own play. It is a matter of sur­ vival at top level . If you do not find your own weaknesses, you opponent will be more than h appy to demon strate wh at you h ave done wrong . Of the earlier World Champion s, one can imagine that Mikh ail Botvinnik would have been extrem ely methodical in going through his post-mortem an alysi s, uncovering both his own mistakes and his opponent's mistakes, and learning from all this. It is enough to rem ember that he was World Ch ampion from 1948 to 1963, a formidable stretch, and th at while he lost m atches again st Smyslov and Tal, he successfully won the return m atches. He was al so a great teacher, and a pioneer of ideas in computer ch ess. At lower level s, one might argue th at games at purely am ateur stren gth might be of only minor interest, since there are often many mistakes, but a pl ayer might not understand why some moves are weak, and other moves are better. Of course, in saying this, my aim is not to try to condemn am ateur chess. I am hoping to pro­ voke interest into trying to encourage ordinary am ateur pl ayers and ambitious

7

Why we L o s e a t C h e s s young players t o think about how t o pl ay better. I am somewhere between the amateur level and the strong grandmaster level . I am an International Master, with a good understanding of the basic ideas of chess, but also the ability to m ake serious mistakes over the board, even ag ainst much weaker opponents. A Dutch 1 M once characterized me as "a chess artist, rather than a practical chess player", a reasonable comment, but it would be nice to discover how to be practical . Possibly in writing this book, it would h elp me to be more practical . But is thi s an excuse, or is it more a case th at while I h ave learnt how to write a decent game of chess, I h ave not really learnt how to play a good game of chess? In term s of life and death in chess, as player and writer, there is something even more important to me. It is a m atter of chance that I am still alive, in 2010, rather than dead. It is a matter of chance th at I am merely partially sighted, rather than blind. It was going to take a lot of h ard work to recover after my stroke in 2004, and I still h ave only partially recovered. I could not read for some time, al­ though now I can read slowly, preferably on large print. Fortun ately, I could see a chess board, just about, and I could therefore play chess after my stroke. My thought processes were still slow after brain damage, and at first I felt doubtful as to whether, if I pl ayed, I could play at over F I D E Elo 2000 strength (about 175 ECF rating). My memory was h owever largely intact, al­ though it was going to be a lot of time before I could connect different thoughts. Aphasia i s still a problem. I can understand wh at i s being said by others, but I could not always string the words together when trying to read. I needed to keep my mind active, and learn to think con structively again, in chess or in anything el se. If anything, my thought processes became much more focussed, as I felt th at I could not waste time. It is difficult enough that I found I h ad to take naps in the afternoon, and th at my thoughts were no longer able to fizz. I needed to think carefully about wh at to do next. Chess was by now much more important in my life, even though I was playing much less. I could no longer think in term s of playing lots of weekenders all over the country (Scotl and and Wales, as well as England), and I h ave become increas­ ingly reluctant to pl ay two g ames in a day. My g ames mentioned in this book, at Bury St Edmunds and Kidlington, give good examples as to why. Creativity i s use­ ful, but if you want to be successful, you need to focus on good technique, and you h ave to respect tiredness. I still wanted to show I could play good chess. Playing a standard nine-round tourn ament was slightly beyond what I was capable of doing with comfort, but m aybe in time I could try thi s agai n ? I h ave to admit th at almost five years after

8

In t ro d u c t i o n m y stroke, I still h ave not summoned m y courage i n playing a nine-rounder, apart from pl aying the occasional Braille event, and a small Middlesex versus junior in­ tern ation al in London . I got too tired. I al so wanted to get back into writing, and for a while I did not even go through my own g ames afterwards. It was so embarrassin g . I had to tell myself th at thi s was t h e result o f tiredness, because o f concentrating o n my books, rather th an a sudden deterioration in my brain. At the time, I was working on my book on Tal, Kasparov and Stein, and I felt confident that while I probably did not analyse com­ pletely accurately in any position (who can ?), I was not yet gaga. So I continued writing, while h oping th at my pl aying strength could improve again. I h ad a g ap in my calendar in 2006/2007, before attempting serious chess analysis on top grandmasters, and in this g ap I was writing up my own g ames, maybe for publication, but primarily for my own interest, and learning again h ow to play good chess, and how to write. Essentially the result is this book, although I was able to revise my comments in 2009, not just because my earlier play and un­ derstanding might not h ave been as good as I would h ave liked, but al so, m ore importantly, because thi s was only an initial draft. Back to Playing Chess

Clearly I wanted to play chess again, but I was not seriously out of touch enough to think that I would be able to reach my peak in chess. My h ope was th at I could still play chess, and not decline too fast. I h ad in fact lost almost a hundred Elo points just before my stroke, and this was at the time a mystery for me. Now, though, it is all very clear. There was accumulating dam age to my brain before the stroke. Even so, I wanted to show that with con structive thinking I could recover most of my peak, despite the slowing down of age, and other problem s. I am not too surprised th at I h ave not fully achieved thi s yet. Other readers, looking to improve their chess, will inevitably be thinking in dif­ ferent ways. In particular, the young player, h aving reached a degree of experi­ ence, will calcul ate quickly, and learn openings speedily, but will not yet h ave the detailed knowledge of positional accuracy. These days I would not be able to calcu­ late lines a dozen moves deep, with sub-variations, but an ambitious teenager would see thi s as the core to chess improvement. If you can calculate quickly, when your opponent can calculate less quickly or deeply, you h ave a clear practical edge. For those over thirty, the player will h ave to modify thought processes. The general procedure would tend to be that, now you cannot calcul ate everything, it is best to use your knowledge and experience to cut out extraneous thought proc-

9

Why we L o s e a t C h e s s esses. For me, I h ave been forced t o take a slightly more extreme approach, as brain dam age means th at I am not able to think quickly. This is recognized in this book. I h ave tended much more th an before to cut down the number of moves to try to analyse, and concentrate instead on thinking about just a few moves, systematically. In other words, do not try to thin k of a long list of possible moves beforehand, trying to assess each position, but start off with a couple of moves to con sider at first, while keeping a quick note of other lines. The idea i s to examine two moves first, normally the most plausibl e moves. N atu­ rally if there is an immediate tactic which must be calcul ated, then examine it. If it can be rejected immediately, move on to other lines. Wh at I am thinking about is, for example, a double piece sacrifice which brings the king out into the open, but can quickly be seen as un sound. If, h owever, there might still be possibilities, then look at it again. You never know, you might h ave a brillian cy. Once we h ave made a quick scan of immediate checks, captures, and other bril­ liancies, al so cutting out any immediate big threat by the opponent, we probably h ave a few moves to be con sidered. These might be attacking moves, or defen sive moves, or positional moves. You need to keep your eyes open . For simplicity, in this test book, we give three alternatives; move A, which I am regarding as the most obvious, then move B, the m ain alternative. If there are two moves to be con sidered, and other moves are irrelevant, that makes life simple. One must, of course, keep in mind th at there are al so possible altern atives, start­ ing with 'Something Else', move C, and then m aybe D onwards. Even if you feel you cannot an alyse in great depth over the board, you still h ave to m ake a deci sion what you are going to play. U sually it is best to concentrate on the most n atural move, A. If you are confident that it is a good move, and any other move (C) seems senseless, you should play it without spending too much thinking time. If, however, you decide that your initial move is not fully satisfac­ tory - m aybe because you feel th at it is promising, but th at there should be some­ thing better, or maybe because it i s ultim ately bad - then you need to think of al ­ tern atives. Remember that for much of the time, the first move you think of i s often the best. Thi s is because you h ave already been thinking about th at idea when looking at the previous position. Maybe your opponent h as pl ayed the reply th at you h ave expected, or m aybe there was a slightly unexpected altern ative, but your possible reply might well seem good and natural . Or m aybe there might be som ething bet­ ter. Think of choice B next anyway, but remember that you must retain your as­ sessment of the evaluation of choice A. Is your first move ending up as slightly bet-

10

In t ro d u c t i o n ter for you? O r just equal ? Or, perhaps the m ost common assessment, slightly worse, or at least m aking you feeling uncomfortable? Or is your initial thought quite simply bad? Then an alyse B, but with a quick flick though to see whether moves C, D, or maybe even beyond could be worth trying later. As some guidance, if you are thinking of analysing a fourth move, or even beyond, you are at ri sk of entering time trouble, sooner or l ater. If you are juggling six possible moves in a given posi­ tion, you will h ave to calcul ate much more th an three times as much than when there are only two moves to be con sidered. You h ave done your basic calculation s, and you must decide whether A or B i s better, and then, for example, whether B or C i s better, then perh aps B or D. In the fin al stage, when you compare B and E, wh at you would really not h ave wanted in retrospect i s to find that B i s better, and th at in trying to analyse lines D and E, m aybe also F as well, you h ave wasted time on the clock. It is a m atter of judgement to decide whether m oves D and E should be ig­ nored. Maybe D looks interesting at first, but a couple of moves l ater, your pawn structure is sh attered, and you do not think th at it is worth defending the line. On the other h and, E might be genuinely tactically interesting, and requires more thought. We must not forget, either, th at an F try might well be worth examining, even though first tim e round it did not seem so effective. Quite clearly there are difficult deci sion s to be made. It is difficult to generalize on how players should find the best thought processes, and find the best move. The ideal is that a player should be able to calcul ate with complete accuracy, but of course a hum an player cannot achieve this in over-the-board play. If anything, it m akes chess far more interesting if the hum an player h as to find good, or in­ deed excellent, moves without the h elp of a computer. It is an exercise for the mind, and top players should quite properly be accorded great respect when find­ ing accurate play and in spired brilliancies. Most of us h ave great imperfections in our chess, but we do not give up the g ame in respon se to them. We need to develop strategies to find ways of finding good moves when we cannot calcul ate everything, and when we do not h ave full understanding of positional chess. I do not pretend that I h ave found the an swer. All I can do is to try to pinpoint ways in which mistakes, and indeed my own mis­ takes, are m ade. This is a preliminary investigation . I h ave indicated in this book 60 mistakes over several month s that I h ave m ade. The main point is th at most of these mis­ takes are not the result of highly complicated and difficult play. Just because I am a master, I can still pl ay rubbi sh chess. At least half of these mistakes could easily

11

Why we L o s e a t C h e s s h ave been avoided by better thought processes. This is easy enough t o establish, to the extent that in the test positions, A, B, and 'something else' (C}, I give improve­ ments on each of my own games. List of Exercises

1.1 Crouch-Oryakhal (White's 5th) 1.2 Crouch-Radovanovic (White's 6th) 1.3 Nurmohamed-Crouch (Black's 7th) 1.4 Wall-Crouch (B7) 2.1 Buckley-Crouch (B9) 2.2 Crouch-Rose (Wll) 2.3 Wall-Crouch (Bl0) 2.4 Crouch-Radovanovic (W12) 3.1 Nurmohamed-Crouch (B12) 3.2 Crouch-Peacock (B12) 3.3 Morris-Crouch (B12) 3.4 Crouch-Lewis (W13) 4.1 Lauterbach-Crouch (B13) 4.2 Crouch-Gait (W14) 4.3 Sen-Crouch (B14) 4.4 Sowray-Crouch (B14) 5.1 Crouch-Hutchinson (W15) 5.2 Crouch-Granat (W16) 5.3 Crouch-Peacock (W16) 5.4 Hebden-Crouch (B16) 6.1 Morris-Crouch (B17) 6.2 Crouch-Hutchinson (18W) 6.3 Hebden-Crouch (B18) 6.4 Lauterbach-Crouch (B18) 7.1 Crouch-Peacock (W19) 7.2 Crouch-Lewis (W20) 7.3 Sowray-Crouch (B20) 7.4 Crouch-Peacock (B21) 8.1 Buckley-Crouch (B21) 8.2 Pert-Crouch (B21)

8.3 Wall-Crouch (B22) 8.4 Nurmohamed-Crouch (W22) 9.1 Crouch-Cox (W24) 9.2 Crouch-Gait (W2 5 ) 9.3 Crouch-Granat (W25) 9.4 Randall-Crouch (B25) 10.1 Morris-Crouch (B25) 10.2 Nurmohamed-Crouch (B26) 10.3 Lauterbach-Crouch (B26) 10.4 Crouch-Gait (W27) 11.1 Crouch-McKenna (W27) 11.2 Crouch-Okike (W27) 11.3 Buckley-Crouch (B27) 11.4 Pert-Crouch (B27) 12.1 Crouch-Cox (W28) 12.2 Crouch-Roberson (W28) 12.3 Cutmore-Crouch (B30) 12.4 Crouch-McKenna (W3 1) 13.1 Randall-Crouch (B31) 13.2 Gregory-Crouch (B32) 13.3 Crouch-Granat (W34) 13.4 Gregory-Crouch (B34) 14.1 Morris-Crouch (B35) 14.2 Cutmore-Crouch (B35) 14.3 Crouch-Roberson (W37) 14.4 Crouch-McKenna (W38) 15.1 Crouch-Granat (W39) 15.2 Buckley-Crouch (B46) 1 5 . 3 Crouch-McKenna (B48) 15.4 Pert-Crouch (B50)

Thi s table is not merely a brief contents page; it is al so a summ ary of research, and a starting point for further examination . There are 60 identified mistakes in

12

In t ro d u c t i o n thi s book, and no doubt I h ave overlooked some further points. I h ave not bothered giving every single slip in each game if I played particularly awfully, for example, in my horrible game again st McKenna (he h ad so m any chances of beatin g me), or even in the l ater stages again st Pert, when I m ade so m any bad moves around the time control, but arguably the worst of these m i stakes was missing an unexpected ch ance of finding a fortuitous draw in the endgame. Ready to start?

We h ave 60 exercises for the reader to con sider. It i s important to rem em ber that thi s is not a quiz. We are not asking you to try to dig out your memory. More to the point is asking readers to think about new and original positions, to try to find the best move, and to avoid identifiable mistakes. I am setting out my own errors, at m aster level, and you are invited to m ake improvements. Also, at another level, you can think about how you can take full advantage of mistakes by your opponent. Not all mistakes get punish ed. Quite often there could easily be half a dozen slips in normal pl ay. If one of the pl ayers win s quickly, that often mean s th at the opponent's mistake is so serious th at a reason able pl ayer should be able to move quickly. When playing through the g ames, take them seriously, but not excessively so. Trying to analyse in great depth, with the h elp of your own brain and computer suggestion s, can be extremely absorbin g . Som e of the position s in this book a re analysed in great depth, sometimes spilling over extra days. The writer likes to aim for perfection, but of course this does not always happen. The reader might by daunted by the thought that it takes a coupl e of pages of inten se analysis to show th at one interesting move eventually turn s out to be better than another. The point is, though, that over the board one can only see a fraction of what could have h appened, and quite often it is possible to say that a player is "lucky" if the critical move turn s out to be good, and "unlucky" if an obscure move turn s out to be an unexpected refutation . A player who loses m ay feel he is unlucky, but it is still a loss and it is important to cut down your unluckiness by working out how to avoid mistakes. For the reader, think about seeing a new position in a g ame at the board. It i s just a n ordin ary position, so w e are not asking you t o find a brilliant sacrifice. W e are asking you t o find a good, ordinary move, avoiding any pitfalls. At an initial glan ce, you will quickly decide whether you think your position is better, or equal, or worse. Or m aybe you just won't know what is h appening and will need to look further before you can m ake any sort of judgement. Once you have sorted out your background information, you can try to decide wh at your

13

Why we L o s e a t Ch e s s next move is going t o b e . I f there is genuinely only o n e sen sible move t o m ake, you can play it immediately. Most of the time, you h ave to think, which is h ard work. You are being asked to find a good move, and for the first part of the exercise, to make a decision quickly. What would you play if you h ad to do something in the next fifteen seconds? In this exercise, you h ave three option s, although the third option is given as "som ething el se", which might be a choice of h alf a dozen rea­ sonable alternatives, particularly in "quiet" position s. Or, on the other h and, there are no reason able altern atives, and it is really only a choice between A and B. As far as possible, I h ave tried not to give a big clue about the best move. Of the three options given, A, B and C, one is the move I actually pl ayed, but was a mistake. The other suggested option i s a natural move, which m ay h ave been right or wrong . You need t o bear in m i n d th at there i s a third possibility, which might o r might not be good. For the initial assessment, decide which move you would play, and write it down . We are not yet asking you which move you would decide to play in a tour­ nament. You will h ave m ore tim e to think l ater. Wh at you are being asked for now are your first impression s. The next stage is to decide which move you would pl ay in a tournament g ame. If you genuinely think that you h ave decided on your choice after a couple of min­ utes, write down your move. In run-of-th e-mill positions, you would h ave on aver­ age three minutes, or less, to decide anyway. In critical and difficult positions, you will want to think for much longer. It often h appen s that a player m akes moves quickly early on in the g ame, and then suddenly slows down . In a critical position, a player will be aware th at the best move m ay keep an advantage, while a second­ best move leads to only equality, and a worse move, attractive but leading to a tactical oversight, may end up as being bad. Take your time, but do not waste time. The clock ticks, and it does not h elp if you are spending 40 minutes on a move if you then have to m ake your last ten moves before the time control in only one minute. In a tourn ament, at some stage you will h ave to m ake a deci sion, and it i s often best, if not necessarily desirable, to save some time for thinking l ater. Use wh atever time you want in this exerci se, then m ake a deci sion and write your move down . Do not be worried if you h ave ch anged your mind since your initial assessment. It m ay well be th at you h ave given yourself extra time to allow extra clear thinking. Maybe you h ave decided that the initial assessment was wrong, inaccurate, or un subtle, and you h ave corrected your thought processes. A more worrying aspect might well be th at you h ave chosen th e correct move i m ­ m ediately, but after some more thought you h ave introduced a n extra error through over-sophi stication. Indeed, on the day after I finished the initial draft, I

14

In t ro d u c t i o n m ade precisely thi s mistake. Another possibility is th at you chose an inaccuracy, and pl ayed it too quickly. This can be very common in the opening if a player saves time by relying on "natu­ ral moves", and the problems come later. I h ave h ad to remind myself that it is no bad thing to spend ten minutes in the opening, rather th an pl ay the first ten moves in a couple of minutes. After going through the exercises, the next stage i s to read closely through my annotation s, in which I h ave m ade use of computer engines and of course a con­ siderable amount of hindsight. I think I h ave learnt a lot from pl aying through my own g am es, and I hope th at you will learn a lot in playing through them too, poor though they m ay be when judged at the top level . After that, it is up to you to start thinking about your own games, and to work out how to improve your play in later encounters. Good luck!

15

Te st On e 1.1 White to play

1.3 Black to play

A) 5 cxb S 8) 5 Ji.e3 C) Something el se?

A) 7 . Ji.cS 8) 7 ... Ji.e7 C) Something el se?

1.2 White to play

1.4 Black to play

A) 6 e3 8) 6 ttJes C) Something el se?

A) 7 . .f6 8) 7 .. a6 C) Something el se?

..

.

.

17

Why we L o s e a t C h e s s

My opponent, and for a time a club colleague, i s what could be described as a 'n atural ' player, or an 'experimen­ tal ' player. He has shown little interest in the opening, or in studying the end­ g ame, but can at times be dangerous in tactical middlegames, and m any club members h ave suffered quick losses to him. It was disconcerting that I was al ­ ready worse as White, within h alf a dozen moves. 1 d4 e6 2 c4 c6!?

leads to a French Advance where Black h as exch anged prematurely, compared with 1 e4 e6 2 d4 dS 3 eS cS 4 c 3 . After 4 ... exds ! , h owever, even the World Ch ampion of the time could not find a convincing edge in M.Tal -A. Bisguier, Bled 1961. Following S eS tDa6 6 tDC3 tDo 7 tDge2 tDe7 8 tDf4 tDfs Black seem s to h ave comfortably equalized, but Tal l ater ground his opponent down in the endg ame. Thi s was reached via a Caro-Kann move order (1 c4 c6 2 e4 e6 3 d4 dS, etc). So 3 ... dS i s playable, but Qryakh al found something more original . He played 3 a6!? in Basm an styl e, but perh aps even more effective th an the 1 e4 a6 2 d4 b S St G eorge's Variation .

Thi s is by no mean s a bad line for Black, and if I develop with 3 tDc3 or 3 tDf3, Bl ack can quickly move into a Semi-Sl av with 3 ... dS. I do not know whether h e would h ave tried some­ thing el se, although I suspect he would h ave done. My instinctive respon se was 3 e4, so th at if 3 ... dS 4 cxdS cxdS, then S e S

It's not so easy for White to set up any clear advantage, and indeed the computer gives only a plus of '0. 19'. All I can sen sibly do i s develop my pieces, and try to ensure th at I will be able to keep a plus. No heroics. 4 tDf3 This cannot be bad, developing the knight to a solid square. However, 4

Test 1. 1 C.Crouch-J.Oryakhal

H i l l i ngdon League

18

2006

...

Te s t On e a4?, to prevent Black from playing ... bS, would be an over-reaction . White would h ave weakened his b4-square. Black naturally plays 4 ... bS, leading to our first test position .

I pl ayed 5 �e3? (B) rather carelessly, and after S bxc4 6 �xc4 (m aybe 6 e S d S 7 exd6 �xd6, but Bl ack is still equal) 6 ... d S ! 7 exd s cxd S 8 �b3 I started to appreciate that while my pieces were better developed, Black h as the better pawn structure, and should be at least equal . Un accountably Black then tried 8 h6?, with a clear loss of tempo, and then after a few later inaccuracies White was able to play for an edge, and later won . White should instead h ave m ade a clear decision over his c-pawn. I did not enjoy the thought of giving away an extra tempo with the pawn before exchanging, namely S cS d6 6 cxd6 �xd6. After 7 �C2 �b7, however, Black's ... c6 move h as got in the way of his pieces. So s cSI (C) is an improve­ ment, possibly with a slight edge. There's al so S cxbS axb S (A), which I ••.

slightly distrusted at the time, with Black keeping the extra pawn in the centre (the c-pawn, as opposed to White's a-pawn). However, Black h as various slight pawn weaknesses on the queen side, m aking the assessment finely balanced. Bl ack m ay well want to pl ay ... dS, but it needs to be well tim ed. Flipping my assessment during the g ame, I would now prefer S c S (C) to S cxbS (A), but either move would be preferable to S �e3 ? ! (B). The rest of the game is of less inter­ est, from my point of view, as my op­ ponent did not play accurately, and I quickly went from a slight disadvan ­ tage to equality, then an edge and a win, with 9 0-0 iLl f6 10 iLl es �d6 11 f4 0-0 12 iLl d2.

Test 1.2 C.Crouch-J.Radovanovic

London Open

2006

...

One of those slight slips in the opening which i s easily missed - here by both sides. The trouble is that both players will want to play quickly in the opening, to avoid the danger of time pressure later on, but if the players do not slow down as soon as something unusual h appen s, one player will un­ expectedly be worse, even after only a few minutes. Play started with 1 d4 d s 2 c4 c6 3 iLl f3 iLl f6 4 iLl c3 dXc4 5 a4, and then S ... �g4, instead of the usual S ... �fS .

19

Why we L o s e a t C h e s s 7 dxe s 'ilVxd1+ 8 'Dxd1 'Dfxd7 9 .1i.xC4 .1i.xf3 10 gxf3 'Dxes 11 .1i.e2, since Black is slightly ahead in development. If my opponent h ad played the best move, I might in the end h ave tried this line, but it leads to no advantage.

Test 1.3 S.Nurmohamed-C.Crouch

Thames Valley League 2006 Since m y stroke, a coupl e of years earlier, I h ad not re-examined thi s posi­ tion, and was taken slightly by surprise. I played the natural 6 e3? (A), but 6 'Des! (B) is more effective, avoiding the pin . Of course, I was well aware of the knight thrust, but was concerned that I would be caught in a m ain line which I h ad not revised. My opponent played 6 ... e6?, which suggests th at he too was not fully aware of the theory. We will see how play continued in Test 2.4. In stead 6 ... e s ! is fully equal . I am sure that had I quickly skimmed through some likely lines before my move, I would h ave decided that since we are still in standard opening play, 7 .1i.xc4 .1i.xf3 8 gxf3 (8 'ilVxf3 ? ! exd4 loses a pawn, without sufficient compen sa­ tion) 8 ... exd4 9 exd4 should be at least playable for White, and would not h ave an alysed much further. After all , saving five or ten minutes in the opening can often be useful later on. I doubt whether I would h ave wanted to think, in advance, of playing

20

Thi s was the start of a nightmare g ame in which my mind was not in focus, I became dizzy, and m ade several mistakes. Fortun ately I became more alert during the end of the session, woken up by the thought that I could easily lose, and was able to negotiate some defensive tactics when he h ad opened up lines again st my king . There was no possible doubt, h owever, th at during the later stages, h e h ad a clear chance of a win . The opening moves started 1 e 4 c5 2 'Df3 e6 3 d4 cxd4 4 'Dxd4 'Df6 5 'Dd2 (too quiet) 5 ... 'Dc6 6 'D4f3 'ilVC7 7 c3.

Te s t On e White has lost a lot of tim e with his knights. Should Black already be play­ ing for an edge? Thi s seem s a little too early. White has played ineffectively, but he starts off with the advantage of the first move, and it often takes more than a couple of quiet moves by White to allow Black to be already better. Usually the m ost practical approach as Black in such a scen ario i s to pl ay quiet but forceful moves early on, and to al ­ low the opponent to try to prove that he is equal. In stead, I quickly played the naive attack 7 ... i.. C 5?! (A). It's a developing move, but it can h ardly be expected that my opponent will be un able to find a good defence to cover the f2square. As we will see in Test 3.1, a few moves l ater, I soon found out, in the main line and in a few variations, th at the bishop is too exposed, and th at White can create counterpl ay with one of b4, ttJb3, or (after ... dS) a pawn ex­ change on dS, and if Black recaptures with the knight, then ttJe4. Black would not necessarily be worse, but h e has lost his initiative. A reality check h elped by using the computer i s th at the position after 7 c3 is close to equal . The hum an pl ayer might well want to argue that Black still keeps the initiative, though, by playing a Scheveningen set-up, with 7... i..e7 (B), and ... d6, when Black i s solid in his centre a n d his pieces can develop quietly and securely. White's pieces cannot be more effectively

placed here than in standard openings. Possibly White can still equalize, or possibly Black with sen sible play can demon strate an edge. Who knows? What i s more certain th at Black has played better than White in the open ­ ing. So 7 ... iLe7 i s the suggested m ain line, but 7...d6 or 7... dS are fully ac­ ceptable too, under 'something else' (C). We sh all return to this g ame l ater.

Test 1.4 G.Wall-C.Crouch

B ritish League (4NCL)

2007

After 1 e4 c5 2 ttJf3 e6 3 c3 ttJf6 4 e5 ttJd 5 Wall played the slightly unusual 5 ttJa 3 .

Going through with the computer a few days later, I saw that he had tried this before. Maybe there i s a lesson h ere? Or maybe not. If I had taken the opportunity of quickly using the oracle an h our before the game, I could have

21

Why we L o s e a t Ch e s s tried t o work out a possible improve­ ment. U sually I prefer not to bother. Chess is a tiring game, and I prefer not to turn a four-hour session into a five­ hour session, through pre-game analy­ sis. In swiss tournaments or team events, it is not possible to examine in leisure the openings a day before. An extra hour resting in bed, or a gentle stroll is often better th an pre-match preparation. Play continued S ... d6 6 tiJC4 dxes 7 tiJcxes, and here I h ad to decide over the board what to play next.

Here the chance of preparation in advance would possibly h ave h elped me to the extent th at I would not h ave been too anxious of a check on the a4e8 diagonal . Indeed, the computer in­ dicates that thi s check i s not too dan ­ gerous, and that Black could quite sim­ ply h ave developed. I pl ayed 7 ... a6?! (B), aiming to con­ solidate in the centre, but also giving away a tempo with a pawn move. Black is close to equality, nevertheless, as we will see in Test 2 . 3 , but it would h ave

22

been useful to end up with complete equality. 7 .. i.. e 7 (C) is the most natural de­ veloping move, after which I probably would h ave been slightly worried by 8 i.. b S+ tiJd7 9 d4 0-0 10 0-0. White keeps slightly more space in the centre, and probably a small but irritating edge after, for example, 10 ... cxd4 1 1 i.. x d7 tiJxd7 12 'iYxd4. In a 4NCL encounter several years earlier, Wall played more directly in the centre, with 8 d4 0-0 9 i.. C 4 cxd4 10 'i¥xd4 f6 1 1 tiJd3 tiJc6 12 iVe4, and his pieces in the centre were irritatingly effective in G .Wall-N. Davies, Briti sh League 2000. 7... tiJd7 (C) is also playable and n atural, but after 8 tiJxd7 i.xd7 9 tiJes i.. d 6 10 tiJxd7 'i¥xd7 1 1 d4 White i s bet­ ter placed in the centre, even if only slightly. Sometimes it m ay be acceptable for Black, even very early on, to take the initiative and play for an edge, even when there is no obvious serious mis­ take by White. Thi s might well h appen, for example, if the player with White tries slow manoeuvring in the opening, and Black sets up some well-timed counterpl ay. The computer suggestion is 7 .../61 (A), and if 8 i.. b S+ then, as a second choice, 8 ... �e7 ! . I am sure that h ad I been a teenager, I would h ave seen thi s a s the obvious a n d natural line, but as one gets older, the inclination is often to try to cut out the more bizarre ideas, since quite often they turn out to be .

Te s t On e wrong . Strange moves still need to be con sidered, though, as well as n atural moves, and it is best not to eliminate too quickly some prom i sing ideas. If then 9 ct:Jg4 e s 10 ct:Je3 e4 11 ct:Jxds+ 'iVxds 1 2 'iVe2, and Bl ack is happy with his position . Once his king has taken the time to escape to f7 after 12 . . . �e6 (though not immediately 12 . . . 'it>f7 ? ? 13 �C4), his other pieces and pawn s will be more active. White therefore can think about aiming for a sacrificial attack, with 9 d4! ? Then perh aps 9 . . :�as (there are other tries) 10 �d3 ! ? fxes 11 �g s+ 'it>e8 12 ct:Jxes g6 1 3 'ilVf3 'ilVC7 14 0-0-0 (or maybe 14 �bs+ ct:Jc6 15 0-0-0 first), and we reach a familiar sacrifice of knight versus pawn, where Black's king is in the open and his pieces are not yet fully developed.

I leave thi s as an exercise in annota­ tion for the reader, whether with the traditional aid of board and set, or with the h elp of the computer. A possible line is 14 . . . cxd4 ! ? l s l::th e l ttJxc3 16 bxc3 'ilVxC3+ 17 'it>bl 'i!Vb4+

with a perpetual, and thi s i s a natural way of finishing. H owever, h ere 16 . . . �a3+? 17 'it>d2 'ilVxC3+ 18 'it>e2 would h ave been unwise, as White's king i s much safer than Black's, with rooks and minor pieces now being level. Bl ack could try in stead 14 . . . i.d6 1 5 ..ibs+ ct:Jc6, and the position would seem to be m ore or less incalcul able over the board after 16 ..id8 'ilVg 7 17 ttJxc6 ..id7. Try it! H owever, pl ayers are allowed to accept sacrifices, and the simple 16 . . . 'it'xd8 1 7 ttJf7+ 'it'e8 18 ttJxh 8 �C4+ 19 'it>bl �d7 i s good for Black. There are m any other lines to be con sidered from move 9. Earlier, White al so h as 8 'ilVa4+.

This time 8 . . . 'it>e7 ? ! takes things too far, since White has 9 �4!, and if 9 . . . g s 10 ttJg 6+!' So instead 8 . . . ttJd7 9 ttJxd7 iLxd7, and in my previous notes, I suggested that there is "comfortable equality for Black". This is a shorthand way of saying that Black m ay already h ave gone beyond equality, and be

23

Why we L o s e a t C h e s s slightly better, but there i s no real point in analysing every line in depth . H ere, for example, 10 i.bs i.xbs 1 1 'iVxbS+ 'iVd7 12 'iVxd7+ 'it'xd7, and 10 iVe4 i.d6 11 d4 fS 12 'iVe2 cxd4 13 tbxd4 '*'f6 both seem promising for Black. In summ ary, 7 .. .f6 ! seem s the best, or at the very least the most promising,

24

although Black must be aware that extreme tactical complication s m ay arise. With best play, Black should end up with an edge, but we need to re­ m ember that few players can play per­ fect chess in a long and inten se battle, and the likelihood is that the better player will win .

T e s t Two 2.1 Black to play

2.3 Black to play

A) 9 . . cxd4 B) 9 ... ..te7 C) Something else?

A) 10 . 'iVC7 B) 10 ... SLe7 C) Something el se?

2.2 White to play

2.4 White to play

A) 11 'iVe2 B) 11 a3 C) Something el se?

A) 12 1tJxg 6 8) 12 0-0 C) Something else?

.

..

25

Why we L o s e a t C h e s s Here I played 7 c5, and I com­ mented th at "7 ... iLe7 appears to be more common. Then after 8 iLd3 iLxd3 9 'ii'x d3 0-0 White can play 10 e4 c 5 with possibl e inter-tran sposition with my m ain line, although White has an extra recapture on d5 after 11 d5 exd5 12 exd5, with a slight edge for White. My thought with an early ... c5 is th at if White plays e4, Black keeps the option with ... cxd4. It might not be needed, but it is there ! " The computer suggests that there mig ht h ave been several equalizing choices at the diagram position . I chose wh at seem s to me the most natural . Pl ay continued 8 iLd 3 iLxd 3 9 'ii'x d 3. ...

Test 2.1

D. Buckley-C.Crouch British League (4NCL) 2006 Play started with 1 d4 ttJf6 2 c4 e6 3 ttJf3 b6 4 a3 .ib7 5 ttJC3 . Fairly standard so far. The next few comments were written at the time. 5 ttJe4 "Here I realized that I h ad not played this variation since my stroke, and I was hazy with my opening. Never mind, I remembered the knight exchange was the move I had previously played, and I recalled that there was some slight problem (in retrospect, it was that White had too easy a time to keep the draw). I just played a move quickly. 5 ... iLe7 is standard. A quick flick through the next couple of games suggests that 6 'ii'c 2 c5 7 e4 is reasonable, and has been played several times. I looked at this, and felt that maybe White has a slight edge, nothing spectacular, but I was h appy to simplify." 6 ttJxe4 iLxe4 7 e3 ...

We h ave now reached the position in the test. It is so important as Bl ack to find a clear equalizing move in the opening, if given a chance. There i s of­ ten a great temptation for the higher­ rated pl ayer as Bl ack to try to keep the tension alive, but so often the pl ayer with White has the chance to keep a modest edge. After th at, it would take m uch m ore work to try to win a game

26

Te s t Two th an if the position i s equal . Wh atever happen s, the only way of playing for a win is to rely on some sort of mistake from the opponent, landing him or her with a slight inferiority. If the opponent is slightly better, it would take a larger technical error to end up worse, than if the opponent is only equal, and even a microscopic slip becomes more signifi­ cant. In a quiet and accurate position, the lesson would seem to be th at it is generally best to play quiet and accu­ rate moves. 9 cxd4! (A) i s simplest and best, but probably only to a slight extent. Then 10 tDxd4 tDc6 should be equal, and could easily transpose later to the stem game - see Test 8.1. 9 ill. e 7?! (B), as I pl ayed, is more provocative. My notes at the time sug ­ gested th at 10 dS exdS 11 cxdS d6, fol ­ lowed by ... ill.f6, might give White a slight edge, but not of any great sig­ nificance. Looking at the position later, 10 dxcS ! is uncomfortable for Bl ack. If 10 ... bxcS, Black's pawn s are difficult to mobilize, while 10 ... ill. x cs involves a loss of tempo with the bishop. After 11 0-0 Black's bishop will probably h ave to return to e7 at some stage. White is slightly better. In th e opening Bl ack is usually slightly worse, with White h avin g more options thanks to th e advantag e of th e move. It i s often best for Bl ack to try to simplify th e central pawn structure, to try to cut down th e op­ pon ent's option s.

Test 2.2 C.Crouch-M.Rose

Kid l i ngton

2007

We start with standard King's In­ dian play with 1 d4 tDf6 2 c4 g6 3 tDc3 ill. g 7 4 e4 d6 5 tDf3 0-0, then 6 �e3.

.•.

...

Thi s i s not quite th e main line, but could easily tran spose into 6 ill. e 2 eS and then 7 ill. e 3 ! ?, which I h ad tried a few times, just before my stroke. This line is still fashionable, and partly as a result of this, I wanted to switch my pl ay slightly, not engage in opening theory. Apart from anything else, I could barely read in the previous cou­ ple of years. He played 6 e5, and I could then transpose, if I wanted, into main stream theory with 7 ..te2 ! ? Then if 7 ... tDg4 8 ill. g s f6, all of 9 ill.h 4, 9 ill. d 2 and even 9 ill. c l h ave been tried. Ag ain, I could not remember why, three years earlier, I h ad decided which of these was best. Another line, experimented by Bent Larsen many years ago, is 7 dxe s dxe s 8 ...

27

Why we L o s e a t Ch e s s 'iVxd8 'iVxd8 9 CLlds, but the slightly unlikely 9 ...l:1d7 ! equalizes easily. So I tried 7 dS, and for what it is worth, the computer gives this as a reasonable edge for White. Al so, we are starting to go beyond mainstream opening theory. After 7 ... CLla6, it is possible th at my reply, 8 .i.d3, was inaccurate. There are, however, alternatives l ater on which might still give White chances of a slight edge, so perh aps it is too early to give question marks. The bishops on e3 and d3 look good and imposing in the centre, except of course that Black's knights will hit back with ... CLlb4 and ... CLlg4. Black n aturally starts with 8 ... CLlg4, then 9 .i.gs f6.

CLlxe4 ..ixb2 16l!bl fxe4 17 ..ixe4 'iVg 7 18 hxg4 .i.xg4 19 iVC2 .i.f6 20 .i.xf6 l:1xf6 2 1 CLlh 2 .i.h s 2 2l:1b 3 ! ?, keeping an edge. In other words, my opening might h ave been unusual, but it was certainly not daft. There is an impor­ tant distinction between playing a completely bad opening, and making a slip slightly l ater. I played instead the inferior 10 ..id2, which i s pl ausible but no more than equal . It allowed Black's pawn push, 10 ...fS, to come too quickly.

Then I completely mi sjudged the tempo of pl ay, with 11 a3?? (B). A few natural developing moves by White and after 11 CLlcS 12 .i.C2? (White needed to give away the bishop-pair), my position collapsed. He exch anged with 12 ... fxe4, then came 13 CLlxe4?!. Battl e-weary, I now saw the sacrifice on f2, but I could not see any way to avoid it. 1 3 .i.xe4 provides more resistance, but even h ere 1 3 ... CLlxf2 ! 14 'it>xf2 CLlxe4+ 15 CLlxe4 'iVh4+ 16 CLlg 3 (16 �e3 leads into the game) 16 ... e4 will not allow White to l ast for long. •••

H ere I probably should h ave played 10 ..ih4! ?, slowing down the .. .fs push . This move i s well known in analogous positions, with the light-squared bishop on e2 rather than on d3. It would take things too far to analyse thi s in detail, but a possible computer­ generated line run s 10 ... 'iVe8 11 0-0 fs 12 exfs gxfs 13 l:1el e4 14 h 3 'iVg 6 1 5

28

Te s t Two Bl ack exch anged 13 ... ctJxe4 14 ..ixe4 ...

again

with

... and now 14...ctJxf2 !. Perhaps the gentlemanly finish would have been to allow the queen sac­ rifice with 15 ..txf2 'iVh4+ 16 'it>e3 ..ih6+ 17 'it>d3 'iVxe4+ 18 'it>xe4 (18 'it>C3 lasts only a little longer after 18 ... l:1.xf3+ 19 gxf3 'iVd4+ 20 ..tc2 ..if5+ 21 ..tCl 'ivxc4+, and mate next move) 18 ... ..if5 mate.

In stead, I wanted to keep some hopes alive. Not for long, though. After the moves 15 ..ig5 ctJxdl 16 ..ixd8 "Lixb2 17 ..ixe7 l:1.f4 18 ctJd2 ctJxe4 19 ..te2 ctJxd2 20 ..txd2 l:1.f2+ it was time to resign (0-1).

So what should I h ave done at move 1 1 ? It was a matter of pride to try to prove that, a couple of rounds l ater, ag ainst Andrew Lewis, my opening was okay, and that it was only a silly miscal ­ culation o n my part that l e d m e t o a loss. Indeed, I could h ave improved a move earlier, with 10 ..ih4! (in stead of 10 ..id2). Or, of course, I could h ave tried one of the several 11th-move al ­ ternatives. So again st Lewis, I played 1 1 'iie 2 (A), but I l ater lost the thread of this game too, and then lost.

Test 2.3 G.Wall-C.Crouch

B ritish League (4NCL)

2007

Not, this time, a question of sh arp tactics, but rather one of finding the best move order. We resume the Wall­ (rouch g ame (1.4), with 8 d4 ctJd7 9 ctJg5. At the time I felt White's move was probably too direct, and I was more worried about his g aining the bishop-

29

Why we L o s e a t C h e s s pair with 9 tDxd7 i.xd7 1 0 tDes cxd4 (10 ... i.e7 11 �s 0-0 12 dxc S i.xcs 1 3 .i.d3 is promising for White) 11 tDxd7 't!Vxd7. Black has a few local difficulties on the king side, so maybe he can switch to the queen side after 12 'iVxd4 a-a-a ! ? : for example, 13 i.e2 "iJlC7 14 0-0 i.d6 15 h3 i.h 2+ 16 �h l .i.es 17 �a7 llVb8 18 't!Vxb8+ i.xb8. The bishop­ pair is to be respected, but Black's knight, bolstered by his pawn on a cen­ tral square, may h old the bal ance. Play continued in stead with 9 ... tDxes 10 dxes, and now I h ad a choice to make.

only then go ... i.cS : for example, 1 3 ... b S 1 4 .i. e 2 .i. b 7 i s comfortably equal, and indeed White could h ave to face a few problems on the 'Meran diagon als', with open bishops on b7 and cS. In the g ame, play instead continued with 10 ... i.e7 11 �es g6 12 �g4 'ii C 7 13 f4 c4 14 a4!? (preventing ... bS) 14....id7, and it i s clear th at Bl ack i s developing much slower than in the alternative line. Black is still not far away from equality, nevertheless, as we will see in Test 8 .3 . Thi s all might seem very technical to the reader, but the accumul ation of minor technical points may quickly accumul ate to points on the board.

Test 2.4 C.Crouch-J.Radovanovic

London Open

I played 10 ... .ie7?! (B), which i s of course a natural developing move, but there are a coupl e of small problems. First, if Black l ater plays ... 'iVC7, he can­ not cover f7, and second, in delaying the bishop move, h e can try in stead to play ... c4 and then ... .ics, without loss of tempo. lO..:f1c7! (A) is more accurate, and if 11 f4 c4 12 'iVh s g6 1 3 'iVg4 I h ave not committed myself to ... .i.e7. Indeed, I can defend on the queenside first, and

30

2006

We h ave discussed thi s position be­ fore (1.2). Normally Black will h ave played ... i.fs rather than ... .ig4. The g ame continued 7 i.xC4 .i.b4 8

Te s t Two �b3 a s 9 t2:ies � h S 10 t2:id 3, with White having good chances of keeping a slight edge, with attacking flexibility in the centre. Still, 10 ... 'iVb6 keeps play in reasonable bal ance. Black tried instead 10 ... t2:ia6 and I replied 11 t2:if4. 11 �xa6 was tempting, but would not h ave led to much after 11 ... bxa6 12 t2:ixb4 �6 1 3 t2:ica2 t2:ids 14 �d2 .l:[b8. Black played 11 ... �g6, reaching the quiz diagram, and the first serious mis­ take, though there h ave been a few minor slips in the earlier pl ay.

12 t2:ixg6? (A) is a remarkably lazy move, saying in effect th at " I h ave done my job, I 've gained the bishop-pair, now I can rest on my laurels." As play went on though, it became obvious th at White h ad great difficulties on the kin g side, with most of his main pieces being temporarily stuck on the queen­ side, and Bl ack quickly being able to pressurize White's king side. 12 0-0 (B) is the obvious altern ative, not allowing Bl ack the open line for his TOok on the h -file. White safely castles, although Bl ack can set up some slight

pressure with 12 ...�C7 . 12 i..d2!? (C), followed quickly per­ h aps by l:tCl, seems the most accurate. There will be some ten se manoeuvring to follow by both sides. White should be able to keep an edge, but certainly no more than on move 1. In stead, after 12 t2:ixg6 hxg6 13 O-O?! gS! White faces unnecessary problems on the h-file.

The immediate problem i s the g8h 2 diagonal . I played 14 t2:ie2 (14 f4 immediately is probably better) 14 ... �d6 15 f4 ( 1 5 t2:ig 3 at the time seemed to passive to me) lS ... t2:ib4 16 �d2 �C7 17 �e1,

31

Why we L o s e a t C h e s s starting t o set up the barriers o n the king side. The reader will no doubt h ave the impression that I am not h appy with most of my moves in thi s game, but somehow my pieces work together. Radovanovic continued his devel­ opment by bringing his rooks into play immediately with 17 ... 0-0-0?, no doubt thinking about doubling on the h -file, but he could have improved. Indeed, White is under pressure af­ ter 17 .. .'!i:JbdS ! , and if 18 .Jtg 3 tt:Je4. White is so busy, understandably so, in cementing his pawn s on the stonewall dark squares that he h as great diffi­ culty covering the light squares. It's not pleasant. I was fortun ate, after he castled, to

32

find a way of aiming for a quick per­ petual by way of 18 .Jtg3 tt:JhS?! {18 ... gxf4! still leaves Black better} 19 fxgS ! tt:Jxg3 20 hxg3 .Jtxg3 21 l:lxf7 .Jth2+ 22 �h1 {23 �fl? ? �xf7+} 22 ... .Jtg3+ 23 �g1 i.h2+ 24 �h1

and a draw with 24 ...i.g3+ {Vz-Yz}.

Te st T h r e e 3.1 Black to play

3.3 Black to play

A) 12 ... bxC3 B) 12 ... i.b7 C) Something el se?

A) 12 ... d4 B) 12 ... aS C) Something else?

3.2 Wh ite to play

3.4 White to play

A) 1 3 'itf2 B) 13 i.d2 C) Something else?

A) 13 exfs B) 13 0-0-0 C) Something el se?

33

Why we L o s e a t C h e s s

Test 3.1 S.Nurmohamed-C.Crouch

Tha mes Va l l ey League

2006

Readers will h ave recognized the opening moves of the game (from Test 1.3). We start from move 7.

Black has just played the in accurate 7 ... iLc s. Then White pl ays the natural 8 iLd 3. I vaguely con sidered 8 ... liJg4 9 0-0 liJce s 10 liJxe s liJxe s, but this seems anti-positional . After 1 1 iLC2 Bl ack has no attack to carry on with, and White will develop quickly. In stead 8 ... dS, or 8 ... 0-0 9 0-0 first, and then 9 ... dS, is playable, with good equalizing chances, but un exciting af­ ter White's previous slow pl ay. I wanted more, and overpressed with 8 ... bS?!, a second doubtful move. White quietly played 9 0-0. I h ad hoped to provoke him with 9 iLxb S ! ? iLxf2+ 1 0 'It>xf2 'iVb6+ 1 1 liJd4 liJxd4 12 cxd4 'iVxb s 1 3 e S liJds 14 liJe4 0-0 with un clear, and m aybe equal, pl ay, but he was evidently not interested.

34

After he quietly castled, I continued my bad-brain day with 9 ... b4?! . There is no point in opening up lines on the queenside, especially as the queen on C7 i s on an awkward square if White's c-pawn is exposed. 9 ... a6 is more the­ matic, waiting to consolidate the pawn s on the queen side, and develop­ ing the pieces. Nurmoh amed played natural devel­ oping chess, with gain of tempo, with 10 liJb3 iLb6 11 ..id2, and it is now my pawn that is under attack on b4, rather th an mine attacking his. I tried to keep my pawn s together, with 11 ... a S . After 11 ... bxC3 12 ..ixC3 ..ib7 Bl ack's position is playable, but I h ave lost tim e with the b-pawn . White h as a slight edge here. Then 12 l:tC1, adding pressure to the b-pawn .

12 ... bxc3? (A) was another miser­ able move, allowing my opponent to open up lines. A pawn exchange a move earlier might just about h ave been acceptabl e, but now Bl ack has lost a tempo by inserting ... as.

Te s t Th r e e Yes, I can g et tired in the evenings, but I should all the sam e pl ay logically and imaginatively. Th at said, 12 ... .§L.b7? (B) 13 cxb4 axb4 14 �xb4 .l:!xa2 1 5 1i.a3 leaves the rook stranded, and is not to be recomm ended either. In stead 12 .. :Wid6 or 12 .. :Wia7 (C) would h ave stopped the position mov­ ing out of control, although undoubt­ edly White keeps a slight edge. After the exch ange on c3, and 13 1i.xc3 �a7 14 ctJe5 ctJxe5 15 1i.xe5 White clearly h ad a substanti al edge. For some reason, I h ad not expected that after 15 ... �b7, he h ad 16 �f3 ! . I suppose that I h ad vaguely expected th at White would h ave wanted to play eS, and then of course the queen on f3 would h ave self-pinned the pawn, but of course the queen h as itself become highly m obile with, for example, at­ tacks with �g 3, or, m ore immediately, working with 1i.xf6 to win a pawn. I felt I h ad to cover the threat on f6, with 16 ... a4 17 ctJd2 1i.d4. I h ad expected, or perh aps hoped, that White would allow the exch ang e of bishops with 18 �g 3 1i.xe s 19 �xe s 0-0 and perh aps a minimal edge for White, but White sacrifices a pawn in­ stead, with excellent compen sation after is 1i.d6! . It is difficult to say why I did not con sider this sacrifice, which keeps the bl ack king stuck in the cen ­ tre. The pawn dropping on b2 was only of secondary importance. More likely perh aps was th at I was so intent on trying to take control of the long di-

agonal that I mi ssed the significance of another dark-squared diagonal .

Even h ere, I was not yet worried, and in any case all I could do was grab the pawn , lS ... 1i.xb2, and see what was going to h appen . Taking the pawn, with g ain of tempo, is useful, but at least as important from my point of view is th at, with the help of a later ... a3, I can keep the bishop on the long diagonal. I doubt wheth er I genuinely be­ lieved at the time that Black was fully equal . It was more the case th at I knew I h ad done something seriously wrong, but th at it was no tim e to resign, and I was pl aying on as effectively as possi­ ble, and waiting for any slight slip by my opponent. White played vigorously with 19 .l:!C7 .l:!cs 20 .l:!xcS+ 1i.xcs 21 ctJC4 a 3 . Even h ere I felt I had chances o f hold­ ing, and playing for more, and it so happened that I was later able to win after scrambling through some tactics. For the next test (8.4), I will ask how later White should h ave won .

35

Why we L o s e a t Ch e s s

Test 3.2 C.Crouch-M.Peacock

Kid l i ngton

2007

Pl ay started with a sh arp theoretical opening, 1 d4 4:Jf6 2 c4 c5 3 d 5 e6 4 4:JC3 exd 5 5 cxd 5 d6 6 e4 g6 7 f4 Jig7 8 ..ltb5+ 4:Jfd7 9 a4 a6 10 ..ltd3 �h4+.

(Briti sh Ch ampion ship, Hove 1997), a draw which could h ave gone either way, but th at was about it. Then sud­ denly, when I started to resume play, the Modern Benoni seemed fashion­ able ag ain, and I started to recognize that Bl ack's play was again of interest. Certainly my bash-bang attack again st Peacock was less than productive. I played 11 xg 2. H e pl ayed in stead 11 'iVb1?! and I began to sense that I could pl ay for rather more th an a quick h alf point. His c-pawn was weak, and so I played 11 ... 'iVc8 12 a3 tiJa6.

Test 4.3 S.Sen-C.Crouch

London Open

2006

The opening is reasonably well known, an English with 1 tiJf3 tiJf6 2 c4 b6 3 g3 �b7 4 �g2 c5 5 0-0 e6 6 tiJc3

48

H ere my opponent thought for a long time, and I started to appreci ate that h e could already be in difficulties. H e g ave up a pawn with 13 i..f4?! to activate his pieces. While h e was thinking, I h ad ex-

Tes t Fo u r pected 1 3 'ilVa2 ! , keeping the pawn on c4 safe, and al so the queen . Thi s should end up in symm etrical equality ag ain, after 1 3 ... dS 14 cxdS tiJxdS 1 5 j.,d2 tiJxc3 16 j.,xC3 j.,dS 17 � 1 tiJcS . There are other possible quiet alternatives, such as 1 3 j.,g s h6 14 j.,xf6 �xf6 1 5 'ilVc2 lId8, m aybe slightly favouring Black (one usually wants to keep the bishop-pair in an open position), or 1 3 'ilVd3 tiJ c S 1 4 'ilVc2 tiJce4 1 5 b3 dS 1 6 tiJxe4 tiJxe4 17 j.,b2 dXc4 18 bXc4 tiJcs, but again White has to work h ard to gain full equality.

White pl ayed 14 :C1, leading to the quiz position . The question i s one of careful and accurate defence, and al as my calculation skills soon folded badly.

Black's problem is that his queen i s under pressure, with no quick escape from the c-file. He needs to sort this one out as soon as possible. 14...'ilVb3?? (C), the only immediate escape from this file, is worse than useless, as 1 5 tiJd2 leaves the queen trapped. 14 :WicB?! (B) is just about playabl e, but after 15 tiJdS tiJxdS (lS ... 'ilVd8 16 tiJxf6+ j.,xf6 17 tiJg 5 j.,xg 5 18 j.,xb7 leaves White m aterial up) 16 lIxc8 J:!.xc8 17 j.,d2 White's queen slightly outweighs Black's rook, knight and pawn . Thi s is, of course, the first line for Black to be con sidered while analys­ ing, but it should only be played in emergency, and only if any other move is even worse. 14 ... tiJb8!? (C) i s almost the last non­ blundering move that a player might expect, and this retreat may be possi­ ble, but is unlikely to be the best. If 15 b4, Bl ack must avoid l S ... dS?? 16 tiJd2 ••

Sen played imaginatively, and as the continuation shows, he an alysed better th an me, but if I h ad analysed well myself, I would h ave kept on top. 13 ...'ilVxc4! is correct! I remembered my recent games ag ain st H ebden and David Buckley (Tests S A and 8.1), and reminded myself that pawn snatches are often vital in chess, and th at failing to taking the pawn is often a m ajor concession . This move i s good. It was on ly the follow-up that was poor. 13 ... lId8 is equal, but nothing more.

49

Why w e L o s e a t C h e s s �a6 17 b S �as 1 8 tLib3, and the queen is trapped, in favour of lS ... aS 16 bxas l:txas 17 'iNxb6, which is level . I doubt if many players would want to examine this line over the board, but these days if the computer suggests something, it is worth con sidering. 14 :iYC5! (C) i s better and the cor­ rect choice, pointing the queen in an­ other direction, with the escape line being to either hS or fS .

tIe more with l S .. J:Uc8, and if 16 b4 'iNfs 17 'iNxfs exfs 18 tLid4 �xg 2 19 �xg 2 g 6, when White's queen side h as been weakened a little further th an in the previous line. I played in stead 14 tLi C 5? (A), but thi s is poorly thought through. ..•

..

Then 1 5 tLidS ? ? 'iNxds 16 tLih4 �e4 i s just a n illusion . The somewh at superior 1 5 h 3 ! cuts down Black's queenside option s. Black can then simply pl ay lS .. :�fS 16 'iNxfs exfs 17 tLid4 �xg 2 18 �xg 2 (18 tLixfs �xh 3 favours Black) 18 ... g6, and this i s more than enough t o demon strate that 14 .. :�cS ! i s better th an any alternative. Black is not worse, but could well be better. His extra pawn i s not looking threatening, but looking closer, one sees th at White might h ave a few weaknesses on the c-file and the light squares on the queenside. Possibly Bl ack could even press a lit-

50

Th e computer at first suggested th at thi s was good, so it is not a totally ridiculous idea, but on further analysi s, it then shows the refutation, admit­ tedly far from in stantly, that I mi ssed. My opponent alas saw it, pl aying 15 b4!. If now l S ... tLib3 ? 16 tLies, and White win s m aterial . If l S ... �xf3 16 �xf3 tLib3 1 7 tLidS ! exdS 18 l:txC4 dxc4 19 �xa8 l:txa8 20 l:ta2, and White i s ahead on m aterial, with queen versus two knights and two pawn s. The m aterial di sadvantage is still rel atively slight, and perh aps thi s is the best chance for Black to set up a rearguard action . H e h as an advanced c-pawn which can be protected by an­ other pawn with 20 ... dS, so White h as to work h ard.

Te s t Fo u r In stead, I pl ayed the move I previ­ ously intended, 15 . ..luce4?!. He pl ayed 16 ctJe5, and tactically Black's position is no longer under control .

If, for example, 16 ... ctJxc3 17 ctJxC4 ctJxb1 18 .1i.xb7 l:tab8 19 .1i.g2 dS 20 ctJb2, and Black's knight is hopelessly trapped, much as in the g ame. In stead, 16 .. :iVxC3 17 l:txC3 ctJxC3 18 "YiVd3 .1i.xg 2 19 'it>xg 2 ctJcdS 20 .1i.d2 is the best of a bad batch, with rook, knight and pawn for the queen, theo­ retically level m aterial. Black's minor pieces are, however, poorly coordi­ nated, blocking each other, and poten­ tially under attack from the e4 push . Moreover, Black's extra pawn is ineffec­ tive in attacking term s. I should h ave pl ayed thi s, but my position is not good. 16 ... "YiVd4?! h astened the collapse. After 17 e3 ctJxc3 18 exd4 ctJxb1 19 �xb7 l:tab8 20 ..ig2 my knight i.s again trapped. Afterwards I felt slightly guilty about not resigning quickly ag ain st my young opponent, but ag ain st this,

there was a quick time limit, and h e was already slightly (but not seriously) short of time.

I carried on with 20 ... ctJxa 3 21 l:txa 3 a s 22 ctJd 3 l:tbc8 2 3 l:tac3 l:txc3 24 l:txc3 axb4 25 l:tb3 l:tc8 26 ctJxb4 l:tC4 27 ctJd 3 l:txd4 28 .1i.e3 l:tC4 29 l:txb6, but it was essentially a m atter of time. White's extra bishop, when given the opportu­ nity, would chew up some of Black's rem aining pawns, and I had to give up on move 43 (1-0).

Test 4.4 P Sowray C Crouch .

-

.

London League

2007

give no analysis of the opening moves, just to say instead th at I felt reason ably h appy in th e opening, with 1 e4 c5 2 ctJf3 e6 3 d4 cxd4 4 ctJxd4 ctJf6 5 ctJC3 d6 6 g4 h6 7 h3 a6 8 ..ig2 "YiVC7 9 .1i.e3 ctJbd 7 10 f4 ctJb6 11 "YiVe2 ctJc4 12 0-0-0 .1i.d 7 13 .1i.f2 l:tc8 14 l:td 3 . This h a s been played before, but not very often. Among the 'something else'

51

Why we L o s e a t C h e s s options i n this quiz, one of the possi­ bilities here is 'draw agreed, M. Katona­ E. Kovacs, Hung arian Team Champion­ ship 1995'. Thi s does not help us very much, neither in theoreti cal term s nor, since Sowray did not offer a draw, in practical term s. I was on my own, try­ ing to find a decent move.

does not do all that much, and wastes time. Bl ack h as been careful not to h ave played ... Jie7 early on, so that if the centre open s up quickly, jumping with the bishop over the currently blocked d6-pawn, Black has ch ances of g aining a tempo with the bishop to b4 or cS, or even sometimes as far as a3. Indeed, 14 . . eS! (B) is clearly the best, and then 15 fxe s (if 15 lLlfS, then simply lS ... exf4) ls ... dxe s 16 lLlb3 (16 lLlfS 'iVas with advantage to Black) 16 ... Jib4 17 Jiel (17 lLldS lLlxdS 18 exds Jib s fa­ vours Bl ack) 17 ... Jixc3 18 Jixc3 Jibs leaves Bl ack's pieces far better coordi­ nated than White's. In the g ame, Sowray lined up the thematic knight sacrifice on the e-file, with 15 :el!?, as we will see in Test 7 . 3 . 1 5 b3 was al so good. White h as gained, either way, some extra time for his ini­ tiative. However, if I h ad played my best move m ore quickly, I would h ave g ained the initiative myself. .

I pl ayed 14 ... b5?! (A), a natural Sicil­ ian move, but h ere it i s not preci se. In many lines it is an immediate respon se for Bl ack to push th e knight on c3 to a worse square, and in so doing, to sof­ ten the defence of the e4-pawn. Here it

52

T est F ive 5.1 White to play

53 White to play

A) 1 5 .l:tadl B) 1 5 l:tfel C) Something el se?

A) 16 lbxd4 B) 16 .l:tel C) Something el se?

5.2 White to play

5.4 Black to play

A) 16 'it'b3 B) 16 l:td4 C} Something el se?

A) 16 ... it'xb2 B) 16 ... l:td8 C) Something else?

53

Why we L o s e a t C h e s s

Test 5.1 C.Crouch-N.Hutchinson

B u ry St. Ed m u nds 2006 1 d4 lLlf6 2 c4 e6 3 lLlC3 �b4 4 e3 0-0 S �d3 cS 6 lLlge2 At this stage, I was not really up to playing standard theoretical chess, and I was often worried that I h ad quite simply forgotten much of my opening knowledge. Here I h ad played this move a few times, a few years ago, but I was aware th at I h ad decided that I was planning to abandon this line, with no real chance of playing for an edge. 5 lLlf3 would h ave given more option s. Play continued, with approximate equality, after 6 ... cxd4 7 exd4 dS 8 0-0 dxc4 9 �xc4 a6 10 a3 ..td6 11 h 3 . White avoided 11 �g 5 ? �xh 2+, but it is clear th at White will sooner or later want to play �g 5 . Bl ack now used the queen side long diagonal with 11 ... bS 12 �a2 �b7 13 ..tgs, and h e continued his develop­ ment with 13 ... lLlbd7 14 'iWd 3 ::I.c8.

54

Now I h ad to decide wh at to do with my rooks. Assuming th at White does not want to block the other rook, the basic choice here i s (from left to right) 15 l1acl, 15 ::I.adl, 15 ::I.fdl or 15 ::I.fel. Quite possibly the other rook will de­ velop in the centre soon after, or quite possibly White h as other plans. The most natural plan i s to develop with ::I.acl and l1fel, but which comes first? In my earlier notes, I suggested 15 'g,fe1 (B), but there is a slight degree of inflexibility in th at if White were quickly to play �bl (for in stance, in one line the computer suggests 15 ... lLlb6 16 �bl ! ?), then the rook on al h as been blocked out of play. Probably thi s will not be a significant defect, and White still keeps a slight edge, but even so, if given the option, it is best to keep flexibility. Th erefore 15 ::I.ac1! (C) is better, and Bl ack still h as to demon strate complete equality. Moving one of the rooks to dl is positionally unin spiring if the pawn centre is closed. Thus I was not im­ pressed with 15 ::I.fd1 . It i s now gener­ ally accepted th at the defence of the isol ated d-pawn should be covered lightly, so th at the other pieces remain active. I decided to play the other rook to dl, is ::I.adl?! (A), with the intention in the m ain line of tactical play with the pawn to d5. It i s imaginative, but as so often with imaginative tactical ideas, a hidden resource m ay suddenly emerge

Te s t F i v e for the defender. In Test 6.2 we sh all see wh at h appen s l ater. In the m ean­ time, it might be of interest to the reader to con sider wh at the pl ans and counter-plans would be after lS .. :�C7 16 f4.

Test 5.2 C.Crouch-R.Granat

Britis h League (4NCL)

2007

Black pl ayed a slightly unusual, but playable, opening with 1 d4 d s 2 c4 .Us 3 cxd s iLxbl.

I immediately pl ayed 4 :xbl?!, and then almost immediately regretted it. 4 "iVa4+! is more accurate, and only then 4 ... c6 5 :xbl. 4 ... b S ? 5 "iVxb S+ c6 6 dxc6, of course, gives nothing. I h ave decided not to give the dia­ gram position as one of the tests, be­ cause if the reader i s already alerted that something is amiss, he or she will look more carefully and find the more accurate move. The problem for the player over the board is to spot

whether t h e obvious move, here a re­ capture, is automatically the best. After 4 ..."iVxd s, I cautiously pl ayed S a 3 . There are a few g ames, I l ater found out, in which White collapsed after 5 "iVa4+ lLlc6, and if 6 lLlf3 eS 7 dxe s iLb4+ 8 iL d 2 "iVe4! (gaining a tempo with the attack on the rook) 9 iYdl 0-0-0. It was not difficult for me to work thi s line out in advance, but after my quieter move, White's advantage of the first move h as gone. Black was not interested now in hit­ ting the bl-rook, with ...iYa2 or ... iYe4, and in stead developed quickly with s ... lLlc6 6 e3 eS. I was relieved th at I was able to exch ange the queen s with 7 dxes iYxdl+ 8 \t>xdl 0-0-0+.

I very vaguely remembered that I had reached this position before, again with the slight error on move 4. On checking up later, I found the game C.Crouch-I.Sakovich, Decin 1996, with probable equality after 9 i.. d 2 lLlxes 10 lLlf3 iLd6 11 lLlxes iLxes 12 iLe2 lLlf6 13 iLf3 l:the8 14 \t>e2. I could try to pretend that my bishop-pair leaves me slightly

55

Why we L o s e a t C h e s s better, but i n fact i t i s not s o significant. I later won, but this was far from forced. Several years later, I played 9 'it>c2, then came 9 . .4:Jxes, with possible threats with ... 4:Jg4. This was difficult to assess. My general preference in such position s would be to prefer the bishop, but in compensation Black is slightly ahead in development. Probably the position remains equal with best play on both sides, but any slight slip on ei­ ther side would start to change the bal­ ance. It is not an 'easy' equality. I played 10 �e2. There h ave been other games in which White h as tried 10 f3 or 10 h 3, but this seem s slow. I considered 10 4:Jf3, but was a bit wor­ ried about 10 ... 4:Jg4, continuing with 11 .tC4 4:Jxf2 12 lIft 4:Je4 13 4:Je5 4:Jh 6 14 4:Jxf7 4:Jxf7 1 5 .l::t xf7. Equal, probably. Black's knight is an irritation, but White has the bishop-pair. Over the next few moves, both play­ ers develop and consolidate with 10 ...4:Jf6 11 4:Jh3 h6 12 b4 .td6 13 .tb2. The computer suggests th at early on Black is slightly better, but th at later th e position is equal . I would suspect th at the l ast few moves were equal throughout. There followed some more quiet m anoeuvring with 13 ... lIhe8 14 lIbdl 4:Jfd7 15 4:Jf4 g6, and we h ave now reached the quiz position. Black's last couple of m oves h ave seemed almost random, at least from White's point of view, and I started to feel more confident. It is often around thi s stage th at a pl ayer might relax his

vigilance. So it proved.

.

56

My assessm ent was th at it was a good time to move one of my pieces, rather than trying to push a king side pawn . 16 'it>b3? (A) looks natural, but I overlooked or underestim ated 16 ... a s! . attacking White's queenside pawn s. I did not like the idea of allowing an i so­ lated pawn on b4, and so played 17 bxa s 4:Jc6 18 4:Jd S . Now Black gains a tempo with check, with 18 ... 4:Jxa s+, as a result of White's unfortunate king m ove. So a retreat, 19 'it>a2 .tes, and Black h as set up a useful passed pawn on the c-file, while also neutralizing White's bishop-pair.

Tes t Five While thi s is far from deci sive, I was starting to feel under pressure, and later m ade mistakes. We shall return to this exercise at a l ater stage (Test 9.3), with an accumulation of minor errors, plus a time trouble blunder, leading to a clear loss. So what possibl e improvements could there be for White ? Thi s all seem s open -ended, with several possible ideas ending up reaching a position of likely equality. One n atural possibility is 16 cj;b1!? (C) 16 ... lt:ib6 17 It:ids It:ibC4 18 i.xe s It:ixe s, equal . At the time, I wanted more. 16 It:idS (C) puts some slight pres­ sure on the opponent, but with quiet play, for example, 16 ... c6, it is unclear that White can achieve anything con­ crete. Maybe the best move is 16 'f1.d4! (B), preparing to play the other rook, either to dl, doubling, or to set up pressure on the c-pawn with :tCl.

There are various quiet moves for Black, m aybe with equality, or maybe

with a fractional edge for White. 16 ... a s would still h ave been possible, as in the game, but after 1 7 bxas It:ic6 18 .l:!.d2 It:ixas 19 :thdl White is ahead on tempi, and possibly with a slight edge. Thus m any of the 'something else' moves are playable, and quite likely to be better than the move I actually played, but my preference is for 16 .l:!.d4.

Test 5.3 C.Crouch-M.Peacock

Kid l i ngton

2007

Returning to a g ame we con sidered back in Test 3.2, here Black played lS ... lt:icS. I pl ayed the natural 16 It:ixd4? (A), but this soon allows Black to equalize with 16 ... lt:ixe4+ 17 itxe4 'iWxe4 18 lt:if3 itg4. We continue this exercise at a later stage - see Test 7.1. 16 l1e1! (B) is more accurate. If 16 ... lt:ixe4+ 17 itxe4 'ilVxe4, I h ad only con sidered 18 'ilVxd4?! 'ilVxd4+ 19 It:ixd4

57

Why we L o s e a t C h e s s tLJd7, with equality, since both d-pawn s drop, with a symmetrical opposite­ coloured bishop position . However, 18 tLJc3 ! substantially improves. After 18 ...�f5 19 �xd4 White's comfortably centralized queen causes problems for Black. I was al so concerned with 16 ... i.. g 4, but 17 1:[a3 tLJxe4+ 18 i.. x e4 �xe4 19 'Yi'xd4 �xd4 20 tLJxd4 leaves White with the more active pieces in a queenless endgame. The best line for Black i s 16 ... �4+! 17 'iit> g l tLJxd3 18 g3 tLJf2 19 gxh4 tLJxdl 20 1:[xdl i.. g 4 21 Wf2 tLJd7 22 ktxd4 f5 .

White is a pawn up, but thi s is only a doubled i solated h -pawn, which h as littl e impact. Black still h as to take som e care in holding the g ame, but after 23 tLJC3 tLJa5 24 1:[a3 fxe4 25 tLJxe4 tLJxe4 26 l:txe4 1:[f5 27 l:tg 3 i.. d l 2 8 1:[g 5 i.. d l 29 i.. d 2 ktC5 h olds the bal ance for Bl ack. At least this way it is Black who h as to keep equality. With my own move for White, I h ad to work h ard myself to achieve the same goal .

58

Test 5.4 M.Hebden-C.Crouch

Metropol ita n Open

2006

The final round of a weekend tour­ nament, with ch ances of equal first, and some prize money, though much less, in real term s, than in the seventies or eighties. I did not relish the thought of play­ ing a second g ame in a day, given the likelihood of dizziness, and tiredness during the end of the session . A pro­ longed g ame ending up in a blitz finish would h ave been a nightmare for me, so I wanted to pl ay quickly, trying to eliminate any elaborate play. My op­ ponent has always believed in playing quickly him self, finding decent moves, if not necessarily the best. He m akes his mistakes, like all of us, but it is often very difficult for his opponent to find any refutation when th e time ticks on . My psychological mistake in this game was to pl ay too quickly against quick but provocative play. 1 d4 tLJf6 2 tLJf3 b6 3 i.. g 5 i.. b 7 4 tLJbd2 c5 Some thirty month s earlier, playing again st the same opponent el sewhere (M.Hebden -C.Crouch, Coventry 2004). I avoided thi s pawn exch ange, playing instead 4 ... e6 5 e4 h 6 6 iLxf6 �xf6 7 iLd3 d6 8 �e2 �e7 9 0-0 g 6 10 e5 i.. g 7 1 1 1:[fel 0-0 12 c3 tLJd7 1 3 iLa6 i.. x a6 14 �xa6 dxe 5 15 tLJxe5 1i.xe 5 16 dxe 5, with perh aps a minimal advantage for

Te s t F i v e White, ending up as a draw. Thinking back to the earlier tour­ nament, it seem s such an incredibly long g ap between 'before' and 'after', with a stroke intervening. It seem s so strange th at after pl aying in Coventry, first equal, then a hundred-mile walk­ ing h oliday, then pl aying another tour­ nament in Oxford, my health was in such a sudden collapse. And indeed the physician s were puzzled too. There were some extrem ely elaborate diag ­ noses, quite often contradictory. I do not want to discuss in great depth wh at was h appening at the time, as this is something of a digression of analysing two g ames of chess. I tried hard to understand what was going on in the thought processes and degree of understanding of the medical doctors, and I was looking for contradictions and possible mistakes. A difficult proc­ ess, but necessary. After all, medics, like chess pl ayers, can often m ake mistakes, and I wanted to assess early on whether a mi stake was being m ade. We can add th at economists and politician s too are highly capable of making serious mistakes, something fairly clear these days for the gen eral public. Does anyone believe th at we are in a stable economy now? Anyway, pl ay carried on with 5 c3 e6 6 e4 h6 7 �xf6 'ilVxf6 8 �d3 cxd4 9 cxd4 CDc6 10 e5 'ilVd8 11 a3 d6 12 'ilVe2 �e7 13 0-0, and if anything, I might well have played better in 2006 th an in 2004.

Thi s h as, o f course, been played be­ fore. In an earlier game, I . Rausis­ J . Plaskett, Port Erin 1998, Plaskett played 13 ... 0-0, not the sort of move that I would generally want to con sider myself, not really liking castling wh en moving the king towards the risk of immediate attack. Black was better, and later won, after 14 'ilVe4 g6 15 'ilVe3 �g 7 16 CDe4 dxe s 17 dxe s 'ilVds 18 liJf6 ? ! 'i!Vb3 19 l1acl l1ad8, but 18 liJO is equal . I chose 13 ... dxe5, and if Plaskett's immediate castling looks to me slightly strange, then my own plan would to m any oth ers seem even more unusual, allowing the possibility of my king h av­ ing to stay in the centre. Quite often it is not a question of whether a player thinks th at one type of move is 'better' or 'worse' than the other, but rather a case of h alf-remembered experiences of encounters from many years ago. I tend not to like going into early mid­ dlegames where my king gets stuck in the corner when I do not h ave extra pieces defending. Moreover, I tend to

59

Why we L o s e a t C h e s s feel uncomfortable about m y king get­ ting forced away from the back ranks. I have h ad a few bad experiences in the past. My plan was 14 dxe5 ttJd4 15 ttJxd4 "iVxd4, so far, equal throughout.

I had expected 16 �b5+ �8 17 b4 reckoning this as a draw, and maybe this is objectively the best. Instead Heb­ den gave away a pawn, with 16 l:i.acl?!. Over the years, indeed over the decades, Hebden has developed the tournament strategy of playing reasonable moves quickly, hoping his moves are okay, and above all, making sure that he does not allow himself to be in time trouble. This can be an effective way of playing, espe­ cially in weekend games, when his op­ ponents get tired and short of time, and he himself remains fresh for the second game of the day. Hebden 's last move was played quickly, and I h ad simply not expected it. Worse, I h ad decided beforeh and that if my opponent was going to play quickly, I would do the same. I did not want to get too tired, especially when

60

getting short of time. The nightm are would h ave been that, as has occasion­ ally h appened, I was to reach a position where he h ad an hour on the clock, and I h ad five minutes for the rest of the game. In better years I could h ave h an­ dled it, but right now, I couldn 't. It did not help that on the Saturday before­ h and, I h ad received a phone message 'urgent' concerning a club event. This was spectacul arly non-urgent, and I should have ignored it. Any m atch cap­ tain will know about this type of stress before a game. So the general picture was th at I was out of sorts, and th at I would h ave been more than h appy, especially as Bl ack, to find a comfortable way of finding quick and steady equality. And maybe thi s was at the root of the psy­ chological error I m ade, quite a com­ mon one. I played 16 l:i.d8? (B), aiming for equality, but ending up worse. ...

White played 17 �b5+ 'iM8 18 l:i.C7, and I was starting to struggle, as we will see further in Test 6.3. U sually I have a

Te s t F i v e counterattacking style of play, not aim­ ing for quick solidity as Black, but in­ stead keeping a sharp eye (these days, alas only my left eye) for any possibility of taking over the initiative after any­ thing that seems slightly inaccurate. In more normal circumstances, I would h ave gobbled the pawn, 16 :�xb2! (A), without much hesitation. .•

Bl ack i s better. T o continue the at­ tack White h as to play 17 .l:!.C7 i.ds 18 �b S+ f8, but his attack seem s specu­ l ative. If, for example, 19 a4 ..td8 20 .l:!.d7 �g s 21 f4 iVd4+ 22 �h l ..txf4, and White cannot go any further. In stead, 19 tf-fcl adds another at­ tacker into play. The computer sug­ gests 19 .. .f6 ! ? and 19 ... ..tcS ! ? as possi­ bilities for an edge, and both moves look reason able enough. Moreover, if I wanted a quiet and safe equality, 19 .. :�ha3 would h ave done. Then 20 .l:!.7c3 (20 .l:!.c8+? ..td8 21 tf.xa8? ? iVXC1+) 20 ... iVb2 (20 ... iVb4?? 21 .l:!.d8+ ..td8 2 2 .l:!.xa8 win s for White, n o w th at Bl ack does not cover the cl-square) 21 .l:!.lc2 iVal+ 2 2 tf.cl iVb2 repeats.

61

Test S i x 6.1 Black to play

63 Black to play

A) 1 7 .. :�e6 B) 17 .. JUe8 C) Something el se?

A) 18 .. :�Vxd2 B) 18 .. :�xb2 C) Something el se?

6.2 White to play

6.4 Black to play

A) 18 4'lxe4 B) 18 dS C) Something el se?

A) 18 ... h 6 B) 18 .. .f6 C) Something el se?

62

Te s t S ix

Test 6.1 C.Morris-C.Crouch

British League (4N C L)

before the time control, before ulti ­ m ately losing. The remaining moves will be con sidered later in Test 10.1.

2006

Black should have played with much more grit, with 1 7 JJj.e8! (B), with the defensive idea of ... ti.Jf8 and ... .1i.. c 8. The other knight, when given the opportu­ nity, can join in with ... ti.JO. White's edge is negligible. I would like to think that had I been in better health, I would have easily seen this idea. The queen exchange was a di saster for Black. Exchanging queens can be useful in defending the king when un­ der attack, but is often less than useless when defending weak pawns. I needed to defend on the kingside with my pieces, rather than exchange my queen. ..

As we saw in Test 3 . 3 , White h as a better pawn structure, and started to set up pressure with his pieces, with 17 'iVfS. Here I pl ayed 17 'iVe6? (A), which I described at the tim e as "a weak and lazy move." Certainly I was feeling very tired that day, and could not even con ­ template, on Remembrance Day, walk­ ing almost next door to the old Coven ­ try Cathedral, heavily bombed during the War. In th e position itself, I was too wor­ ried about pressure on my kingside, and underestim ated my possible queenside pawn weaknesses. After 18 .id3 'iVxfS 19 i.xfs .l:!.ad8 20 .l:!.acl ti.JC7 21 iLxd 7 ! .l:!.xd7 2 2 ti.Ja4 ti.Ja8 my posi­ tion was crumbling. There were twi sts and turn s later on , and I was even briefly able to get back into the game, if temporarily, just ...

Test 6.2 C.Crouch-N.Hutchinson

B u ry St Ed m u nds

2006

Continuing from Test 5.1, my plan is highly ambitious, with thoughts of a pawn breakthrough with either d5 or f5.

63

Why we L o s e a t C h e s s Either i t works or i t doesn't. To the best of my calculating capability, I felt I was doing well, but I missed a resource sev­ eral moves later in a critical position. It turns out that there were good alterna­ tives for Black anyway, so my earlier play was too ambitious.

Black played 16 .. :�c6. It would be prem ature to break the diagonal by force with 17 d s ?, as after 17 ... exds 18 ttJxds i.. c s+ 19 �h l ttJxds 20 i.. x ds 'iVxds 21 'iVxds i.. x ds 22 1:!.xds ttJb6 the pawn structure is roughly symmetrical, but Black's pieces are far better placed, and so he h as an advantage. So 17 1:!.f2.

64

Bl ack could now h ave considered the ultra-solid 17 ... h 6 ! ? 18 i..h 4 1:!.fe8 19 fS eS 20 dxe s i.. x es 21 ttJd4, followed by finding one of the good queen re­ plies. I h ave to admit that White's king­ side looks far too loose, and th at I would h ave needed to work h ard to try to hold the position . Black's kingside is extremely safe, despite White's at­ tempt to start an attack on th at side. There are, as we shall see, two good moves for Bl ack, and therefore I cannot justifiably claim that I am 'unlucky' in h aving chosen the line starting with 15 1:!.adl, and with the unexpected coun­ terattack l ater on . In stead, I misjudged the position, m aking both positional and tactical errors. Black in stead pl ayed the more direct and obvious 17 ... ttJe4. Now I should h ave pl ayed the drawish 18 ttJxe4 (A) 18 ... 'iVxe4 19 'iVxe4 i.. x e4 20 dS i.. x ds 21 i.. x ds exds 22 1:!.xdS 1:!.c6, and quite probably Black would h ave been h appy with a h alf-point. I had prepared, imaginatively but incorrectly, is d 5 ? (B), and play seemed smooth enough for m e after lS... ttJxf2? 19 dxc6 ttJxd 3 20 cxd7 i.. C 5+ 21 �h2 ttJxb2 22 dxcS'iV i.. xcs 2 3 1:!.d2 ttJC4 24 i.. x C4 bXc4 2 5 a4, and I eventually won with my extra knight versus two pawns. While h e was thinking about his taking the knight on f2, I wondered what would h appen after lS ... exd 5, which at first seem s like a blunder. I h ad thought th at everything was

Tes t S ix covered, and I h ad not really con sidered that there could be a danger for m e, but then I saw a possibl e problem just before he m ade a move, and back at home I saw that it would h ave been a major problem, and th at I could h ave lost two games (the other again st Gregory - Test 1 3 . 2 ) on the same birth ­ day.

I could h ave h ad a slalom run with 19 tiJxe4 dxe4 20 "tIYxd6, when I felt I was safe a piece up, but Black has le­ thal counterpl ay with 20 ... e3 ! ! .

"tIYxc6 exf2+ 2 2 �xf2 ..ixc6. If White wants to move the rook in­ stead, the only try i s 21 Itf3 �xd6 22 .l::!. x d6 i.xf3 . Black should win, although it m ay take time, after either 23 gxf3 .l::!. c 2 24 .l::!. x d7 Itxe2 2 5 ..ie7 Itxb2 2 6 ..ixf8 �xf8 2 7 l:!.xf7+ � e 8 28 .l::!.f5 .l::!. x a2 29 .l::!. e 5 + �f7 30 .l::!. x e3 �f6 (level m ate­ rial, but White's king and pawn s are dreadful), or 23 tiJd4 ItC1+ 24 �h 2 ..ih 5 2 5 .l::!. x d7 e2 2 6 tiJxe2 ..ixe2 (exch ange up, and Black can squeeze the bishop with ... ..ic4, after 2 7 ..ie7 .l::!. e 8). But I won the g ame. Was I 'lucky', in the sense th at he missed his chance of winning, or 'unlucky' in th at while I calculated an interesting line in ad­ vance, there was an unclear tactic many moves on ? It depends on the strength s of the two players. At very top grandm aster level, all thi s would h ave been a string of blunders, and of course I am aware of th at. For the time being, I h ad the excuse of illness, and few things can be worse th an brain dam age for a chess pl ayer. But I am starting to TUn out of excuses now ...

Test 6.3 M.Hebden-C.Crouch

Metropol ita n Open

I h ad missed th at ! And of course h e missed it too. White is a clear exch ange down, without compen sation, after 2 1

2006

Despite earlier events (see Test 5 .4), my position should not collapse. Unfor­ tunately i t did. I would h ave suspected that I h ad done something wrong, but I still h ave play, h aving a useful bishop-

65

Why we L o s e a t C h e s s pair, a good open file, and an active queen . That said, I al so h ave obvious disadvantages, with my king side pieces being clogged up, and White h aving a rook on the seventh .

Again, I played too quickly, and m ade the obvious capture, 18 ... "ii'x d2? (A), assuming that I was at least con ­ formably equal . I missed something a few moves along the line. White won after 19 l:[xb7 "ii'x e2 20 �xe2 gs (better is 20 ... g6, but White is still on top after 2 1 b4) 21 l:[c1 l:[d2 22 �f1.

Now I missed a critical pin after 22 ... l:[xb2?? 23 ..ihS!, and I immediately resigned (1-0), in view of 23 ... �C5 24

66

.uxc 5 ! bXc5 25 l:[xb2 . 23 ... ..ixa3 24 ItcS+ �g7 2 5 l:[xf7 m ate i s even quicker. Black has to try to improve. My im­ mediate reaction afterwards was to try to find an improvement with 18 . :Wixb2 (B) 19 Itxb7 'iix d2, swiping the b-pawn (not though 19 ... Itxd2? 20 ItbS+ .lidS 21 ItxdS+, with a cross-pin). White is still much better, h owever, after 20 "ii'x d2 Itxd2 2 1 l:[c1 g5 22 Ikc7 l:[d1+ 2 3 ..itl .lixa3 24 .uxa7 ..i C 5 2 5 .uxf7+ �eS 26 Itfb7. Black i s a tempo down after taking the b-pawn, and White ag ain keeps a strong attack with two rooks and bishop. Here 21 ... ..idS provides more resis­ tance. Then 2 2 .uxa7 g 6 23 a4. .

Can Black hold this? Offh and, I do not know. In practical term s, White would be more th an h appy to carry on trying to chip away for several hours, or, in a quickplay finish (as h ere), wait­ ing for Bl ack's positi on to collapse. Even if Black manages to fin ally completed his development, with ... �g 7 and a rook move, he still has problems with his pawn s on f7 and g6 (also e6, if

Tes t S ix Black's king i s on g 7), and his bishop i s not mobile. This, however, i s a first impression . If Black were to try 2 3 .. J:td5 ! 24 .l::t d 7 (24 f4? ! g5 gives counterplay) 24 .. J:txd7 25 iLxd7, with the first pair of rooks gone, Black increases his possibilities of hold­ ing. He is of course not equal yet. Bl ack can al so try to set up a differ­ ent pawn structure in the endgame, with lB iLxg2!? (C) 19 'it'xg 2 "iVxd2 20 'i¥xd2 l:txd2 2 1 l:!.xa7.

( 2 5 .l::t c 8 'it'g6 i s comfortable for Black) 2 5 ... iLh4 26 l:!.cxb6 l:!.xf2+ 27 'it'g l l:!.a2 28 l:!.xe6 l:!.a2+, with a draw, following a reason able degree of accuracy by Bl ack. Thi s would seem to be the most ac­ curate line for Black, boldly aiming for equality with counterplay, rather than h oping th at the opponent cannot find a way of keeping a slight edge. There i s another way for Black, keeping m aterial on the board with lB i..oB (C) 19 4:Jb1 ! .

Clearly Black will not be able to take the pawn (21 .. J:txb2 ? ? 2 2 l:!.a8+), and so we now h ave a more dyn amic pawn structure, with strength s and weak­ nesses on both sides. White will h ave excellent chances of creating a danger­ ous passed pawn on the queenside, while Bl ack will need to set up coun­ terpl ay on the kingside. It is important to recognize for Black that he must not just sit on the extra pawn on the king­ side, but that he needs to play actively. Therefore 21 ... g 5 22 b4 (saving the pawn, and al so preventing ... iLc5) 22 ... g4! 2 3 .l::tb 7 iLd8 24 .l::t c 1 'it'g 7 2 5 l:!.c6

The knight retreat is unexpected, and indeed it was pointed out to me by computer. The knight soon bounces back though, and after 4:JC3 l ater, Bl ack h as no control of the d5- and e6squares. White i s better, Black still h av­ ing problems with his development. There are several tries for Black here, but none seem s to equalize. For exam ­ ple, Bl ack can start with 19 ... "iVd5 (19 ... iLc5 20 4:JC3 squashes Black's counterpl ay) 20 f3 iLC5+ 2 1 �h 1. Then 2 1 ... iLe3 2 2 "iVxe3 "iVxb5 is an attempt to break the natural course of play, and if the natural 23 l:!.fc1 "iVxb2

...

•••

67

Why we L o s e a t C h e s s 2 4 �f4 g 5 2 5 �4, Black holds the position with 25 .. J:tC7 ! . Once thi s has been appreciated, 23 lIg 1 ! is quickly seen as a good move, and if 23 ... �xb 2 ? (but other moves are not very good) 24 �f4 f5 2 5 � 5 lIe7 26 lId8+, and m ate next move. Black can try instead 21 ... a5 22 4Jc3 �d4 23 lId1 �4 24 lIxd8+ �xd8 2 5 lId7, and White keeps a steady edge. The test position i s more compli­ cated th an it looks, and time pressure was beginning to be a problem. I did not h ave time to analyse in depth, and found one of the worst moves. 18 ... �xg 2 ! i s the best, although thi s would take good nerves. 18 ... �xb2 i s pl ayable, but not 18 ... �xd2?, when I did not see a tactic later on .

slight pressure for the last few moves, but my next move,

18 ... h6?! (A), is highly compromis­ ing: a pawn weakness. Before too long, Bl ack decides h e has to pl ay .. .f6 as well as ... h6, and then he has light-squared weaknesses in front of the king . 18 ../61 ( B ) i s much better, and is about equal . .

Test 6.4 I.Lauterbach-C.Crouch

Britis h League (4NC L)

2007

We resume pl ay from 4.1 with 17 4Je2 4Jd8 18 h4. I h ave been under

68

Black will be able to keep the pawn on h 7 . Any h 5-h 6 push by White can usually be countered quite easily. No detailed analysi s here. Just play through the game, and imagine wh at would h ave h appened if Black h ad de­ layed ... h 6 .

Tes t S ix Another possibility i s l.B ..tDe6?! (C), with a trap. If 19 iifS .tc8 ! ? 20 iixdS ? CiJe], and the queen i s unexpectedly about to be trapped. 21 'ilYxC4 is the only move, but 2 1 ... .ta6 skewers the knight. An attractive variation, but the simple 20 hS keeps an edge for White. In the g ame, White quietly re­ treated with 19 .td2. .

Black does not h ave any imm ediate problems just yet, and the computer gives it as equal , but there are will be difficulties ten or twenty moves along the line. In other words, this i s a posi­ tion al battle, with advantage to White, rather than a tactical struggle. There are two basic problems with Black's pawn structure. First, Black i s suffering from covering what i s in ef­ fect an isol ated pawn in the centre. His (-pawn has moved too far, and cannot defend the ds-pawn, nor even do any­ thing to attack White's pieces and pawns. Black is forced to work out h ow to defend the ds-pawn . Second, White has g ained space on the king side, with pawn s as well as pieces. This suggests

that Black i s forced to defend his king­ side structure as well as the central pawn . White can think of a possible attack again st the king . I pl ayed 1 9 .tb7, in part t o cover the ds-pawn, al so perh aps to try ... .tc6 later, to cut down any m anoeuvring by White's queen and rook. It seem s a slightly lazy move. I was in pain suffer­ ing after a nasty fall, as well as longer­ term illnesses, and I was not playing energetically. Of course the only way of losing a game of chess is to m ake bad moves, and so somehow the chess pl ayer, when under pressure, still h as to work h ard. Lauterbach played better th an I did in the middlegame. 19 ... CiJe6 is more relevant, and if 20 hS CiJe7 21 CiJes ? ! ..txes 2 2 dxe s CiJcs, and Bl ack soon m aintains the balance with his pieces. 21 CiJh4! ? keeps a slight edge for White, though. Lauterbach played 20 h S . . . .

I played 2 0 CiJh8?!, which may seem strange and unnatural, placing the knight into a passive corner. It is not as bad as it looks, and I did not like . . .

69

Why we L o s e a t C h e s s the alternative. Black's mi stake came later. I felt the knight was, in several respects, worse with 20 . ..tlJe7 than on h8, blocking the e-file.

There i s al so the questi on of m a­ noeuvrability. The knight on e7 can go to c8, but then wh at next? At least on h 8, Black can find a better square with .. .f6 an d ...elJf7. But, remembering ear­ lier comments, Black would h ave done much better with 18 .. .f6 ! (kicking the bishop out), rather th an 18 ... h 6 ? ! . O n specifics, after 20 . ..tlJe7, i f White plays 21 ..tf4 ttJe6 22 ..txd6 iNxd6 2 3 ttJe s ? ! , following the idea of the m ain line, then Bl ack equalizes with 23 ... ttJc6, a good argument for the ... ttJe7 ap­ proach . 23 l:!ab 1 ! keeps White an edge though, forcing Black to work out h ow he h as to face b3 (or b4), or ttJes, or, after 23 ... ttJc6, then 24 ttJh4. I suspect th at I might in stead h ave been worried about 21 ttJes ..txe s 2 2 dxe s, opening up a square for the knight on d4. Thi s does not seem all that effective after 2 2 ... ttJfS ! . If 2 3 ..th 3 ttJd4 24 ttJxd4 iNxh 3 (threatening the

70

h s -pawn) 2 S iNfs ..tc8 26 iNxh 3 ..txh 3 2 7 f3 ..td7, and Black should equalize, with the idea of .. .f6. So the obvious 20 ... ttJe7 is better than 20 ...ttJh 8, but only because of a self-pin with ... ttJfS, not so obvious. I could easily h ave added this as a test position, but there would be too many interruptions if there are questions to be asked on each successive move, for each minor slip. We need to keep the flow going. Without the ... ttJfS self-pin, Black's knights would h ave been clumsy, the knights on d8 and e7 not working well together, and not helping the other pieces. This was what I remember being worried about at the time. Back to the m ain line. Lauterbach played 21 ..tf4, exchanging off Black's better bishop. Then 21 ttJe6 22 ..txd6 iNxd6 leaves White with a slight, but annoying, edge. ...

She played 23 ttJes, which instinc­ tively surprised me. As Nimzowitsch used to say, "the threat is stronger than the execution". The knight is not attack­ ing anything, apart from the easily cov-

Te s t Six ered lz'lxc4, and Black can hit back, later on, with .. .f6 followed by ... lZ'lf7. Probably White should leave the knight at home, allowing Black to decide whether to try .. .f6, without any tempo gain, and have to decide whether it is playable or bad. 23 l:tab1! would be a way to test her op­ ponent. Looking at this now, 20 ... lZ'lh 8 was over-elaborate, and does not do the job, so deserves its ' ? ! '. After her knight advance, I played

23. . .l:te7, then came 24 'ii'd 2?!. Maybe 2 4 lZ'lg4! l:tfe8 2 5 lZ'l e 3 would have more effectively justified White's lZ'le 5, keep­ ing an edge. I was not sure what White was doing with her queen move. Obvi­ ously White is not worse after the text, but she could have achieved even more. Play continued with 24 ...l:tfe8 2 5 Wf1 f6 26 lZ'lg4, and at last I felt I was fully equal . We sh all resume the game later in Test 10. 3 .

71

Test S eve n 7.1 White to play

73 Black to play

A) 19 J::t a 3 8) 19 tLig s C) Something el se?

A) 20... ii.a3+ B) 2o . f6 C) Something else?

7.2 White to play

7.4 White to play

A) 20 h 3 8 ) 20 .l:!.d1 C) Something else?

A) 2 1 h 3 8 ) 2 1 'iVd4 C) Something el se?

72

.

Te s t S e v e n

Test 7.1 C.Crouch-M.Peacock Kidlington

2007

Following on from Test 5 . 3 , this is an extrem ely difficult position to try to assess. I played 19 .l:1a 3 (A) over the board, after some thought, but found thi s un satisfactory. My initial idea was 19 .l:1e1 (C), but after 19 ... i.. xf3 it looked as though Black was doing well. 19 tLJgs (B), a suggestion by the computer, looked good, and on seeing this, I regretted that I did not look at it more closely. That said, looking at it ag ain, it is not so convincing. These are the three most active options, but the quieter continuation s might turn out to be more effective, if the more com ­ bative moves do not impress. The one really clear assessm ent i s th at 19 �d4?? i.. xf3 is a blunder, losing a piece. The position is to be regarded, at least initially, as ·unclear'. It is unlikely that anyone is clearly worse, as there

are strength s and weaknesses for both sides. If the position is genuinely un ­ clear, the implication i s th at unless ei­ ther player m akes a mistake from thi s position, t h e final result should e n d up in a draw, with neither player being able to force a win . This i s much as at the start of the game. The one way of losing is to make a mistake. The best way of avoiding loss i s to avoid mis­ takes. Caution, not aggression, is therefore the watch -word. Let us go through the position in greater depth, after the ini­ tial assessm ent, and consider 19 .i:!.e1 (C) 19 ... i.. xf3 . My notes at the time ex­ amined only 20 gxf3, keeping the queen s on the board. After 20 .. :��Yf5 2 1 "iid4 tLJd7, White h a s slightly more ac­ tivity th an Black with the pieces, but his pawn structure is badly dam aged. " E qual, but no more", I wrote earlier, and without trying to an alyse in ex­ treme depth, this seem s fair enough. It is interesting th at in my notes at the time, I did not even mention 20 �xf3 ! ?, protecting the pawn structure.

73

Why we L o s e a t Ch e s s White is then slightly better after 20 ... 'ilVxf3+?! 21 �xf3, moving towards a favourable endgame. I feel sure th at I would have assessed thi s position cor­ rectly. I h ave no fear of a queen ex­ ch ange. Indeed, I would almost cer­ tainly h ave decided that 20 ... 'ilVd4+! i s best, a n d highly unpromising if White is thinking of playing for an edge. After, for example, 21 i.. e 3 'ilVxb2+ 22 �g l tte8 2 3 ttab1 'ilVC3 White will later re­ cover the pawn on b7, but this will lose time, and Black's knight will reach good central squares after ... ttJd7. There is a clear danger th at White could easily end up worse, and thi s is exactly wh at I wanted to avoid. We move next to the line I actually played, 19 tta 3?! (A) 19 ttJd7 20 tte1 (but not 20 tte3 ? 'ilVxf4 2 1 l:Ie7 �fS, and Bl ack win s a pawn). Then 20 'ilVb4! proved uncomfortable for me.

Sometimes it is better to use a com­ pletely new perspective. I h ad been as­ suming th at I needed to bring the rooks into pl ay quickly, giving myself the chance of actively developing the bishop. I took it for granted th at the knight was pinned, and th at at some stage Black would exch ange on f3, if required. But maybe the knight can h ave an impact? It took me a long time to think about 19 ttJgS!? (B), even after the gam e, and even with the h elp of the computer.

...

...

Black is still keeping pressure on the b2-square. I did not h andle thi s posi­ tion well, and I was fortun ate th at he opted for a perpetual when short of time. But that is a stage of Test 7.4.

74

If 19 ... i.. x d1 ? 20 ttJxe4 i.. c 2 (20 ... i..b 3 ? 21 tta3 wastes time) 2 1 ttJxd6 with a big plus for White. Bl ack needs to improve, with 19 .. :YWfS . Then perh aps 20 'ilVd4 (20 1!Vh3 ttc8 is much slower) 20 ... l:Ie8 2 1 h 3 'ilVc2+ 22 �g 3 i..fs 2 3 'ilVf6 'ilVd3+ 24 �h 2 'ilVxds 2 s ttd1 ttJd7 ( 2 S ... ttxd1? ? 26 'ilVxf7+) 26 �xg6+ hxg6 2 7 l:txdS. Even h ere, 2 7 ... ttJcS i s slightly better for Black. White's bishop still remain s undeveloped, and his rook still obstructed. This suggests, as an alternative, 19

Tes t S e ve n 1i.e3 (C), but after 19 .. .'�Jd7 followed by ... tt:Jf6, the knight suddenly reaches a good square. There are various other possibilities for White, including 19 as, 19 b3 and 19 l:.fl, but none seem s completely secure. Bl ack quickly plays ... tt:Jd7. By now, it is easy to see that analysis can easily run around in circles. After some thought, I played what seems to be the natural move, having found nothing better over the board. It looks like that there ought to be an improve­ ment for White, but even in later analy­ sis, I cannot find a clear drawing line, let alone a win. So perhaps I was worse? Around the roundabout, White's origi­ nal exit, 19 l:.a3, now looks the best, but I had to find good moves later on.

now do I respond? I n the end, I played 18 iLC2!?, which I described as a "wimp out" in an earlier set of analysis. Play continued with 18 :iVf6 19 iLC3 iLd7, which was critical, and the point at which White needed to improve. We sh all examine thi s below, although of course the con scientious and thorough reader will already h ave examined this. Sometimes it i s appropriate to grab the exch ange, but sometimes it should be delayed. It i s always a difficult deci ­ sion to m ake. White certainly needs to consider 18 1i.xf4! ? exf4 19 h3 "iYf6 (maybe the simple 19 ... tt:Je s, with rea­ sonable compen sation for the ex­ ch ange), and indeed I con sidered it, and rejected it. ..

Test 7.2 Colin Crouch - Andrew Lewis

Kid l i ngton

2007

As we saw in Test 3 .4, Lewis h as of­ fered an exchan ge sacrifice, but how

This game was played a few years back, before my writing on Tal, Stein and Kasparov, and also my book on wins by 2 700+ players in Modern Chess: Move by Move. I would like to think that by now I would have been more fully aware of the idea that even in the wild­ est positions, the best play by both sides should end up in a draw, unless of

75

Why we L o s e a t C h e s s course one of the players is already bet­ ter. What this in practical terms would mean is that if your position feels okay before the opponent's attempted sacri­ fice, and if you see no clear advantage for your opponent, then you are at lib­ erty to accept the sacrifice, however dangerous the opposing sacrifice might appear. 'Fortune favours the bold', and bold defending, as well as bold attack­ ing, should be praised. Bl ack h as no win after 19 .. :�f6 ? ! , which sets u p an apparently extremely dangerous attack, but it turn s out that Black has no more than a draw after 20 hxg4 �al+ 2 1 i..b l i.. x g4 (he h as to aim for ... idS) 2 2 l:I.dl ii.b2+ 2 3 �C2 i.. C 3 24 �Cl i..b 2+ 2S �Cl. And, second, th at the sneaky 24 .l::i. c l ! ! gives White an unexpected advantage after 24 ... iLfS+ 2S �dl i.. x bl 26 �e6+ �h 8 2 7 �f7. This line would h ave been extremely difficult to foresee for either side. So 18 i.. xf4 is extremely promising, but 18 iLC2 !? �f6 19 iLC3 iLd7 i s al so fully playable, provided I play accu­ rately.

76

I h ave avoided the tactical battle, but now I h ave to con sider the posi­ tional battle. I should not be worse, but my next move is poor. I played 20 l:I.dl? (B), but I soon was forced to appreci ate that the rook was on an unfortunate square, and that after 20 ...l:I.f8 21 h3?! (mistimed, but even h ere 2 1 l:I.hel l:I.xf3 ! ? 2 2 gxf3 �xf3 2 3 �xf3 l:I.xf3 24 l:I.d3 l:I.xf2 clearly fa­ vours Black) 21 ...iLlh6 22 l:I.d2 iLlf5 23 iLlel iLld4 my position was gradually falling apart. There was not much to be done after 24 iLxd4 exd4 25 iLld 3 i.. h 6 26 �bl l:1.e8 27 �dl l:1.fe4 28 f3 l:I.e3 29 �Cl i..f 5 30 �xa 3 l:1.xf3 3 1 l:I.ddl l:I.g3 32 'iia 7 l:I.xg2 33 'iix b7 .l::i. e e2 34 �b8+ �f8 3 5 �xf8+ �xf8 0-1. During the g ame, I was regretting th at 20 h3!? (A) would h ave been m ore accurate, so that after 20 ... iLlh 6, White h as a wider choice, other th an 2 1 l:I.dl ? ! , in reply. 2 1 l:1.el would certainly h ave improved, trying to bol ster the e4square, but Black is still better: for ex­ ample, 21 ... iLlfS 22 iLld2 iLld4 23 iLxd4 exd4 24 iLle4 'iVe s 2 S i.. d 3 i..fs 26 f3 (26 �C2 i.. x e4 2 7 .:t.xe4 .:t.xe4 2 8 i.. x e4 d3 29 i.. x d3 �g S+ favours Black) 2 6 ... iLxe4 27 fxe4 (27 i.. x e4? d3 28 i.. x d3 �al+ 29 �d2 'iix a2+) 27 ... .l::i. af8 28 .:t.hfl l:1.xfl 29 .:t.xfl .:t.xfl+ 30 'iVxfl �g S+ and White is uncomfortable. In such lines, Black has excellent play on the dark squares, with opposite-coloured bishops, but White has no corresponding pl ay on the light squares, his own pawn s blocking his bishop and queen.

Te s t S e ve n So wh at else i s there? Clearly I need to bring my pieces into pl ay, but in the game I soon h ad to give protection with l::t d 2, thereby losing time. This suggests an improvement, with 20 �fl!? (C).

h as a flight square for both knight and rook, but also that the es-pawn is no longer pinned. If, for example, 2 1 h3 e4! 22 .txe4 .txc3 2 3 hxg4 l::t e 8 24 tiJg s l::tfxe4 2 S tiJxe4 'iWf4+, and Black ends up with a winning advantage. There are various other possibilities for White, but while the knight is on g4, and creates pressure, White faces problems. The best way for White, it seem s, is to kick the knight out immediately, with another 'something else' move, namely 20 h3!? (in stead of 20 .t:.d1 or 20 .t:.f1), and then 20 ... tiJh 6 21 l::tf1 ! . The only problem is th at my intention was to avoid pushing the knight away to a better central square.

Suddenly White's quiet pl ay with ii.c2 and .tC3 m akes more sense. If he has time to regroup with, for example, 20 ... l::tf8 21 h3 tiJh 6 22 .te4 .tfS 2 3 tiJd2, White h a s finally con solidated h i s centre, with perh aps a slight edge. However, 20 ...'iVf8 !, a suggestion from John Emms, is an improvement for Black.

The point is not just that Black now

If then 21 ... tiJfs, White would probably be quite h appy to break the central tension with 2 2 g 3 tiJd4 2 3 tiJxd4 ( 2 3 .txd4 l:txf3 24 i.. e 4 l:txb 3 ! favours Black) 2 3 . . . exd4 2 4 gxf4 dXC3. Thi s might not necessarily be the most critical position where there is a choice of several moves, but it is certainly one of them. Thinking in term s of a player trying to analyse this position over the

77

Why we L o s e a t C h e s s board, the natural inclination would be to think th at Black is better, despite being the exch ange down . Black has two advanced pawn s on the queenside, whereas White has a weakened pawn structure there, making it difficult for White to bring his pieces into play. If the bishop moves, for example, ... c2 would be a reflex reply. If the reader is finding it confusing over the intricate changes in move or­ der, and the sudden emergence of tac­ tics, then me too. I still feel th at it is almost a m atter of principle that White ought to be able to keep some sort of edge in the King's Indian, with the ad­ vanced central pawn structure, but thi s turn s out t o b e extremely difficult to prove. Every time that White feel s that h e has a secure edge, there always seem s to be some unexpected way for Bl ack to hit back. Positional chess is so often ex­ tremely delicately bal anced, particu­ l arly when it involves sh arp position al pl ay. It might often be the case th at a player might h ave four natural moves with which to carry out a m anoeuvre, but it is far from clear which precise move order needs to be pl ayed, given the case th at the opponent might al so h ave four possible manoeuvres. Of course, at any stage one of the pl ayers might well ch ange the balance of pl ay, possibly ending up with tactics, or pos­ sibly with a slightly varied new strate­ gic pl an . A move not only puts a piece on to a new square, but al so removes a

78

piece from its previous square. For the moment, playing ag ainst complicated King's Indian openings, I h ave been playing the Torre (1 d4 tDf6 2 tDf3 g 6 3 �g 5), before trying again something more ambitious.

Test 7.3 P Sowray C Crouch .

-

London

.

2007

Thi s position ought to be fully play­ able for Black, one might think, but after a few more moves White started to emerge ahead, and later won . It was important for me to try to work out how and wh ere I went wrong . My as­ sumption during my initial an alysis was th at I defended far too cautiously, and that I should h ave fought fire for fire. After further scrutiny, I now h ave come to the view that the next few moves for Black were correct, ending up with a slight but tenable disadvan ­ tage. The attempt for Bl ack to play for an advantage, in m any lines, would

Te s t S e v e n have ended up as an advantage for White, often after tactical play. I played 1s ... e s ! ?, a move l ater than I should h ave done, as we saw in Test 4.4. Black's ... bS proved to be unneces­ sary. Instead, Bl ack could h ave tried taking up the ch allenge with lS ... b4 16 liJdS exds 17 exds+ d8 18 liJc6+ ..txc6 19 dxc6.

Such minor piece sacrifices in the Sicilian are extrem ely scary, and I felt I wanted to avoid this. Black i s not threatened with checkm ate yet, so there i s still some time for flexibility, and the chance to pl ay 19 .. :�as ! . Then a quick and unexpected repetition with 20 l:td4! liJb6 21 �e s ! liJC4! 22 �e2 liJb6, difficult to envisage a few moves in advance. In stead, 20 ..td4? �xa2 2 1 i.xf6+ C7 ! 2 1 b3 liJa3 would h ave lost quickly for White. A third possibility i s the simple developing approach with lS ... ..te7 ! ? 16 b3, then either 16 ... b4 or 16 ... liJa3, with tough play, probably ending up around equal . Black's quiet move would per­ haps h ave been the way to play for a

win, but i t would take a con siderable amount of hindsight to think about this line. I felt comfortable with my central pawn push, but 16 b 3 ! ! took me out of my comfort zone.

The big problem from Bl ack's point of view is that he weakens the e-file if he captures on d4. White then can open up the e-file, with eS or liJds, when he is taking the initiative. I did not expect my opponent's move, but now wh at can I do? This was a difficult choice. For a long time, I concluded that my reply, 16 ... exd4, was a mistake, but it looks okay, and it was the next move th at was my mistake. The computer suggests, after a bit of prodding, th at 16 ... liJa3 17 liJdS �d8 gives an advantage for Bl ack, but this seem s unlikely. Computer analysis tends to concentrate on trying to avoid the loss of a piece, but strong hum an players will want to keep the attack going, overriding loss of material . In­ deed, 18 l:td2 exd4 19 ..txd4 leaves

79

Why we L o s e a t Ch e s s Bl ack uncomfortable, for example after 19 ... 3Le6 20 'it'b2 b4 21 3Lxf6 gxf6 2 2 'ilVxa6. 16 ... 'ilVas is slightly more accurate,

All sorts of pieces were flying around in my mind, and eventually I decided th at it was too ri sky for me to try to keep the extra piece and win . I was more worried that I would end up losing. So I blocked the e-file with 17 ..ltJes ! 18 fxes dxes anyway, and later lost. My l ater reaction after the game was th at returning the piece was a mistake, leaving me with a positional disadvantage. Surely I could h ave pl ayed more ambitiously, with an un­ clear position, and with ch ances of keeping the extra piece? Of course, one cannot m ake an as­ sessment completely abstractly, and precise evaluation of pieces and varia­ tion s needs to be con sidered. In home analysis, I examined several lines, both from thi s position, and in earlier posi­ tion s. I decided th at in the end Black h ad gone wrong on move 17, with the possibil ity of a complicated perpetual . I al so decided that even with best play, Black h as not got an objective advan­ tage. I cannot play too ambitiously. It was only a long tim e later th at I started to look again at the quiet line of the g ame, deciding in the end that with an improvement, Black's play would h ave led to close equality. In other words, my assessment in the early part of the g am e was correct (apart, of course, from the imprecise 14 ... bS), but that there was a serious mi stake l ater on . We shall consider the improvement soon, but first it i s time to examine the wilder lines. .

in the sense th at after 17 bxc4 bxc4, White can only choose between 18 .l:!.f3 and 18 .l:!.g 3, and cannot play 18 tte3 ? in view of 18 ... exd4. This m ay be com ­ pared with 16 ... exd4 17 3Lxd4 'ilVas 18 bxc4 bxC4, when White h as the choice between 19 .l:!.e3, 19 .l:!.f3 and 19 .l:!.g 3 . Even so, 19 .l:!.g 3 3Le7 20 g s hxg s 2 1 .l:!.xg s 'ilVa3+ 2 2 'it'dl still keeps an edge for White. Anyway, after my immediate cap­ ture on d4, he took back with 17 3Lxd4.

80

Te s t S e ve n It is, of course, a bad defen sive mis­ take to retreat m eekly with 17 ... liJb6 ? White crashes through in t h e centre with 18 e 5 . Black needs to play actively, either returning the extra piece, or us­ ing it in a counterattack ag ain st the white king . So 17 .. :�a5 ! ? needs to be examined.

We m ay quickly reject 18 liJd5 ? as premature: 18 ... liJxd5 19 exd5+ �d8 20 bXc4 lIxC4 is safe for Bl ack, and he has an edge. White's king is not so safe. 18 e 5 ? ! i s the m ost immediately ag­ gressive line, but Black again h as quick counterpl ay, with 18 .. :�a3+ 19 �dl ttJb2+ 20 �d2. Then 2o ... liJxd3 ? would be too greedy, in view of 2 1 exf6+ liJe5 22 .txe 5 dxe 5 23 'iVxe 5+ .te6 24 fxg 7, with advantage to White. In stead, 2o ... dxe5 21 .txe 5 ! (Black is better after either 21 �xe 5+ iJ.. e 7, or 21 fxe 5 liJxg4 22 liJxd3 23 'iVxd3 iJ.. x g4), and it is be­ coming even sharper. The computer suggests two win ­ ning moves, but 2 1...liJxd3 ? is in the end far too entertaining to be wise: 2 2 iJ.. d 6+ iJ.. e 6 2 3 iJ.. x a3 liJxel 24 iJ.. xf8

liJxg 2 2 5 iJ.. x g 7 liJxf4 2 6 �f3 lIxC3 2 7 'iVf4 i s g ood for White. 21 ... JJ.. e 6 ! is much better, and good, but a nightm are for a player trying to analyse over the board. Or, for the Tal sch ool of attacking chess, many would take a positive delight in playing through such lines, knowing that when neither player can analyse to a clear result, artistry in chess prevails over m ere calculation. Pl ay might continue with 22 f5 liJxd3 23 'iVxd3 l:td8 24 JJ.. c 6+ �e7 2 5 .txf6+ g xf6 2 6 liJd5+ �d6 27 �g 3+ (27 fxe6 �xc6 2 8 e7 i.. x e7 win s for Black) 2 7 ... �xc6 28 'iVC7+ �xd5, which eventually leaves White running out of checks.

It would take far more th an a quick glance for Black to convince him self th at his king is safe in the centre, but he h as some defen sive ammunition though h aving an extra rook plus two extra bishops. Al so, perh aps even more significantly, White's king is on an open and exposed square, with checks and a discovered check by Black's king cutting down White's option s. If, for

81

Why we L o s e a t C h e s s example, 30 c4+ bXc4 3 1 bxc4+ 'it>d4 3 2 �f4+? (but 3 2 'iVxd8+ 'iVd6 i s al so a win for Black) 32 ... 'it>cS+, and yes, thi s is a counter-check, so White i s not allowed to play �f2+. The mistake took pl ace much earlier in this line. Back to 17 ... 'iVas, and now 18 bXC4 ! .

B y recovering the piece quickly, White slows down his attack, and al so his opponent's attack. We are now in a sh arp positional set-up, rath er than all­ out tactics. The computer suggests that the position is equal, but after further examination, White keeps an irritating edge. Black is not equal, and the move I actually played, 17 ... tDes ! , was closer to full equality. H ere 18 ... .l::t x c4 does not work, as White h as a resource with 19 eS J::i. x d4 20 exf6+ 'it>d8 21 .l::t x d4 �xC3 22 'iVe3 ! , choking Black's counterpl ay. Therefore, 18 ... bxC4 (or 18 ... 'iVa3+ 19 'it>dl bXC4), but White still keeps an edge after ei­ ther 19 .l::t e 3 or 19 .l::t g 3 Ji.e7 20 gs hxg s 2 1 J::i. x g s, a side-line we h ave noted be­ fore, after a tran sposition .

82

After 1 9 J::i. e 3 'iVa3+ 2 0 'it>dl Ji.e7 2 1 eS dxes 2 2 J::i. x es Ji.e6 White stands well, the most direct way being to sac­ rifice the exchange with 23 J::i. x e6 fxe6 24 'iVxe6. Not pleasant for Bl ack. So we return to the m ain line, and Bl ack's return sacrifice at move 17. White regains the knight, then 19 tDd s tDxd s 20 exd s .

I was understandably concerned that I had done nothing to develop my king­ side pieces since move S, and that my king was still stuck in the centre. I was greatly relieved that I could gain a tempo with 20 .ia3+? (A), and then castle quickly. This unfortunately proved to be a serious mistake. After 21 Ji.b2 Ji.xb2+ 22 'it>xb2 0-0 23 d6 'iVcs 24 �d2! (a quiet queen move which I h ad underestimated), White has excellent central control, and an advanced and dangerous passed pawn. Black h as none of the usual Sicilian counterplay on the queenside, although of course I tried for some. We return to the finish later. 20 f6! (B) is not so far from being level . Black keeps his bishop in play, ...

...

Tes t S e ve n making it far m ore difficult for White to advance his passed pawn .

Perh aps it would be only just for White to be able to keep a slight edge, given that h e i s better developed, and has a good passed pawn . So 2 1 .l:i.C3 ! i.a3+ 2 2 bl "ifd6 2 3 .l:i.xc8+ .txc8 24 .l:i.dl 0-0 25 .tal .td7, and the struggle is likely to continue after the adjourn ­ ment. Bl ack h as not yet equalized. A computer suggestion gives 2 6 h4 l:!.e8 27 g5 f5, and it i s not so clear that White has broken through after either 28 gxh 6 �xh 6, or 28 � 5 e4. It is time to return to the m ain line, after 24 �d2.

The next few moves were played at some speed, with 2 5 d7 l:!.cd8 26 l:!.d6 b4 27 l:!.d1 as 28 .tc6 "ifb6 29 "ife2 "ifcS 30 l:!.ld3 hS 31 gxh s .tfS 32 .te4 .txe4 3 3 �xe4 fs.

I h ave m anaged to keep up some counterplay, hoping that there is going to be an in accuracy by my opponent during the time scramble, m aybe due to some pressure in front of his king, and above all, pushing the two passed pawn s in the centre. I did not expect my position to hold, but White was clearly nervous of my pawn s. 34 "ifc4+ ! "ifxc4 35 bxc4 slaloms through with White's second pawn . Both players h ave two connected passed pawn s, but it is not difficult to work out that White's pawn s are fur­ ther ahead, closer to promotion . For example, 35 .. .f4 36 c5 f3 37 l:tdl .l:i.f4 3 8 c6 .l:i.C4 3 9 l:!.el e 4 4 0 .l:i.xe4 l:txe4 4 1 c 7 l:!.ee8 4 2 dxe8"if l:!.xe8 43 .l:i.xd8. White played instead 34 �h4?!, which gave me a little encouragement, but once White h as the opportunity of attacking with h6, Black is likely to run

83

Why we L o s e a t C h e s s out of time. Then 34 ...e 4 3 5 l:td1 led to my final mistake, which is worth re­ cording for psychological interest, but not really worth giving as an exercise.

35 ... a4? 36 bxa4 g ave away a pawn for nothing. White has to capture of course, as otherwise ... a3+ is likely to be winning. There are so m any examples in chess of giving up a pawn to break up the opponent's pawn structure (in­ cluding 30 ... hS in this g ame), and sometimes such moves can be played almost automatically. H ere, all I m an ­ aged t o achieve was t o weaken m y b4pawn, while giving my opponent a passed a-pawn . I was annoyed afterwards with my misjudgement, but in the end it did not m ake much difference to the final out­ come. The tactics remain the same. Play continued 36 J!Ve5+ 3 7 '.t>b1 (see diagra m) 37 ...f4, and 1-0. A strange position to resign a game, except for the point that this was an evening London League game without a quickplay finish. I resigned at the ad­ journment. There is no point in travel -

ling to central London when with the help of computer an alysi s (and, of course, my opponent was al so more th an competent to analyse without computer help), it is easy enough to find a win by the opponent.

The liveliest line is 37 ... e3 38 'iVC4+ Wh 8 39 l:te6 "iVg 3 40 h 6 g xh 6 41 "iVd4+ Wg8 42 'Yi'b6, and White wins.

Test 7.4 C.Crouch-M.Peacock

Kid l i ngton

2007

••

84

Following on from Test 7.1, I was

Te s t S e ve n understandably nervous about thi s position , m y pieces all being o n the edge, except for the pinned knight and the exposed king, whereas Black's pieces are on good positions or on open lines. My position is not yet drastically bad, but I need to be careful, and cer­ tainly I am not better. Someh ow I need to activate my pieces, or to exchange some m aterial . I chose the second idea, playing the sim­ ple 21 h3?! (A) 21 ii.xf3 22 l::i. xf3 l::i.fe8, but my position was worse, and I was fortunate that h e eventually decided to take a perpetual when he was better, but short of time. The problem with the move I tried to defend with i s that it i s using u p another pawn move, when really I needed to concentrate on bring­ ing my pieces together. Even 21 rj;;g3!? (C) is a possibility, in­ viting Black to exchange on f3, while also m oving the king to a safer square quickly.

can b e useful . If, for example, Black were to carry on with the line of the g ame, with 2 1 ... ii.xf3 2 2 l::i. xf3 l::i.fe8, White now h as 2 3 Itfe 3 ! , and the f­ pawn i s safe. 2 1 ... ii.fs 2 2 Itae3 also keeps White's pieces in play. 21 'iVd4! (B) is a much simpler line, indeed almost too simple if White is still hoping to outplay his opponent.

...

There is no check on the g l-a7 di­ agonal, and al so the less obvious point that setting up an extra defence on f4

The queens come off immediately, with 21 .. :iVxd4 22 tLlxd4, and then 22 ... tLlf6 23 l::i.b 3 .l::i.fe8 24 .l::i. x b7 .l::i. x el 2 S rj;; x el tLlxds 2 6 fS ! , and once White's bishop is in full play, he comfortably holds the balance. 21 Wid2?! (C) i s worth considering, especially if Black is not interested in steady equality. If 2 1 .. :�xd2 ? 22 tLlxd2 ! , White suddenly jumps t o good squares with the knight, but 2 1 .. .'iVc s+ 2 2 '.tg 3 ii.xf3 2 3 l::i. xf3 tLlf6 is about equal . There is a third option . In stead of immedi­ ately exch anging, or escaping with the queen, Bl ack can keep the tension with 21 ... a s ! , when he seem s slightly better. There are then a couple of dubious gambit ideas. If 21 il.d2? (C) 21 .. :iVxb2

85

Why we L o s e a t Ch e s s 2 2 l:tb3 �a2, the queen is o n the edge, but then what next? 2 3 l:txb7 .i.xf3 24 gxf3 tLlC5 win s the exch ange for Black. Or 21 �ae3?! (C) 21 ... �xf4 22 �e7 �f5, and White does not h ave quite sufficient compen sation for the pawn . There were clearly quite a few al­ ternatives to con sider, but offering to exch ange the queens with 2 1 �d4 looks the best and safest. We return to the position as di s­ cussed on move 22, and then 2 3 �xe8+ �xe8.

.i.b2 �e1+ 29 �f2.

Both king s look to be in trouble, but a critical attacking piece in either side is pinned. My opponent thought until the last minute. As the earlier minutes ticked by, I started to feel more relaxed. If he could find a win in stantly, then so be it. If he h as to take a few extra minutes, he would not be able to find a more complicated win . Eventually he decided it was time to take the draw with 29 �e3+?! 30 �f3 �e1+ 31 �f2 1/1-1/1. I still h ad the feeling that I should probably h ave been losing, and in later analysis I was able to find something, with the h elp of an endgame zugzwan g . But who would be able to find the line, the first part of the proc­ ess, and then spend th e extra tim e to confirm th at the critical line is indeed a zugzwan g ? The critical l i n e is 2 9 . . .�f1 ! (or, after repetition, 3 1 ... l:tfl) 30 .i.xf6+ �xf6 3 1 � 8+ �e7 3 2 �d4, when the king and pawn endgame is equal if the queens and rooks are exchanged. ...

White is 'almost' close to being equal, but it turn s out th at he i s not quite there. I h ad misjudged thi s posi­ tion, and none of my pieces are on a good square. The only hope, I felt, was to try to take control of the long diago­ nal , with 24 b3, then .i.b2, when given the chance. Black played 24 tLlf6 (24 ... tLlC5 i s also promising, and if 2 5 .i.b2 tLlxa4), and then came 2 5 �g3 ! ? O n other m oves, he will soon win a pawn . I was h oping that Black would not be able to find a winning plan after 25 J1e1 26 �C2 �h1 27 �c8+ Wg7 28 ...

..

86

Tes t S e ve n

However, it turns out that Black can still win with 32 ... h S ! ! . If then 3 3 �d2 ??, Black win s with a cross-pin after 3 3 ... h4+ 34 'it>h 2 �xf2. If in stead, 3 3 as h4+ 34 Wf3 �d1+ 35 'iWxd1 1:Ixd1 36 Wg4 1:IxdS 37 'it>xh4 .l:!.xas, and Black will eventually grind away for a win in the pawn-up rook

ending. Even the inoffen sive a-pawn ends up in a zug zwang trap. Or if 33 h4 as, and White's pawn s are now blocked, and h e will lose to zug zwang in a king and pawn ending after 34 'iVd2 �xf2+ 35 �xf2 lhf2 36 Wxf2 Wf6 3 7 Wf3 WfS 3 8 g 3 f6 3 9 'it>e3 gS 40 fxg s fxg s 41 hxg s wxg s 42 Wf3 WfS, and a pawn drops. A long varia­ tion, but al so elementary. Thi s leaves 3 3 'it>f3 . Instead of look­ ing for complications with another zugzwang, there is a simple line with 33 ... 'iVd1+! 34 1:Ixd1 1:Ixd1, when a pawn drops, and Black eventually win s the rook endgame. Thi s would h ave been an excellent win for my opponent, except th at he missed it.

87

Test E i g h t 8.1 Black to play

8.3 Black to play

A) 2 1 ... �gS B) 21 .. .fS C) Something el se?

A) 22 ...'it>d7 B) 2 2 ...WfS C) Something else?

8.2 Black to play

8.4 White to play

A) 2 1 ... �as B) 21 .. . �a7 C) Something el se?

A) 2 2 ..ib4 B) 2 2 'iVg 3 C) Something else?

88

Tes t E ig h t

Test 8.1 D.Buckley-C.Crouch

British League (4NCL) 2006

Continuing from Test 2.1, White h as just castled, and Black now castled him self, 10 ...0-0. Both m oves seem a bit too automatic, and indeed lazy. As dis­ cussed earlier in this game, White should try d5, or even dxC5, if given th e chance, and Black should h ave avoided this by exchanging on d4. I h ad the higher Elo rating, and so in theory it i s more likely that he would m ake a slight mistake at some stage than I would. Thi s edge is, however, slight if the posi­ tion i s drawish, with little chance of m aking mistakes. The stronger player therefore h as the incentive to try to keep the position lively, but it is easy to overpress. Wh at h appen s, for example, if the opponent does not m ake mis­ takes? Quite often he will keep a slight edge, when with best pl ay the position will remain equal at best. White him self makes another minor

in accuracy, with 11 e4. Clearly thi s does not m ake his position worse, but the point i s th at White still h ad prospects of keeping an edge. 11 .l:Id1 ! ? seem s promisin g . If then 11 ... cxd4 12 'iVxd4!, and White has gained a clear tempo when compared with ttd1 instead of e4, the pawn push being of little rele­ vance. Bl ack, of course, has altern atives, but after, for example, 11 ... d6, White is now better placed for 12 N, and if 12 ... cxd4 13 tt'lxd4 tt'ld7 ? ! there follows 14 tt'lc6 and White wins a pawn . Black h as to find something else, but he is not equal . In the game, I decided it was high tim e to exch ange, with 11 cxd4 12 tt'lxd4 tt'lc6. Black is h appy now after 13 tt'lxc6 dxc6 14 'iVe2 'iVC7, and could think of aiming for an edge, so White develops with 13 iLe3. Then 13 .. J:tcB, showing Black's intention is to focus on White's slightly weakened 'English pawn ' on c4. •••

After 14 .l:Iac1, I h ad several reasonable altern atives, many of these ending

89

Why we L o s e a t C h e s s up a s equality. I tried 1 4 'ilVC7, telling myself th at there was no reason to move the knight just yet. .••

came lively, continuing with 15 ... CZJxd4 16 'ilVxd4 (maybe 16 iLxd4! ?, probably equal) 16 ... .i.f6 17 'ilVd2. Now I saw the glint of an initiative with 17 :�e5, si­ multaneously attacking two pawn s. .•

Even so, the immediate knight ex­ ch ange, 14 .. .'�Jxd4, and only then ...'ilVC7, is a safer way of equalizin g . The computer al so suggests 14 .. .f5, which I admit I h ad not seriously con­ sidered. After 1 5 exf5 CZJxd4 16 iLxd4 .l:i.xf5, Black h as too m any 'pawn is­ l ands', and White is slightly better. So m aybe I was correct in not analysing deeper on this over the board. My opponent too developed quietly, with 15 l:rfdl, with equal pl ay. Perh aps though he should h ave tried for a slight edge with 1 5 CZJb5 'ifl>8. Then not 16 'ilVxd7 ?? l:rfd8, and the queen is trapped, but 16 .l:i.fdl or 16 f4 might still give White a slight push, provided he can keep his pawns secure. By now I h ad weighed my opponent up, and decided th at he would h ave been more th an h appy with a secure draw. This gave me slightly more op­ portunities to push for a slight break of the symmetry. Before long, play be-

90

Then 18 i.d4 'ii'x e4 (if 18 ... 'iVxd4 19 'ilVxd4 iLxd4 20 l:rxd4 with a slight edge for White, but 18 ... 'ilVg 5 can be con sid­ ered) 19 i.xf6 gxf6. I felt reasonably confident h ere, reasoning that it was now up to White to prove that he could keep the bal ance. He thought for a long time, and I saw nothing m ore fearful th an 20 'ilVxd7 l:rxC4 2 1 'iUxa7, and then I saw 21 ... 'iUd4!, with a likely draw after 22 .l:i.bl 'iVc 5 2 3 b4 'ilVC7 24 'iUxC7 l:rxC7. I h ave to admit that I h ad no sense of real dang er, but 20 l:rC3 ! aimed for checkmate, and forced m e to think h ard. If, for example, 20 ... l:rxc4?? 2 1 l:rg 3+ 'it'h 8 2 2 'ilVh 6, which is the e n d of g ame. 20 ... .l:i.c5 2 1 .l:i.g 3+ l:rg 5 2 2 .l:i.xg 5+ fxg 5 2 3 'ilVxg 5 + 'it'h 8 24 'ilVf6+ 'it'g 8 25 h3 'ilVxC4 26 l:rd4 'ilVcl+ 2 7 'it'h 2 'ilVC7+ 2 8 g3 al so win s for White. I was briefly h orrified until I realized th at the king could slip away with

Te s t E ig h t 20 Wh8!, avoiding the rook check. A close escape. ...

White tried 21 �xd7, and I was still hoping to try to pl ay for a win . In stead 2 1 lIg 3 .l:!.g 8 (the reason why Black must m ove his king first) 2 2 �xd7 could easily tran spose into 2 1 �xd7 (see l ater), with, for example, 22 ... �e2 23 !ttl .l:i.xg 3 24 h xg 3 .l:i.f8 2 5 �xa7 �xb2. Black's position i s slightly uncomfortable, but it seem s to hold. 2 1 �6 �g 6 2 2 'iYxg 6 fxg 6 2 3 !txd7 .l:i.fd8 24 !txd8+ .l:i.xd8 25 Wf1 is a straightforward draw if both players pl ay sensibly. In the g ame, Bl ack now jumped too quickly for the attack. After 21 fS? (B), there was more a sen se of relief, th at I was not getting checkm ated, rather th an the pleasure of wild attack. I h ad not really thought about dropping pawn s, m aybe on the basis that ' I was a pawn up, everything goes up and eve­ rything goes down, life goes on .' But 2 2 'Yi'xa7! i s n o w quite simply good. White is simply a pawn ahead, and a second i s about to drop.

There i s not much that needs to be noted, except that 22 ... .l:i.xc4?? 23 �e7 leads to a quick checkmate, and that 2 2 ... .l:!.fd8 2 3 .l:i.f1 leaves White m ore than comfortable. If pawn s are not necessarily the soul of chess, chewing up your opponent's pawns provides good nutrition . Buckley instead tried 22 'Yi'e7?! �es, and the crisis h ad passed. Play was sh arp, and I later h ad a strong attack, without either player handling the po­ sition particularly well. We resume th e an alysis later in Test 11.3. Earlier, Black should h ave tried 21. .l:!.g8! (A) •••

...

91

Why we L o s e a t C h e s s 2 2 :tg 3 ( 2 2 g 3 llgd8 ! ? 2 3 'iYxd8+ llxd8 24 llxd8+ 'itg7 25 :tcl �e2 26 b4 'iVb2 27 llcdl 'iYxa3 should draw) 22 ... llxg 3 23 hxg 3 , with a position we have discussed above. Then 23 ... llf8 24 �xa7 'iYe2 25 llfl 'iYxb2 26 'iYe7 'itg 7 keeps the balance.

that it i s not all th at effective after ac­ curate pl ay by Bl ack. Pl ay continued 16 ... axb4 17 exd6 .1i.xd6 18 llel �b8 19 �b5 .1i.xg3 20 hxg3 'iYb7.

Test 8.2 N.Pert-C.Crouch

B u ry St Ed m u nds

2006

Some quiet pl ay early on, with 1 d4 tbf6 2 tbf3 b6 3 e4 .1i.b7 4 tbe3 e6 5 a3 tbe4 6 tbxe4 .1i.xe4 7 .1i.f4 .1i.e7 8 e3 0-0 9 .1i.d3 .1i.xd 3 10 'iYxd 3 d6 11 0-0 tbd7 12 lladl e5 13 .1i.g3 Wie7, then 14 b4!? I h ad expected 14 d5, and this seem s to give a slight edge for White, after ei­ ther 14 ... e5 1 5 e4, or 14 ... exd5 1 5 Wixd5 . Pl ay continued with 1 4. . .exb4 15 axb4 a 5 .

Now Pert played 1 6 c 5 , which I was impressed with at the time, and which I awarded an excl amation m ark in my notes just afterwards. The trouble i s

92

50 White recovers th e pawn, and

Black is stuck with an isol ated pawn . It is difficult, however, to take the second pawn, and if Bl ack can keep it, and pro­ tect it with care, White will h ave to watch out for the pawn being ad­ vanced, and/or one of the open lines for the black rooks being a threat. White played 2 1 llbl. Now I took the wrong plan , with 2 1 ... lla5?! (A) 22 'iYxb4 �e4?!. The problem is th at Bl ack is moving his pieces far away from the weak isolated pawn, and he is going to h ave to work h ard to cover the pawn . Black cannot attack very h ard on the king side, and so White soon keeps the initiative. There are two basic methods of de­ fending again st positional pressure. One is to set up counterplay, where the opponent is weak. The other i s to con­ solidate, as firmly as possible ('overpro­ tection'), so that the defender covers all

Te s t E ig h t weaknesses on his own part of the board, with later thoughts of his own counterplay. On thi s occasion, since White has no genuine weaknesses on the kingside, Black should h ave over­ protected on the queen side. Therefore, (B) 21 ::071 (and indeed there m ay be others, such as 21 .. JUc8 or 2 1.. . .l:!.a2) 2 2 .l:!.xb4 .l::tfa8, and Black's pieces are tightly compacted, but with ch ances of springing out again. ..

The knight protects the isol ated pawn, the queen and rook cover th e knight, and either the queen or rook could bounce back into play, and none of Bl ack's pieces is confined to passiv­ ity. Carrying on, with for example 2 3 �cl �al 24 .l:!.bc4 .l:!.8a5 2 5 �c6 .l:!.xcl+ 26 �xcl �xc6 2 7 �xc6 �al+ 2 8 'it>h 2 'it>f8, Bl ack can find simplification, and even the better pawn structure. After all, if Black can keep his passed b-pawn, and White's king is a long way from the queen side, it i s Black who h as the bet­ ter ch ances. In addition, White's dou­ bled pawns are often a slight weak-

ness, and m aybe something more sig­ nificant. It would be rare in an end­ g ame for pawn s on f2, g2 and h2 (or h 3 ) to be less useful than pawn s on f2, g 2 and g 3 . Back i n the m ain line, and pl ay con ­ tinued with 2 3 �e7 �d 5 24 �al tiJf6.

In my contemporary notes, I sug ­ gested that " Black has played actively, and the position should probably be a draw." Thi s now seems wildly optimis­ tic. Compare this diagram and the test position, and it is clear that Bl ack has gone downhill. I now m ade several slight mistakes before the time control, and these all added up to a losing posi­ tion . Then unexpectedly h e missed something, in the quickpl ay finish, and suddenly I h ad the chance of holding the draw, but I was too short of time to see this. So there are still some more question s to follow in Test 11.4. After pl aying through several lines, it seem s th at Black is slightly worse, and will h ave to work h ard and accu­ rately to hold the balance. After 2 5 tiJe5, I played the weak

93

Why we L o s e a t C h e s s 2 5 ...I;!.a5?!, and soon fell into difficul­ ties after 26 I;!.xa 5 bxa 5 27 I;!. a 1, when my pawn soon fell, and I h ad to scrab­ ble around for any sort of compensa­ tion . Possibly I could h ave tried to re­ engage my pieces with 2S ... I;!.dd8, but even here White is standing well after 26 I;!.fc1 4:JdS . Then after 27 'iVb7 ! White has taken control of the seventh rank, and this is why Black should h ave tried 21 ... I;!.a7 much earlier.

Test 8.3 G.Wall-C.Crouch

B riti s h League (4NC L)

2007

We h ave seen thi s position before in Test 2 . 3 . Both players h ave pl ayed slightly in accurately, but it is now about equal . I continued to develop with 14... iLd7, which at first might look a slightly pas­ sive set-up for the bishops, but White's bishops h ave even less development opportunity. Black needs to get his rooks working together as soon as possible.

94

White continued his plan of attack­ ing on the kingside, with 15 h4. I h ave needed to con serve my energy in term s of brain power. Translation : I pl ayed l azily, because of my health at the time. If there is a n atural and obvious move, I would play quickly. Often thi s i s by no means a bad idea, for any pl ayer, but it i s important to h ave a sixth sense of feeling that on certain occasion s the pl ayer must slow down, whether it is a sen se of danger, or whether, as here, there is an advantage to be g ained from finding the less obvious move. Here I pl ayed 15 ... h6, pushing the at­ tackers quickly away, and gaining some time to try to develop on good squares. I suspect th at in my younger days, I would h ave preferred l S .. .fS ! ? 16 exf6 iLxf6, quickly opening up the centre, and counterattacking . Both king s are still in the centre, but the important difference, in term s of trying to take the initiative, is that Bl ack is able to castle quickly, bringing the rooks into the centre, whereas White's bishops are blocking the rooks and king s .

Te s t E ig h t Here 17 ttJxe6? �d6 18 fS gxfs 19 �S+ �e7 favours Black, but 17 ttJe4 iLe7 18 hS leads to sh arp pl ay. Black could continue either with h and-to­ hand pawn fighting with 18 ... e S , or consolidate his king side with 18 ... �g 8. Either seem s reasonable. My notes at the time stated that: "My regret i s not so much th at I did not play this, m ore th at I did n ot even con­ sider it." Probably the alternative move is interesting, but not necessarily bet­ ter. The g ame continued with quiet po­ sitional play, with 16 ttJe4 ..tc6. I sensed that if I was trying to overprotect my­ self by defending the pawn on c4, I would lose m omentum with my pieces. If now 17 iLxC4 ttJb6 18 iLd3 ttJxa4, and Black m ay h ave a slight edge. So in stead 17 'ife2, and now it is time to protect the c-pawn with 17 b5. White simplified with 18 axb5 axb5 19 l::t x a8+ ..txa8 20 g3 h 5 . ...

th at for the time being, the emph asis will be on the pieces. The knights on both sides are on good squares, but I can only take advantage of thi s after careless play by the opponent. I was more interested in term s of taking the initiative, with the queen and bishops, and m aking use of the a8-h l and a7-g 1 diagonals. What I did not like, though, was the positioning of my kin g . If Black were to be castled king side, White could possibly regenerate a pawn at­ tack, with g4. If my king did not move, then obviously I could not bring the rook into play. Thi s leaves the possibil ­ ity of bringing the king to the queen­ side, probably safe enough, but al so losing time. First, both sides continue their de­ velopment, with 2 1 iLg2 ..tc6 22 0-0, reaching the quiz position, and now it was time for me to decide what to do with th e king .

2 2 0-0 ( C } keeps the king away from the open queenside, but after 2 3 g4 hxg4 2 4 �xg4 �g7, followed by ... l::th 8, Black h as partially covered . . .

All the pawn s o n both sides are well covered, and there is unlikely to be a pawn rush on either side. This suggests

95

Why we L o s e a t C h e s s White's attack. White can, of course, try to continue the attack with 2 5 h 5 .l:th 8 2 6 hxg6 l:th4 2 7 'iWg4 fxg 6 28 lbg 5, and h e i s better. This is the sort of position that I was worried about, and decided to avoid. So it is a simple king move, but where? 22 ri;j8! (B) is best, and then de­ fending with ... ri;g 7.

pl ayed 2 3 ... l:ta8, still with equality, but in stead continued wastin g time with the king, with 23 ri;c8?, and then 24 'it'h2 'it'b7?! (m aybe 24 ... .l:!.d8 ! ?, although White is now better). I h ad completely missed the force of 2 5 b 3 ! cxb3. ...

••.

Black h as now gained a full tempo for manoeuvre when compared with ... 0-0, ... ri;g 7 and ... .l:th 8, and is safe, with ch ances of an edge. If, for example, 2 3 l:tdl ri;g 7 24 i. e 3 lbxe3 2 5 "iVxe3 .l:td8, and Bl ack is h appy, although at the moment it i s too early to claim an edge. My text move, 22 ...�d7?! (A), i s bad, and it gets even worse when Black con­ tinues with the king m anoeuvre to b7. It i s not so much that the king's final square is necessarily worse, but rather th at Bl ack has foregone two decent m oves with other pieces. At the end of his plan, White ends up with a signifi­ cant advantage. After 2 3 l:tdl, Black could h ave

96

Now h e tried 26 C4?!, and it looked highly effective, but the quieter 26 �2 ! ?, with a positional grind, seem s better in the longe term, with 2 6 ... .l:i.a8 27 "iVxb3 .l:!.a4 (otherwise c4 for White) 28 lbg 5 i.xg 5 29 hxg 5, and Black will h ave to work h ard to hold the bal ance. Play continued with 26 ... bxC4 27 �xc4, and now Black's king i s exposed to the pieces.

Tes t E ig h t I h ad to defend the b-pawn with 27 JWb6 for as long as possible, other­ wise my position would collapse. I was hoping th at m aybe I h ad chances of scrambling for a draw. It was now White, rather than Black, who started to take the initiative on the long di­ agonal, with 28 ttJC3 .l:.d8 (if 28 ... b2?, then simply 29 �xb2). Thi s i s a typical sort of frustration position for the attacker. There are many obvious choices to play for a win, but nothing quite seem s to work, with the position fizzling out to a draw on best play. Wall goes for the apparently deci sive fin al effort - and loses! ..

White must remember h e i s a pawn down, and so he must move quickly. 29 J::!. d 3 ! ? leads to tactical play, but soon ends up in a drawn endgame after 29 ... b2 30 �xb2 'ilVxb2 3 1 ttJxdS exdS 3 2 J::!.b 3+ W C 7 3 3 'ilVxc6+ Wxc6 34 .l:.xb2. Black cannot afford to lose the isol ated d-pawn, but White i s unable to force the pieces away. An opposite-col oured bishop endg ame with rooks. There i s no obvious improvement,

so thi s position seem s at least pl ayable for Black. Now 29 ttJxd S exd s led to a critical position.

White can try 30 .l:.xdS, and if 30 ... b2?? 31 .l:.b S ! . Entertaining vi sual chess. Instead, 30 ... �xdS 3 1 i.. x dS+ J::i. x dS 3 2 'ilVxdS+ Wc8 33 'ilVxf7 b2 34 'ilVC4+ 'it>b7 3S 'ilVe4+ Wb8 3 6 �xb2+ 'ilVxb2 3 7 'it>h 3, and various assorted checking lines, leaves White under pressure, with only two pawn s versus the bishop. Maybe White can still hold on, or m aybe not, but it is difficult. The best line is the ultra-positional 30 'ilVd3 ! ! , which probably neither player would h ave con sidered over the board, ignoring Bl ack's two isolated passed pawns. White's pieces are now far more centralized than Black's, and his extra kingside pawn is much more effective th an Black's pawn s, even though he is a pawn down. Black would be unable to keep both isolated pawn s together, and when one drops, ' White would still keep the positional advan ­ tage. Black would suffer the problem of th e isolated pawn, while White would

97

Why we L o s e a t C h e s s still b e able t o take advantage of his extra king side pawn, with e6 and/or fS pawn breaks. In stead, Wall blundered with 30 ii.xd 5??, and now 30 .. J:txd 5 ! 3 1 1::tx d 5 b2, winning for Black (and indeed 31 ... 'ii'f2+ 32 c,t>h 3 'ii'f3 is even quicker). The most likely explan ation for White's blunder, perh aps, was that he saw both of his captures on dS as leading to a likely advantage for him after a double­ exch ange (indeed a triple-exch ange), but missed that in playing the wrong move order, his opponent h ad the op­ portunity of an intermediate m ove. A common blunder, which we h ave seen many times in this volume. Here Wall loses the g ame. After 3 2 ii.xb2 'ii'x b2+ 3 3 c,t>h3 'ii' b l 34 c,t>h2 'ii'e l, White's rook cannot move, and Bl ack can capture the ex­ ch ange at leisure, staying the exch ange up. On 3 5 'ii' b 3+ c,t>C7 3 6 'ii'd l 'ii'f 2+ 3 7 c,t>h3 ii.xd 5 38 'ii'x d 5 'ii'g l I knew that I would be given a long string of checks, but I felt I would survive the perpetual and win .

So it proved, although naturally it took a long time for Black to take com­ plete contro1. Play eventually finished with 39 'ii'c4+ c,t>b6 40 'ii' b 3+ c,t>a5 (40 ... c,t>cS i s more efficient) 41 'ii'C 3+ c,t>b5 42 'ii'd 3+ c,t>a5 43 'ii'd 2+ ii.b4 44 'ii'd 8+ c,t>a4 45 'ii'd 7+ c,t>a3 46 'ii'd 3+ c,t>b2 47 'ii'e 2+ c,t>c3 48 'iWf3+ c,t>d4 49 'iWc6 'ii'fl+ 50 'it'h2 'ii'e 2+ 51 c,t>h3 'ii'g 4+ 52 c,t>h2 'ii'e 6 5 3 'ii'c 2 'ii'C4 54 'ii'f 2+ �d 5 5 5 'ii'f 3+ 'it'e6 56 'ii' b 7 ii.a 5 57 'ii' b l 'it'e7 58 'ii' b 7+ 'ii'C 7 59 'ii'a 6 ii.b6 60 'ii'a 8 'ii'C 5 61 c,t>h3 'ii'g l (now Black's king finds breathing space on g7) 62 'iWf3 c,t>f8 63 g4 hxg4+ 64 'ii'x g4 'ii'f l+ 6 5 c,t>h2 ii.e3 0-1

Test 8.4 S.Nurmohamed-C.Crouch

Thames Va l ley League

2006

Following on from Test 3 . 1, this tim e the puzzle is to find the winning idea for my opponent. I h ave pl ayed dreadfully so far, and he should be able to find good ch ances of a substanti al

98

Te s t E ig h t advantage just with good and logical pl ay. There might well be a forced win somewhere, though, without any chance of Black wriggling out. Nurmoh amed immediately pl ayed for a direct attack again st my king, with 2 2 .tb4?! (A), and it looks ex­ tremely threatening, but Black was able to squeeze out with 22 :�a4! 2 3 ttJd6+ �f8, and now White h as no time for a winning discovered attack, as White's bishop is threatened. ••

We return to thi s later i n Test 10. 2. White i s clearly not worse, but as pl ay continued, any slight mi stake by White could easily turn m atters around. Perh aps the easiest way of demon­ strating how Bl ack h as gone wrong over the l ast few moves i s by showing wh at i s h appening after the best move, 22 �g3! (B). The queen and bishop are no longer tied down, allowing them to maintain maximum flexibility. Black's king i s stuck in the centre, whereas o n g 8 rather th an e8 the king i s much safer. It i s not just solely the question of the king.

After 24 .tC5 'Yi'c6 2 5 'iVe3 �g8 Black has covered the immediate danger, and was even able to win, after some ten se pl ay. Wh at would finally decide after 2 2 'Yi'g 3 i s that Black cannot protect his g 7 pawn . If White invades there, the rest of Black's position crumbles. 22 ... g6 2 3 ttJxb2 axb2 24 �e s, for example, takes complete control for White on the b2h 8 and b4-f8 diagonals, with the bl ack king being stuck in the middle. Or 22 ... l:1.g8 23 eS ttJdS 24 .txh 7, and White joins in on a different diagonal. Let us return to the main line, and

99

Why we L o s e a t C h e s s the diagram after 2 S .. :>t>g8. Then he played 26 es?, natural enough, but I had expected 26 .i.bS ! '¥Va8 and now 2 7 eS, which i s more ac­ curate timing. White then doesn 't h ave to worry about the bishop on c S .

Bl ack has a coupl e o f altern atives h ere, but neither holds the balance. If 2 7 . ..t2Jds, the move I was thinking of, then 2 8 '¥Vf3 fS 29 exf6 .i.xf6, but now 30 J::t C 1 ! , and Black is effectively in zugzwan g . Or 27 ... tLJg4 28 '¥Vg s h S 29 '¥Ve7 ! (29 tLJxf7? ! i s good for entertain-

100

m ent, but after 29 ... .i.a6 30 tLJxh 8 �xbS 3 1 tLJg6 .i.xfl 3 2 tLJe7+ ..t>h 8 3 3 '¥Vxh S+ tLJh 6 34 'iit xfl �e4 Black holds) 29 ... tLJxes 30 f4 '¥VdS 31 fxe s '¥VxcS+ 3 2 'iith 1, and White h a s a clearly winning advantage. It could be argued that since White was winning anyway after either 22 .i.b4? ! or 22 �g 3, why should the anno­ tator be criticizing either m ove? In the long run, Bl ack i s dead. The question is really one of practical aesthetics. The pl ayer with a winning advantage needs to convert it into a win as quickly and clearly as possible. If the attacker messes around, Black can mess around himself, muddying the waters. Then there might be chances of counterplay. In practical term s, Black would lose very quickly after 22 �g 3, and would h ave slight practical chances after 26 .i.b s, if somehow White miscalculates, but in the g ame, with 26 e s ?, the posi­ tion was suddenly messy.

Test N i n e 9.1 White to play

9 .3 White to play

A) 24 llb 5 8) 24 'iWe2 C) Something el se?

A) 2 8 tiJd3 8) 28 tiJc4 C) Something el se?

9 .2 White to play

9 .4 Black to play

A) 2 5 e 5 8 ) 2 5 .l:!.f1 C) Something else?

A) 2 5 ... ..txh l 8) 2 5 ... �C2 C) Something el se?

10 1

Why we L o s e a t C h e s s

Test 9.1 C.Crouch-J.Cox

London League 2006 There is plenty of sh arp and un ­ usual play in this opening, and inevita­ bly both pl ayers make mi stakes, espe­ cially given that we were pl aying an evening game in the London League, where time limits are quick. Quite pos­ sibly a deep theoretical an alysis m ay pinpoint a suggested improvement in the opening part of the g ame. There was always a suspicion in thi s g ame, after all, th at in a complicated line White was quite often in danger of ending up worse, and this should not happen . It i s Black, not White, who has to work h ard to equalize. I do not wish to concentrate on Nimzo-Indian opening theory with much depth, trying to assess sub­ variation s in move A12 1, or wh atever, on move 8 or 11 or 14. Thi s book is con ­ centrating far more on searching for any clear mistakes th at players should be capable of avoiding over the board. The more interesting question, from our point of view, is not opening the­ ory, but rather, what mistakes are made over the board, and h ow such mistakes can be eradicated. We start with a Nimzo-Indian, with 1 d4 tiJf6 2 c4 e6 3 tiJC3 .i.b4 4 e3 c5 5 .i.d3 .i.xC3+ 6 bXc3 d6 7 tiJe2 e5 8 0-0 e4 9 .i.c2 .i.e6. Cox was playing quickly, so I was

10 2

sure th at he h ad experience in this po­ sition. In contrast, this was new to me. Clearly h e was deliberately provoking my dS push, and I decided th at here I should be provoked, in return setting up counterpl ay on e4.

So, 10 d5 .i.f5, and h ere I immedi­ ately played 11 f3 !? My opponent sug ­ gested afterwards th at White could effectively g ain a tempo with 11 tiJg 3 ! ? .i.g6 12 f3 exf3 1 3 gxf3 . This would cer­ tainly be so if Bl ack were to exch ange on c2, but if h e were to wait, h e can try som ething different, maybe 13 .. :iVaS ! ? I n the g ame, Bl ack played 1 1... exf3. My original intention was 12 .i.xfS fxe2 1 3 �xe2, with a good bishop-pair except th at when I looked more closely, I started to wonder wh at I was going to do with the dark-squared bishop, and al so how White could prevent Bl ack from moving the knight to a strong square on e S . So I ch anged my mind, preferring to keep the pawn structure more flexible with 12 gxf3 .i.xc2 13 �xc2 tiJbd7. We h ave very close to an equal posi-

Te s t N i n e tion . I like the extra pawn in the centre, but I am worried about the doubled isolated pawns.

I am not too worried about the slight exposure of White's king, but I am not quite sure wh at to do with the bishop. I played 14 tiJg3, which might or might not be fully accurate. I do not know. The computer suggests th at 14 e4, 14 l:!.bl and 14 �3 are al so equal . Any of these might al so be equally good, very slightly better, or very slightly worse. Probably, as we sh all soon see, 14 e4 would lead to a tran s­ position . For myself, it h ardly seem s worth­ while examining extremely subtle dif­ ferences when analysing afterwards. Most players will tend to g ain more from looking for more basic mistakes, and trying if possible to cut down on simil ar types of error. If I were to reach thi s position again, I suspect that I might well be thinking of 14 l:!.bl ! ? in stead, forcing Bl ack to con sider wh at to do with his b-pawn . Thi s i s a reasonable possibility, not

necessarily a clear improvement. Black tried 14 ... g6, to cover any tiJfS attack. Then 15 e4 �e7. Of course there could be a tran sposition with 14 e4 �e7 15 tiJg 3 g 6, and thi s in m any ways looks the more n atural way of playing it, but it would end up the same. White now h as an obvious way of taking the initiative with 16 f4. This seemed so obvious and natural a way of playing such a move, that I did not con sider any alternatives. The 4V2 pawn push in the centre looked good, provided I do not create too m any extra weaknesses for my king . A reality check suggests that, ac­ cording to the computer, there are at least a dozen alternatives, probably at least equal, and in some cases even better. 16 l:!.bl ! is suggested as a slight improvement, forcing Black to decide wh at to do with his b-pawn . If 16 ... tiJb6, then 17 a4! , and if 17 ... tiJxC4?! 18 �3 tiJb6 19 as, with advantage to White. Maybe at about the fourth attempt of an alysing this position, this seem s the best, with ch ances of keeping an edge. Backtracking a couple of moves, maybe Bl ack could h ave developed more sol­ idly with ... �C7, rather th an ... g 6 and ... �e7. Who knows? Cox plays combatively, with 16 . . 0-0-0!? There are several other choices, maybe castling on the oth er side, with 16 ... 0-0. Now there is an obvious possibility of the centre blowing up with a pawn sacrifice on e S . Or maybe White can .

103

Why we L o s e a t C h e s s hold the balance, perh aps only tempo­ rarily if he has a better occasion to break with e S .

I did not h ave t o think about thi s position too h ard. 17 e s ! is a highly thematic central pawn sacrifice, in such lines as the Modern Benoni and the Four Pawn s Attack in the King's Indian . White gives up a pawn, win ­ n i n g Black's d-pawn, but losing his e ­ and f-pawn s, quickly bringing h i s bishop and rook into active play, and setting up a strong passed d-pawn . Quite often there is a question of tim ­ i n g i n such pawn explosion s. I f you don 't sacrifice or g ambit immediately, it is often best not to do so at all . I did not like the idea of Black restraining White's e-pawn with ... J::i. d eS or .. .lUeS or ... ctJg4; and earlier, the only sen sible continuation to my f4 push was an e S push . So i t h a s t o b e done immediately. Several months later, I wondered whether I could have improved my game, by developing with 17 l1b1. The idea is that when Black too plays a quiet move, White's rook is now on an active

104

file, and he can play eS with an extra attacking piece in play. This seems to make sense until one realizes that if Black can prevent or reduce the impact of eS, White would then be under pres­ sure. Therefore 17 ... ctJh S ! , a useful move in the game as well. Whether White ex­ changes on h S , allowing Black the open g-file, or whether he allows the ex­ change on g3, White is under positional pressure. If he cannot find any good gambit play on eS, his e4-pawn is weak, once the knights have gone. So again, White h as to gambit im­ mediately, or not at all . Back t o the sacrifice, and 17 ... dxes 18 fxes.

Now the captures on e S seem far too dangerous, and I expected a knight move. Indeed, lS .. :iWxes ? gives White far too big an attack after 19 ..tf4 fie7 20 fia4. We see an example of the '4V2 pawn attack', the standard four pawn attack, but with a fifth doubled pawn on c3. Thi s pawn covers the d4-square again st the queen, m aking fewer es­ cape routes from e S .

Te s t N i n e Here I was expecting 18 ... lDg4, to which I planned to play 19 lDe4 lDgxe s (19 ... lDdxeS 20 �g s f6 2 1 �xf6 lDxf6 2 2 .l:txf6 tran sposes) 20 �g s f6 2 1 �xf6 lDxf6 2 2 .l:txf6. I assumed this was go­ ing to be slightly better for White, al ­ though now it seem s th at it i s only equal . For example, 22 ... .l:thf8 23 .l:Iafl l::t xf6 24 .l:txf6 .l:tf8 2 S d6 �d8 26 l::t xf8 'iYxf8 2 7 'iYa4 lDf3+ 28 cj;>g 2 lDh4+ 29 cj;>g l, and Bl ack would be advised to take the draw. 19 �f4! ? is also to be considered. In stead, Cox went for counterattack with 1S ... lDh S ! ?

Now 19 lDxh S gxh s 20 i.f4 lDxes 2 1 �fS+ lDd7 2 2 .l:Iae1 i s probably equal, for example, with 2 2 ... l::th g8+ 2 3 cj;>h 1 "iVh4 24 .l:te4 'i¥f6 2 S 'i¥xf6 lDxf6 2 6 l::t e 2. I chose the m ore aggressive idea, 19 lDe4!. White keeps a slight edge after 19 ... lDxe s 20 �g s f6 21 .l:txf6. Black tried in stead 19 ... �xes?!, when he ad­ mitted afterwards th at he h ad mi ssed my next move, 20 lDgS !. Thi s i s the high point of my attack. I

win the exch ange because of the un ­ stoppable knight threat on f7. Al so, take note of the position of the bl ack queen, on an apparently open board.

The only way the queen can move i s backwards, th anks in part t o White's doubled pawn s. Black would, of course, welcome the chance of a queen check from g4, but there i s no way to attack th at square, even in two or more moves. White's pieces and pawn s are well coordinated. Cox thought for a long time, and decided th at while the rook h as to go, he needs to keep the extra pawn, with 20 ... fS 2 1 lDf7 �f6 22 lDxhS .l:IxhS.

105

Why we L o s e a t Ch e s s W e have just gone through a phase of sacrifice and counter-sacrifice, with the initiative swinging from one side to th e other. White has sacrificed a pawn, in return obtaining good piece activity; Black has returned the exch ange, eliminating White's more attacking pieces. Now Bl ack has the more effec­ tive piece and pawn form ation, even though he is down in m aterial . White is not under immediate attack on the kingside, but h e h as to be careful about wh at might h appen in the future. I played the natural developing move, 2 3 J::i. b l!?, trying to create pres­ sure on the queenside. I underesti ­ mated his reply, 2 3 :iVa6!, when Black has covered all his queen side pieces, and threaten s to take the critical C4pawn, and with all likelihood, White's open lines in the centre and king side. ..

At the time, I felt no real option other th an to block the c4-a6 diagonal with 24 J::i. b S (A), which at least squeezes Black's queen to the edge, but it looked far from convincing. If at all possible, the defender needs

106

to centralize, covering king side, queen­ side and centre. Indeed, 24 'ifie2! (B) is best. Black's queen i s on the edge, so White can create counterpl ay on the rest of the board. I was worried that Black was about to g ain time by attack­ ing the queen, with 24 ... 4:Jhf6, but 2 5 .l:i.e1 ! covers the immediate threat. If 25 ... J::i. e 8??, then obviously 26 "iVxe8+. White is fighting the pressure, not merely defending. Instead, if 2 5 ... 4:Je4 2 6 J::i.b 3 (there are alternatives, probably ending up level) 2 6 .. .tt:Jb6 2 7 J::i. a 3 "ivxc4 28 "iVxC4 4:JxC4 29 J::i. x a7 �b8 30 .l:i.a4 b5 31 ..if4+ �b7 3 2 .l:tb1 4:Jb6 3 3 .l:i.xe4 fxe4 34 .l:txb5 J::i. c 8 35 J::i.b 3, with a likely draw. Should we be disappointed that White cannot find an edg e ? Probably not. The position is starting off as finely bal anced, and if both players keep the delicate bal ance, then the proper result should be a draw. It is not too difficult to appreciate that in my an alysi s later, I h ave tried several possible alternatives, either to keep an edge for White, or if not, then clear equality. This is the best line I h ave tried, and it keeps play sim­ ple. Probably we should add that a move earlier (omitting 23 J::i.b 1 "iVa6), 2 3 "iVe2 ! ? i s certainly a choice for White. Possibly Bl ack could try 23 ... 4:Je s 24 .lte3 b 6 ! ?, followed by a l ater .. .f4, with reasonable compen sation for the ex­ change sacrifice. We now return to the m ain line, and 24 J::i. b S?! J::i. e S.

Te s t N i n e

It i s starting to be difficult for White to coordinate his pieces, especially as Bl ack is intending to play .. J:!.e4. I de­ cided it was time to set up a counter­ exch ange sacrifice with 25 iYf2, and a possible .l:i.xcS+. Black played 25 ... ttJhf6?!, which is natural and logi­ cal enough, centralizing a knight on the edge.

Even so, inaccuracies are starting to creep in on both sides, in the l ast ten moves before the tim e control i s reached. 2S ... .l:i.e4! would h ave been highly ef­ fective, taking control of the fifth rank, with an attack of the pawn on c4, a

possibly nasty check on g4, and also an effective defence again st i..f4. The rook's remit goes further after 2 6 .l:i.xcS+ ttJxcS 2 7 'iYxcS+ �d7 2 8 iYbS+? (28 i.a3 iYb6 is better, but Black i s bet­ ter) 28 ... 'iYxbs 29 cxb S .l:i.a4, cutting open White's pawn s. Then 26 i..f4!? was a difficult deci­ sion, and, of course, tim e was getting short. The bishop i s on a good diagonal, but it can al so be threatened and kicked out. The alternative is an immediate ex­ ch ange sacrifice, with 26 .l:i.xcS+ ttJxcS 27 �xcS+ 'it>d7 ! (rather than 27 ... �b8 28 i..f4+ 'it>a8 29 �C7, when White is at least equal, and quite likely somewh at better).

If now 28 i..f4, we see a good reason why Black has del ayed ... .l:i.e4. He plays 28 ... .l:i.c8, knocking out White's queen and bishop attack, winning a doubled pawn . White can try in stead 28 i.. a 3, but he still has not got a genuine attack, and after 28 ... ttJg4 Black is better, with thoughts of ... ttJe3 or ... ttJe s . Before too long, White will notice the lack

107

Why we L o s e a t C h e s s of the g-pawn, whether i n a middle­ game, or, with the passed f-pawn com­ ing through, the endgame. After the text move, Black blocked off White's queen side attack by 26 ... b6, avoiding any sacrifice on c 5 . 26 ... tDe4? would h ave been too combative, and would lose after 27 Itel g 5 ? 28 ':xe4 Itxe4 29 .i:!.xC5+ tDxC5 30 'iVxc5+ d7 3 1 'iVC7+ �e8 3 2 'iVc8+ �f7 3 3 'iVxf5+. White pl ayed 27 .tg3, with the none-too-subtle thought of placing the queen in front on the diagonal, and trying to give m ate. I did not like the idea of giving up a few pawns with 2 7 Itel Itxel+ 2 8 'iVxel 'iVxa2 29 'iVe6 'ivxc4 30 'iVc6+ d8. There might be a coupl e more checks, but then wh at? If, for example, 3 1 .tC7+ �e8 3 2 'iVd6+ �f8 3 3 .td6+ �g 7 34 'iVe7+ �h 6, and White run s out of checks. Al so, if 3 1 d6 �e8, and again the king escapes.

In the g ame, Black forces the re­ moval of White's queen, with 27 tDe4, but where should the queen move to? Thi s is Test 1 2 . 1 ••.

108

Test 9.2 C.Crouch-P.Gait

H i l l i ngdon League 2006

As we saw in Test 4.2, my opponent h ad just missed a winning opportunity, which I only bel atedly saw myself. We are now back into norm al pl ay, with 17 .te2, and I was slightly better, in term s of technical chess analysi s. The psy­ chology of the g ame was highly against me, though, as I was ultra-nervous throughout the rest of the g ame, wor­ ried that every move I h ad to think about was a blunder, and th at every move, good or bad, was on an eggshell . I was able to convince myself, though, th at the first move was fine. He played norm ally with 17 ... h6, then came a few reasonable moves, with 18 g4 .tg6 19 g5 hxg5 20 tDxg5 tDh7 2 1 .txe4 'iVf6. Clearly we are now in a ph ase of tactics. I h ave won a pawn, but it will take tim e to con soli­ date, given the openness of the centre, and the ten sion on the bl-h7 diagon al .

Te s t N i n e

In my notes just after the g ame, I was critical of 2 2 f3, missing the idea of his next move. 22 i.xg 6 ! �xg 6+ 2 3 ttJe4 is certainly easier to pl ay for White. Whether it is 'objectively' better, in the sense that it i s the best outcom e after the best m oves by both sides, is not so clear. G ait sacrificed the exch ange with 22 ...l::i.x e4! 23 fxe4 i.e7, creating confu­ sion in the position . I am m aterial ahead, and I h ave the bigger pawn cen­ tre, but I now h ave problems with my minor pieces, with my knight being precariously pinned. It is possible too th at one of White's central pawns will drop, allowing Black to continue the counterpl ay el sewhere. The position is extremely difficult to h andle, and I recognized this at the time. My next move, 24 'it'al, seem s correct, escaping any pin s and checks on the light-squared diagonal, but my follow-up move was a mistake. In stead, 24 l::i.fl looks promising, but after 24 .. :iYxd4 2 5 i.b4 i.xg 5 26 �xh 7+ i.xh 7 2 7 l::i. x d4, Black keeps in touch

with an unexpected combination of fork and skewer, with 2 7 ... C 5 ! 2 8 i.xc5 i.e3, keeping the m aterial bal ance. Not, of course, 29 l:td5 ?, allowing a fork with the other bishop with 29 ... i.xe4+. There is al so another possible bishop fork, with 24 .l:!.gl 'ilVxd4 25 i.b4 i.xg 5 26 'ilVxh 7+ 'it'xh 7 2 7 l:txd4 .ie3 ! , level . Clearly it is not a straightforward win after Black h as given up his rook. After 24 ... ttJxg5, we h ave reached the test position. The next move ended up in the difference between an advan ­ tage and a quick loss.

I pl ayed 2 5 e5? (A), overlooking th at after 2 5 ...'ilVf5, the geometry is again in favour of Black after 2 6 l::i.fl 'ilVxh 3 ! 27 i.xg 5 'ilVxh4 28 'ilVxh4 i.xh4. Nominally, Bl ack's m aterial advantage is slight (two bishops versus rook and pawn), but the two bishops are powerful, with no good way of combating them, and no doubt the passed g-pawn will even­ tually push through . S o I changed mid-stream with 2 6 i.xg5 i.xg5, when I a m worse, but with still a chance of making a battl e. My

10 9

Why we L o s e a t C h e s s next move was a blunder, though, which forced me to resign quickly, as we'll see in Test lOA. Earlier, 25 '1J./l! (B) is much better, with a clear advantage for White after 2 S .. :�xd4 26 �xg s �xg s 27 "iiix g s "iiix C4 28 e s ! (better th an the obvious 28 '1J.d8+). Black h as only a pawn in re­ turn for the exch ange, and his bishop is ineffective, while White's m ajor pieces, plus a possible advance with the h ­ pawn, create great problems. Again, after my earlier big mistake, I got nervous, and made further mis­ takes and blunders.

Test 93 C.Crouch-R.Granat

Britis h League (4NCL)

2007

Continuing from Test S.2, there fol­ lowed some obvious development plans with the next couple of moves, with 20 '1J.c1 c6 21 '1J.hd1 .ixb2 22 'it'xb2. His next move, 22 ... 'it'b8, i s again obvi­ ous, but perh aps not quite accurate,

1 10

involving a slight loss of time. White's knight on dS looks strong, but does not immediately threaten anything, and so there is no need to kick it out just now. 22 ... b 6 ! improves, with the idea of ... 'it'b7 (a more active square), and ... 4JcS . If White were to try 2 3 4Jb4, hit­ ting the c6-pawn, then after 23 ... 4JcS, Black h as clearly gained time in com ­ parison with the line in the game. This pawn move may at first look anti­ positional, weakening the queenside pawn structure (especially the pawn on c6), but sometimes pawn pushes can create more th an compen sating strength s. Black's king and two knights work well together, with the h elp of the two pawn s covering squares on the fourth rank. In the game, White h ad to move the knight, 2 3 4Jb4, and Black played the useful but less aggressive 2 3 ...4Jb6. Looking on the computer afterwards, I felt slightly mystified th at White is given an edge h ere, whereas earlier on Black is given the edge. Maybe the posi­ tion i s in the end about equal, although over the board I was startin g to get worried, as his knights are moving to good and dangerous squares, notably with ... 4Ja4+. Black still, of course, has his passed pawn, a worry if we reach an endgame. I became over-cautious over the next few moves, and thi s af­ fected a critical choice a few moves on . We go quickly towards the test posi­ tion, with 24 'it'c2 'it'C7 25 4Jd 3 4JbC4 26 a4 4Jb6 and now 27 4Jb2, so that the

Tes t N i n e of the g ame. The intention for White here should h ave been to exch ange a pair of knights.

knight covers both a4 and c4.

Then 27 lbd S . I was perh aps too highly concerned with the idea of Bl ack being able to reach b4 with a knight, so I tried to cover with 28 lbd 3? (A), with the h ope perh aps that he would finish with a draw, with 2 8 ... lbb6 29 lbb2. He was more than h appy, though, to try to play for a win with 28 b6!, with his knights and pawn s coordinat­ ing well for both attack and defence. We shall examine the rest of the game later in Test 1 3 . 3 . 2 8 lbC4! (B) would h ave been much better, even giving chances of a slight edge for White. It i s well known th at a pair of bishops will often work better together than bishop and knight. In the game, for example, White would h ave enjoyed h aving the bishop-pair, cover­ ing squares of both colours (imagine a bishop on b2, for example, rather than a knight). What is often less fully exam ­ ined is the knight-pair. Sometimes it can be useful for the knights th at be­ tween them th ey can press again st squares of both colours, as in the rest ...

...

If now, for example, 28 ...lbxC4 29 �xc4, the bishop h as good play, whereas Black's knight accomplishes little. Black might try in stead 28 ... lbb4+ 29 'it'b2 b6 30 lbxas bxas 31 ii.C4 ttxd1 32 ttxd1 tte7, but he has lost coordina­ tion with his queenside pawn s, and White has the better minor piece. White in fact h as a slight edge, because of Bl ack's loss of time on move 2 2 . There are, of course, possible alter­ natives for Black, but it is not easy to find quick equality. At the time, I asked myself, "Why did I reject my simple and direct move? Partly because his knight on as was so much on a limb that it looked crazy to try to exch ange it. But his knight is use­ ful in th at White always needs to cover the b3- and c4-squares. Exch ange th at knight, use the light-squared bishop again, and White can start becoming active. Or at the very least he is not hoping for repetition ."

111

Why we L o s e a t C h e s s

Test 9.4 R.Randall-C.Crouch

London League 2006 was still feeling very unwell, slightly less th an two years after my stroke. I needed both physical exercise and mental stimulation to try to bring myself back to life. For the London League m atches th at year, I generally walked three miles to Edgware to catch some exercise, took a bus into town and then walked a fur­ ther mile or so to the Briti sh Museum . On the bus, I would catch up on some Sudoku puzzles. The bus journey from Edgware to Baker Street was about the right length of time to complete the exercise, although occasionally I would make a mistake, based on an incorrect assumption . This, of course, was frus­ trating. On th at day, I noticed I could not concentrate on the puzzle, suggesting perh aps that I was unlikely to be on form in my chess that evening. I am reason ably h appy with my early moves, but sure enough I made a couple of ridiculous blunders later on . Pl ay started with 1 e4 c5 2 ct:Jf3 e6 3 d4 cxd4 4 ct:Jxd4 ct:Jf6 5 ct:JC3 d6 6 g4 h6 7 h4 a6 8 �g2 g6 9 g5 hxg5 10 �xg5 �e7 11 �d2 ct:Jbd7 12 0-0-0 e5 13 ct:Jb3 b5 14 f3 �b7 15 ..t>bl b4 16 ct:Je2 as 17 �h3 l:!.a6, when Bl ack was doing fine, though I h ave yet to prove an advan­ tage.

112

White pl ayed 18 C3?, a common po­ sitional error - and I am sure I h ave done thi s myself m any times. Black is intending to rush his queen side pawns quickly, and th at can indeed be fright­ ening. White has to remember that Black is not thinking of promoting his pawn s, but rather he i s aiming for pawn-versus-pawn contact in front of the king, aiming to open up lines. Un­ fortunately, White's pawn push accel ­ erates the battl e, in favour of Black. 18 �xd7+! would h ave been better,

and if 18 .. :�Vxd7 ? 19 �xf6 �xf6 20 ct:JcS, winning the exchange, but 18 ... ct:Jxd7 19 l:!.hg l is about equal .

Te s t N i n e I h ad the luxury of setting up a mi­ nor piece sacrifice in order to break up White's centre pawn s, with 18 ... a4 19 lbbc1 lbxe4 20 fxe4 iLxe4+ 21 'it>a1 lbc5.

My original intention was naturally to capture the rook, 2S i..xhl (A), with thoughts of a significant plus. But my eyes were dim , and I could only see with h alf of one eye, and occasion ally do not 'see' things in my mind. Proba­ bly the biggest single problem for me when starting to resume playing chess after my stroke was that I was unable to see the whole board, both visually and metaphorically. Occasionally I saw ph antom s, thinking m aybe th at one or two of my pieces or pawn s were on different squares, rather than where they were actually at. I could not see the whole board, and because of re­ striction s of my vision, I could not see the bottom right-h and square. I h ad to rely on memory. My biggest blind spot was on the bottom right-h and corner, and here of course my queen on a8 was absolutely critical to my calculation s. In earlier years, I would h ave been able to see the position as a whole. In 2006, though, all I could do was remember th at my queen was on a8, simultaneously cov•..

Black's attack i s massive. 22 cxb4 i s t h e only move. In stead, 2 2 .i.xe7 lbb3+ 2 3 lbxb3 axb3 24 a3 'iVa8 would h ave forced m ate on the a-file. The problem of White's c3 push is not just th at Black has opened up the ... b4xC3 pawn ex­ ch ange, although in some cases it might h ave proved useful . The much bigger problem is th at White has badly weakened his b3-square, allowing po­ tenti ally winning tacti cs with ... lbb3+. So I continued with 22 lbb3+ 23 lbxb3 axb3 24 a 3 'iVa8, with a m ating threat on a3, and al so the threat of winning the rook on h i . Then White has to play 25 'iVC1. It was di sconcerting that Black was not already winning, and th at I h ad given up a piece for a couple of pawn s and an attack. For complete security, I would want to give checkm ate, or at least to gain material . ••.

1 13

Why we L o s e a t C h e s s ering both the h l-square and the a-file. It was difficult at the time to cover both the long diagonal plus a-file, as well as what was going on in the rest of the board. Indeed, I was un able to visualize the whole board. I could see part of the board, and such visualization is, of course, extremely important for chess pl ayers. I was unable as yet to find ways to visualize and see different parts of the board, but later I was able to learn how to improve on this. What I 'saw' was 2 S ... ..txh l 2 6 'iWcS+ 'iWxcs 2 7 l:!.xcS+ ..tdS 2S .l:i.xdS, m ate. Except this is a complete illusion . He cannot move his rook to cS. He can move his queen, 'iWcl-cS, but he cannot move the rook as well, l:tCl-CS, in a sin­ gle m ove. The queen and rook are not doubled on the c-file. Logically this i s absurd, since the queen i s firmly an ­ chored on cl, to protect the sacrifice on a3, and there i s no rook coverage on the c-file, just behind the queen . Maybe for some readers thi s might at first be of relevance only to partially­ sighted players. Try though to think of a complicated combination, when you h ave to think of possibilities h alf a dozen moves in advance, when som e of the pieces on either side h ave m oved on, and other pieces and pawn s h ave disappeared. The player would not be able to see the position in advance, as the pieces are not yet there. What i s

1 14

important i s to try to 'vi sualize', to imagine with cl arity wh at the position is a few moves along the line. My problem with my eyesight was certainly a significant problem in this g ame, but ultim ately the deci sive prob­ lem is visualization, and thi s was some­ thing I clearly h ad to work h ard on . I played 2 S ..tc2?? (8), inexplicable, until you know the full thought proc­ ess. I felt I h ad to cover th e c-file, but this was irrelevant and wasted pre­ cious time. ...

In the g ame, White played 26 l:!.d2, with advantage, and my bishop on c2 was not even very effective. Al as, there were blunders on both sides before White later won . We return to the g am e later in Test 13.1. The n atural 2 S ... ..txh l, as noted, is good, while preparing this with 25 /6 (C} 2 6 ..te3 ..txh l 2 7 l:!.xh l dS might even be fractionally better (avoiding 2S ... i.xh l 2 6 i.xe7 ! ?). ...

T e st Te n 10. 1 Black to play

10.3 Black to play

A) 2 S ... g 6 B) 2 S ... �d7 C) Something el se?

A) 26 . . .4:Jg S B) 2 6 ...�d7 C) Something el se?

10.2 Black to play

10.4 White to play

A) 26 ... 4:Jg4 B) 26 ... 4:JdS C) Something el se?

A) 2 7 �g4 B) 2 7 �g 3 C) Something else?

1 15

Why we L o s e a t C h e s s

Test 10.1 C.Morris-C.Crouch

B ritis h League (4NCL)

2006

As we observed in Test 6.1, White i s clearly better, Black h aving queenside weaknesses. He tried 2 3 Ite3?!. This keeps an edge but I was more anxious at the time about 23 e6 ! ?: for example, 23 .. J:te7 24 exf7+ Itfxf7 25 h4 g6 26 .l:txe7 .l:i.xe7 27 .l:tc3. It is not quite as impressive as it earlier looked, though, and White only h as a slight edge. Maybe there i s another improve­ ment? 23 tLlh4! looks better.

It might seem anti-positional at first, placing both knights on the edges, but the knights both threaten to move into the centre. If Black decides later that he has to pl ay ... b5, then the two knights on c5 and f5 will be mighty. Black can try 23 ... g6, but after 24 f4, White i s starting a king side pawn roller. White is much better. I was still under pressure in the game, and eventually hit on the plan of 2 3 .. J�e7! 24 tLlh4 ..ic8!, covering the f5square. Bl ack would not, of course, h ave h ad the tim e to defend in this way if White h ad tried 2 3 tLlh4 a move ear­ lier. White pl ayed 25 f4 (not 25 .l:i.xc6 ?? ..id7), and on my next move, I quickly undid my previous defensive play.

2s ... g6? (A) was a serious conces­ sion, and should only h ave been pl ayed if absolutely necessary. Before too long, I was being ground down, after 26 Itb3 bs 27 tLlcs tLlC7 28 Itf3 Wh7 29 fs gxfs 30 tLlb3. There were a couple of twists to follow, though, which we sh all see later in Test 14. 1.

1 16

Tes t Te n Black should earlier h ave played 25 .i.d7! (B), in stead. ...

Thi s is a difficult line and Black still has to defend carefully, but the alter­ native is the danger of a position al loss.

Test 10.2 S Nurmohamed-C.Crouch

Tha mes Va l ley League

Somehow he h as unravelled his pieces after an extremely unpromising start. Just an extra h alf-tempo, a logical but slightly l azy move by the opponent, can sometimes make a big difference. Play might continue with 26 f5. Then Black can try active counterplay with 26 ... c5 27 tLlC3 tLlC7 (27 ... cxd4 28 tLlxd5) 28 tLle2 C4 (28 ... cxd4 29 tLlxd4 tLlb5 30 tLlhf3 with a slight edge for White) 29 J:::tg 3, and White's kingside pawn push is more effective than Black's queenside pawn push. Black's king is in danger after a sacrifice on g 5 after, for example, 29 ... .l:!.ee8 30 f6 g5 31 tLlf3. 2 6 ... J:::t c 8! i s a tighter defence, with the hope th at if White's king side initia­ tive can eventually be slowed down, Black can at some stage set up coun ­ terplay with ... c 5 . Black is obviously not better, but it is not so clear that he is substantially worse. There i s a possible draw by repetition after 27 tLlc3 tLlC7 28 tLla4 tLla8, but White could still try for more.

2006

The good news from Test 8.4 is that now I h ave a reasonable position, hav­ ing been in severe trouble. Over the board, I felt I had reasonable chances, especially if he made any further slips in a complicated position . The com ­ puter analysis suggests that of the two main lines for Black (26 ... tLld5 and 26 ... tLlg4, both hitting the queen on e3), one is equal after best, and compli­ cated. play, while the other is al so compli cated but there is a win for White. In a quick time limit, I would not be expected to calculate the critical positions in depth . It would always be a case of positional sense, in stinct, and about as much as you could an alyse given the context of the ticking clock.

1 17

Why we L o s e a t C h e s s S o which is better? I concentrated on thinking of where the best counter­ attacking and defensive move for the knight is. I felt at the time that 26 ... ti'lg4?! (A) looked best, with the aim of taking White's bishop on c S . The alternative was 26. . .ti'ld5! (B), aiming to take the knight to one of the four central squares a move earlier. This seem s to be a draw. I was almost certainly worried about 27 �g s g6 28 ..Ite4!, and it looks like Black's king will be swallowed after the central knight has gone.

There are two saving resources for Black, ending up as a draw. Either 28 ... h6 2 9 ..Itxds hxg s 30 iLxc6 iLxe s ! , and Bl ack h as time t o give a perpetual with ... ..Itxh 2+. Or 2 8 .. :�WxcS 2 9 ..ItxdS, seemingly terminal , but Black h olds with the help of a couple of pins, with 29 .. .f6 ! 30 �xf6 �xd6 31 .l:tc1 �f8 3 2 'tWxf8+ �xf8 3 3 .l:txc8+ �g 7 34 l:Ixh 8 �xh 8 3 5 Ji.b7 ..Itxes 3 6 ..Itc8 �g7 3 7 ..Itxd7 �f6, ending up with a quiet draw in an opposite-coloured bishop ending .

1 18

2 7 .. .f6 is another possibility, with less bite for Bl ack than in the similar set-up in the main line with ... ti'lg4. There is a quick draw after 28 'iVh s g6 29 ..Itxg 6 �xcS 30 iLe8 �g 7 ! , forcing a perpetual after 3 1 �f7+ �h 6 3 2 'iVh S+. White could al so try 30 'iVh 6 ! ? (threat­ ening ..Itf7 m ate), and then 30 ... �xd6 3 1 exd6 hxg6 32 �xg 6+ �f8 with un­ clear play, with queen and pawn versus three minor pieces. My in stinct is to favour Black slightly, as White does not h ave much ammunition to favour the queen . The perpetual would seem preferabl e. 2 6 ... ti'ldS therefore gives three ways of m aintaining the bal ance, which would clearly h ave been progress for me after my poor opening. We now con sider the game's 26 ... ti'lg4?!, ambitious but not quite as steady. Then there is forced play with 27 �g5 f6.

Now Nurmoh amed aimed for a tac­ tical flourish with 28 �h 5? g6 29 ..Itxg6, but I h ad seen thi s before playing ... ti'lg4.

Te s t Te n Sometimes, especially in time trou­ ble, it is easier to envisage tactics on either side, rather th an to con sider quiet positional chess. Here 28 'iVxg4 'ifxC5 29 liJe8 ! gives White a clear ad­ vantage. Then 29 ... 'iWe7 30 liJxf6+ 'iitf8 3 1 liJxh 7+ 'iit g 8, with several ways of keeping the upper h and, for example, 3 2 �e4. I m anaged to escape, with 29 . . liJxes, covering the f7-square. .

White could still force a draw with 30 i.e4 �xc5 31 �e8+ 'it>g 7 32 'iVe7+ 'it>g 8 33 'ife8+, but he was still aiming to force a win. There is also a more imagi­ native draw after 30 i.xh7+!? 'iit g 7 ! (30. . .l:txh 7 ? 3 1 'iVe8+ 'it> g 7 3 2 liJf5+! exf5 33 i.f8+ 'it>g8 34 i.h 6 mate) 31 liJe8+ l:txe8 32 'i'xe8 'i'xC5 33 'iVg 8+ 'it>h 6 34 �8 liJf7 (34 .. .f5 ! ?) 35 �g8 liJe5, which would have been on just about the edge of my thoug ht processes. In stead he tried 30 i.f7+ 'iit g 7 3 1 i.e3, playable but more complicated. I t is now White who h as t o find a way of holding the draw after 31 ... h61, as seen well in advance.

All Black's pieces around his king are suddenly extremely well covered. And yes, I played much better in the second h alf of the game th an in the first h alf. White lost co-ordination after 3 2 liJe8+? 'iitf8 3 3 i.g6 'iit e 7 34 liJg7 i.b7, adjourned but White resigned before the resumption, (0-1). In stead, he could still have kept a draw with 3 2 J:td1 ! (protection for the knight) 3 2 ... i.b7 3 3 i.xh 6+ l:Ixh 6 34 liJe8+ 'iith 7 35 liJxf6+ 'iit g 7 3 6 liJe8+ �h 7 3 7 liJf6+.

Test 10.3 I.Lauterbach-C.Crouch

B ritish Lega ue (4NCL)

2007

As soon as I felt I h ad made myself comfortably equal, as we saw towards the end of Test 6.4, I made a comfort­ able but l azy move, and found myself worse. Sometimes it is difficult to read­ just when a player h as moved from being level to being worse. It is difficult

1 19

Why we L o s e a t Ch e s s t o work out whether a player should then aim for careful equality, or whether it is possible to play for more. Here I got it wrong .

Perh aps the tightest pl an for both sides is a quick repetition of moves with 26 :Wid7 (B) 27 iLh 3 �c6 28 iLg 2 �d7. This, over the board, would h ave been an unexpected finish, but it is al so logical enough. White needs to keep the bishop on the line of the queen, while Black wants to escape from the bishop. There are other examples of to-ing and fro-ing in thi s type of closed posi­ tion, and another repetition would be 27 ct:Je3 ct:Jg s 28 ct:Jf4 ct:Je4 29 iVC1 (29 iLxe4? ! .l:Ixe4 favours Bl ack) 29 ... iLc6 30 'it>f1 fS 31 iLh 3 ct:Jg s 3 2 �g2 ( 3 2 �xfS ? .l::t x e3 win s material for Bl ack) 3 2 ... ct:Jg S. I wanted to block out Black's bishop, with 26 . ct:JgS? (A), but in the geometry of chess, a knight cannot simultane­ ously defend both the light squares and the dark squares, and White was able to take over with 27 �f4!. If now the queen s get exchanged, Bl ack's dS-

pawn will soon collapse, for example, with 27 .. :�xf4 28 ct:Jxf4 .l:Ixe1+ 29 .l:Ixe1 ct:Je4 30 ct:Je3 ct:Jf7 31 .l::t e 2 ! , but not 31 ct:JfxdS ? ? �xdS 32 ct:JxdS ct:Jd2+. So I m ade a queen retreat, 27 .. :�d7, and 28 ct:Je3 left White with a signifi­ cant plus. I h ave, of course, lost time.

••

.

120

With 28 ct:Je4 I still felt reason ably comfortable, but can I even hold it? Certainly I needed to activate my knight in the centre. If I tried to play slowly with, for example, 28 ... ct:Jhf7 or 28 ... �c6, White would take control in the centre with 29 �fS, quite probably eventually winning the ds-pawn . •••

.

After 29 ct:JfS, there was a quick

Te s t Te n simplification with 29 ttJd2+ 30 'it'xd2 'it'xfs 3 1 ttJf4, and the prospect of fur­ ther simplification . The trouble is that the backward pawn on d5 will continue to be weak, before simplification, and perh aps even more dangerously, after simplification. ..•

.l:i.xe1+ 3 3 .l:i.xe1 'iYxa4 34 1i.e6+ ttJf7 3 5 ..txd5 1i.xd5 3 6 ttJxd5 ttJg 5 3 7 ttJe7+ 'it>h 7 3 8 'it'e2 i s not so clear) 32 ... .l:i.xe7 33 .l:!.e1 .l:i.xe1+ 34 'ilVxe1 ttJf7 35 1i.h 3 f5 36 ttJg6, and White is h eading for a win. Fin ally, we find 3 1.. . .l:!.xe1+ 3 2 .l:!.xe1 .l:txe1+ 33 'ilVxe1 1i.c6 34 ..th 3 'ilVe4 3 5 �xe4 dxe4 3 6 1i.e6+ ttJf7, and, of course, White wins, with a possible lively checkmate after 3 7 d5 1i.xa4 3 8 d 6 'it>f8 3 9 d 7 ttJd8 ? ! (39 . . . 1i.xd7 40 1i.xd7 lasts longer) 40 ttJg6 m ate. I cannot remember h ow deeply I tried to analyse over the board, but cer­ tainly I saw enough to appreciate that quiet play does not work, and so I broke up the position with 31 gS ! ? At last my knight was back in play, or at the very least I was able to exch ange my passive knight for her more active knight with 32 hxg6 ttJxg6 33 ttJxg6 'it'xg6, then a further exchange with 34 .l:!.xe7 .l:!.xe7. ...

I belatedly appreciated that Black was in serious trouble. Quiet play, I de­ cided, was losing . If, for example, 3 1 ...ttJf7 32 �h 3 'it'g 5 (or 32 .. J::t x e1+ 3 3 l:txe1 l:txe1+ 34 'it'xe1 'it'e4 3 5 'it'xe4 dXe4 36 1i.e6+ 'it>f8 37 1i.xC4 ttJd6, and Black is a clear pawn down) 3 3 1i.d7 l:txe1+ 34 l:txe1 l:te4 35 'iYd1 f5 36 1i.e6 i.c6, when Black already has problems on the light squares, and White can speed up his simplifying attack with 3 7 i.xd5 1i.xd5 3 8 ttJxd5 .l:!.xe1+ 3 9 'it'xe1 "Yi'xh 5 40 'it'e8+ 'it>h 7 41 'it>g 2 ! , with a nasty little zugzwang (41...'it'g 6 42 ttJe7). Or 31 ... ..tc6 3 2 �h 3 'it'g 5 33 .l:i.xe7 l:txe7 34 .l:i.e1, which again leaves White clearly on top. The knight in the corner has still not developed. Likewise, 31 ... 'ilVd7 32 .l:!.xe 7 ! ( 3 2 ..th 3

My pawn structure was even worse than before, but at least my pieces were reasonably active, and I was hop-

121

Why we L o s e a t C h e s s i n g th at I could squeeze a draw. Thi s indeed happened after 3 5 l:te1? l:txe1+ 36 "iix e1 Wf7 (now Bl ack can hold) 37 iLf3 "iif5 38 "iie 2 'i!Ve6 39 "iid 2 "iif 5 40 iVe2 iLc6 (maybe 40 ... iVe6 and offer a draw?) 41 iLg4 iVb1+ 42 Wg2 iVe4+ 43 'i!Vxe4 dxe4 44 iLe2 iLd5 45 Wh3 f5 46 Wh4 wg6 47 iLh5+ Wf6 48 iLe8 We7 49 iLg6 'it'f6 50 iLe8 We7 51 iLg6 Wf6 Yz-Yz. White h ad simplified a little too far. In stead, 3 5 "iif4! is winning for White, as my opponent noted after the game.

41 iLd1 l:rd2 42 �e1 l:td3 43 iLe2. Lau­ terbach should certainly h ave won this g ame. Chess i s an absorbing game, and while I was feeling niggling pain after falling down a long set of stairs at the hotel, I was still able to pl ay a reason­ able game of chess, not h aving, for ex­ ample, a losing position after the first dozen moves. White, h owever, played better than me.

Test 10.4 C.Crouch-P.Gait

H i l l i ngdon League

The queen threaten s to invade with "iid 6, chewing up the weak b6-pawn. If then 3S .. .fS, the timing is right for 3 6 l:te1 l:txe1+ 3 7 �xe1 'i!Ve6+ 3 8 �d2, and White h as a classic good bishop versus bad bishop win . 3 S ... "iic 2 would b e a n attempt to counterattack, but White can, of course, grab the pawn with 36 "iixh 6. Bl ack runs out of play after 3 6 ... �f7 ( 3 6 ... "iix b2 3 7 "iig 6+) 3 7 'i!Vc1 l:te2 3 8 "iix c2 l:txc2 39 l:tb1, and Black's rook gets trapped after 39 ... iLc6 40 iLf3 �e6

122

2006

h ave mercifully m ade very few one-move blunders, where I m ake a move with a piece and h e just takes it, or I miss th at he was attacking my piece. Perh aps one of my more memo­ rable, or forgettable of my g ames, was in a quickplay at Richmond a few years back, when pl aying again st Rasa Noe­ inkeviciute. I moved my king to an ad­ joining file, and immediately she stopped the clocks, and she went up to the arbiter. I simply could not begin to understand what was going on. It was a few minutes l ater, when the arbiter h ad arrived, th at I saw wh at the prob­ lem was. I h ad moved my king into check, along a file, and I could not see the g ap between her rook and my king. Thi s was a bizarre finish. It was not just a question of my very poor eyesight at the time, but rather

Te s t Te n th at my brain was clearly not in full focus on chess. Young children m ay occasion ally try to move a bishop or queen on a diagonal , and m ay try to switch from a light square to a dark square, or vice versa. I was making a similar mistake with the rook, seeing her rook moving from the c-file to the b-file. The end result was a draw, after a time penalty for my illegal move. I was amazed th at I h ad got away with it, thinking th at if I m ake a mistake, she should be allowed to take my piece in a quickpl ay finish. Apparently though the arbiter was correct. Thi s was a quickpl ay finish, not an actual blitz game. Back to the G ait game. Not a one­ move blunder, but a two-move blun ­ der.

I tried 27 'iVg4? (A), seeing my oppo­ nent's move, but missing that after the exch ange h e has a win th e following move, with 27 'iVxg4 28 hxg4 .te2, trapping the rook. White resigned (O-i). White could h ave fought longer with 2 7 flig3 (B) 2 7 ... .th S 28 .l::.b l .th 6, but one can feel certain th at Black would eventually win . Bl ack's bi shop­ pair is extremely strong . ...

123

Test E l eve n 11. 1 White to play

113 Black to play

A) 2 7 ..txh 6 B) 2 7 .l:f.xa8 C ) Something el se?

A) 2 7 ... e5 B) 2 7 . ..1:1g 5 C) Something el se?

11.2 White to play

11.4 Black to play

A) 2 7 fxg 6 B) 2 7 lIxh 6 C ) Something else?

A) 2 7 .. Yiilc 2 B) 2 7 ... h 5 C) Something el se?

124

.

Te s t E l e v e n

Test 11.1

C.Crouch-J.McKenna

London Open

2006

It's difficult to get much worse than this, for any player of reason able strength . I do not pretend th at I was pl aying particul arly well in the early part of the g am e, but then he made mi stakes too. Out of kindness to both pl ayers, we do not cover the first 26 moves. My position looks extrem ely promising just at the mom ent, but h e could h ave equalized comfortably i n the opening. B y now, I was clearly in a winning position, with a powerful rook on the seventh and a protected passed pawn just behind. I could even take his own passed pawn without too much difficulty. In this exercise, I deliberately put in an obvious blunder to set up a false trail for the reader, with 2 7 iah6? (A) 27 ... J::t e 1+ 28 'iVfl J::t xf1+ 29 'it>xfl, but even this is winning for White; his passed pawn dominates.

There are several other ways of pl aying the position for a comfortable win for White, but not the casual 2 7 ClJxc6? ( C ) 2 7 ... .i.xc6 28 .i.xc6 l::t x e3 2 9 fxe3 �g 3, for example, allowing Black too much counterplay. After 30 J::t e 7, there is a perpetual with 30 ... �e1+ 31 'it>h 2 ttJg4+ 3 2 hxg4 'iYh4+ 33 'it>g l 'iVe1+ 34 'it>h 2. The other capture on c6 i s al so un ­ clear after 2 7 .i.xc6 (C) 2 7 ... l::t x e3 2 8 fxe3 .i.xc6 29 ttJxc6 �g 3 . It i s possibl e th at one or two lines for White might still keep some sort of edge after deli­ cate m anoeuvring, but White should be aiming for something simpler and clearer. All thi s suggests 27 'f1C7! (C), blocking Bl ack's diagonal to the kingside.

Bl ack's rook and knight h ave some activity, but Black's queen and bishop now do nothing. With care, White should win comfortably. The next question is why I blun­ dered with 27 l::tx a8?? (B). Part of thi s was that I would h ave been frustrated by not finding anything to set up a win

125

Why we L o s e a t C h e s s with an immediate capture o n c6, and decided th at it was time to eliminate the bishop on a8. Partly too I would have lost concentration . I h ave a win ­ ning position, my opponent is clearly on the lower h alf of the board, so why should I use up energy in such an easy position ? The trouble i s that an easily winning position plus lack of concen­ tration adds up to the loss of grading points. I pl ayed my sacrifice, and then after 27 .. J1Vxa8 I suddenly noticed that not only h ave I given up the exch ange, but I h ave allowed Black to control the a-file.

All thi s broke my equanimity. I set up a couple of passed pawn s with 28 ii.xc6 .l:txe3 29 'iWxe3 ii.xe3 30 ii.xa8 i.. x d4, but Black too h ad strong passed pawn s, and after another bad move, which we'll see in Test 12.4, I was los­ ing. In stead, if I could h ave found a way of regaining my composure, I would still h ave h ad chances of keeping a slight edge with 28 'iWxC 3 ! . I h ad still clearly not recovered from

126

my stroke, and was still m aking many serious blunders, m ainly through poor vision, both metaphorically and liter­ ally. Here I forgot th at Black could at­ tack on the a-file once White's rook was no longer on a7.

Test 11.2

C.Crouch-D.Okike

Thames Va l l ey League 2007

White to play. It is not difficult to work out that this was a Nimzo-Indian, with the well-known doubled c-pawns. Those pawn s can at times end up as weak, but h ere it i s not so clear that Bl ack can take any advantage of the slight positional defect. On the con­ trary, White is better, with more space for his pieces, and has excellent ch ances of attack on the kin g side, Bl ack's pieces being con strained on the l ast two ranks. The next few moves run smoothly, beginning 20 ii.gs f6 21 ii.e3, forcing a pawn to advance in front of the king,

Te s t E l e v e n setting up a new weakness. The prob­ lem for Black is not so much the weak­ ness of g6 itself, but more the threat of a sacrifice on h6, forcing Black's king into extremely open play. He tried, m aybe with great anxiety, 21 ... g5, covering the g 6-square, but also an open invitation for White to sacrifice, or a dangerous pawn ex­ change with h4. The computer explores the idea of defending with 21 ... .l:tc8, but after 22 .l:tf3 .l:tfd8 2 3 l:tg 3 'iif7 24 'iih 4 �f8, there i s a sacrifice o n a different square with 25 l:txg 7+ �xg 7 26 ii.xh 6. White ties thing s up after, for example, 26 .. :iVg 6 27 ii.e3+ 'it>g8 28 l:th 8+ ct;f7 29 l:th 6 �g 8 30 l::rh 7+ ct;f8 3 1 l::rh 8 . For the next few moves, White qui­ etly undermines Black's king side with 22 h4 l:tg8 23 l:tf3 l:tg6 24 l:th3 ct;g7.

Now I played 2 5 ii.e2. This was some­ thing of a waiting move, aiming for a better diagonal for the bishop when the storm arrives. There was a possibility of combinative play with 2 5 hxg 5 hxg 5 26 l:txh6 l:txh 6 27 �xg 5 l:tg6, but I could not find any further breakthrough. With

rooks o n the board ( 2 8 i1.. x e7 l:txg4 2 9 ii.xd6 l:tg6), o r with the rooks off (28 l::rh 7+ 'it>g 8 29 l:txe7 .l:txe7 30 i1.. x e7 .l:txg4 31 .1Lxd6 l:tg6 32 .1LC7 .l:ta6, and Black is better), Black can at least hold. White's bishop-pair does not cooperate. This was a disappointment for me, as I had pressed my attack as far as I could. I pl ayed the bishop move, aiming for a later .1Lg4, but now my e-pawn was slightly weakened. Maybe instead 25 l:th 1 ! ?, as suggested by the com ­ puter. The computer then suggests holding moves, such as 'it>h 2 or l::r h 2 , waitin g for t h e opponent to create a weakness. A more constructive pl an would perh aps be hxg 5 followed by � 3 , when at least we can see what White i s aiming for. In the g ame, Black should h ave tried 25 ... l:th 8 ! , avoiding any unnecessary concession s. White can try for an edge with 2 6 �e6 �xe6 2 7 dxe6, but there is no great chance of certainty for an edge in over-the-board play. A steadier line might be 26 l:th 1 �d8 27 ii.d3, still perh aps keeping an edge, but White is circling around with his pieces. Over the l ast few moves, White has never quite been able to m ake his breakthrough . Wh at is surprising is th at it is Black who open s up the posi­ tion, with 2 5 ...f5 26 exf5 tLlf6. There are tactics for both sides, with White's queen and rook being threat­ ened, and Black's rook and two of the pawn s in front of his king are about to be stripped away. It is now time for

127

Why we L o s e a t C h e s s White t o calculate with clarity. Unfor­ tunately I did not do this well.

27 �h6! (B) i s straightforward, m aking sure th at White can get at the black king without any pawn obstruc­ tions. 27 ... liJxg4 2 8 l:ixg 6+ �f7 29 .ltxg4 drops White's queen, but h e has a rook, bishop and two pawns in return, more than sufficient compen sation, espe­ cially as White keeps a serious attack against the king . Indeed, 29 .. .1:[g 8 30 hxg 5 �xg 6 31 �h 7+ l:ig 7 3 2 �xg 7+ �xg 7 33 f6+ "i!Vxf6 34 gxf6+ 'It>xf6 soon win s for White, with his extra piece. In stead I m ade a bad stumble, with 27 fxg6? (A) 27 ... liJxg4 28 �xg4.

128

Time trouble was approaching, this being an evening m atch, and my op­ ponent tried the n atural 28 ... �xg6?, but I was soon able to gain a winning advantage with 29 hxg5 liJf7 30 gxh6 'It>h7 31 .ltf5+ 'It>h8 32 l:ig3 l:ig8 33 l:ixg8+ 'It>xg8 34 .ltf2 e4 35 .lte6 'It>h7 36 .lth4 'ilC7 3 7 .ltf5+ �g8 38 h7+ �g7 39 h8'iV+ liJxh8 40 l:ih7+ 1-0. Instead, 28 ... liJxC4! keeps play com­ plicated. Maybe White should eventu­ ally end up winning, but it would be more due to luck than judgement. White h ad a much clearer option .

My notes at the time suggested that I was probably thinking of 2 9 .ltxg 5 ? ! , but then after 2 9 . . .h xg 5 30 l:ih 7+ �xg 6 3 1 l:ixe7 l:ixe7 Black h as recovered much of his ground. In stead, 29 l:ixh 6 'YWd8 ! ? 30 l:ih 7+ 'It>xg 6 31 .ltxg 5 "i!Vxd5 32 .l:!.d7 'iig 8 ( 3 2 ... 'iie 4? ! 3 3 .l:!.f3 "iVxg4 34 l:if6+ leads to an attractive checkmate, with 34 ... 'It>xh 5 3 5 l:ih 7 m ate) 3 3 .l:Ig 3 iVh 8 ! 34 i. c l ! "iVxh4 3 5 i.h 5+ �f5 3 6 l:if3+ �e6 3 7 l:ih 7 'YWel+ 3 8 l:ifl "i!Vxfl+ 39 'It>xfl eventually leads to a win for

Tes t E l e v e n White, but who can doubt th at White, and indeed Bl ack, could h ave pl ayed it so accurately? This was one of m any club games I played against David Okike over a few years, playing again st Kin g s Head and Hayes. I won most of them, but h e h as scored a win and a draw against me. The strangest encounter was a few month s after thi s one. I walked through Oxford Circus be­ fore the start of a London League m atch, then realized th at I was starting slightly l ate, so I got up the stairs at the venue, quickly m ade my way through the dimly lit room, m ade my first move as Black, ... c7-c5 on the dark squares, pressed the clock, sat down, then sud­ denly noticed th at I h ad played closer to th e edge, light square to light square, 1 e2-e4 b7-b 5 ? ?, in stead of the intended 1 ... c 5 . M y opponent briefly wondered what I had planned with my unexpected opening move, when it suddenly be­ came clear from my shock that I h ad blundered on move 1, a mixture of poor eyesight (I am only partially sighted) and being distracted by having to rush to the venue. I did not ask for the game to be restarted at the beginning, and I suspect that it would have been im­ proper for me to do so. My opponent sportingly suggested that we could start the game at the beginning anyway, be­ ing aware that I was barely able to see, and that he would rather have an inter­ esting game rather than a hollow win.

Much l ater, when looking though my games for thi s book, I m ade the second big visual blunder in thi s g ame, pressing the wrong button on my computer (no, it wasn't a '(' button too long to expl ain, but I'm sure th at the memoirs of Grandmaster James Plaskett will give the full story). Any­ way, what h appened is th at I m anaged to delete my game. Almost certainly it would be possible to retrieve it, but I h ave not got round to this. I was able to remind myself that I m anaged to get a good position in the 1 ... C5 opening, or a Sicilian, then mispl ayed the queenside, then he mispl ayed it, and then finally a win for me. A long shaggy dog story. There are, of course, m any more an ­ ecdotes from events for blind and par­ tially sighted players. Alas, such things are inevitable if half the players cannot see, and the other h alf can barely see, and can easily bump into obstructions. The cl assic question is wh at h appen s when a blind player plays a deaf pl ayer. My own nightmare is that my brain dam age not only makes me only par­ tially sighted, but al so th at I suffer from a degree of aphasia. I can usually understand wh at is being said, but it can be very difficult to bring my words together, and I am terrified of saying the wrong move, and making a big mistake. To make a move over the board, .l::!.fe1, is easy enough, and it is easy enough also to write it on the scoresheet, but then to have to say

129

Why we L o s e a t C h e s s ' Rook H ector Eva One' takes more time, and the intern ation al Germ an Braille version, 'Turm Hector Eva Eins', takes several seconds to work out. By the end of all this, I cannot remember what I have written down on my scoresheet. Fortun ately, Braille pl ayers know their own allowances. After all, every player has his or her own disability.

Test 11.3

D.Buckley-C.Crouch

B ritis h league (4NCL)

2006

pawn push weakens his overall king­ side pawn structure. If he was worried about the back row, a much simpler option was 2 3 f4! ? �C5+ 24 'iyxc5, and then after either recapture, White would try 25 �d7. White has a clear edge in the double rook and pawn endgame, Black's pawn structure being weakened. If White was worried about counterattacks, then surely forcing an advantageous endgame was the safest way of not losing the game in a complicated mid­ dlegame. 2 3 �xa7 ! ? can still be tried. Then 23 ... �cd8 24 �f1 f4 2 5 �xb6 �b8 26 �c6 �xb2 27 �d3 will give no m ating attack for Black, just a little pressure, whereas White still h as two threaten­ ing pawn s. After thi s reprieve, I threatened to set up an attack with 23 f4, also keep­ ing the queens on the board. Maybe, after 23 ... a5, I would h ave been worried about 24 �d5, but after 24 ... �g 7 2 5 ttd7 f4, Bl ack h a s reasonable chances of h olding the balance. ...

Continuing from Test 8.1, Bl ack's king side attack is threatening, but far from decisive, and if White is able to keep an extra pawn, and perh aps an extra passed pawn l ater, then the onus for Black i s to keep the bal ance. Here White pl ayed slightly nerv­ ously with 23 h3?!, with no discernible intention other th an to create an es­ cape square for the king on the back row. Black i s not threatening anything on the end rank though, and White's

130

Te s t E l e v e n He now tried 24 ltd 7, doubling on the seventh . It i s possible th at h e h ad previously considered trying 24 ltdS 'iYg 7 25 ltg s 'iYh 6, but thi s seem s messy. If 2 6 ltbS 'iYg 6, White's best op­ tion is 27 ltg s 'ii'h 6, with repetition., as 27 'iYxa7 ? ltg 8 28 'iYb7 ltcd8 ! i s ex­ tremely dangerous, with threats of ... 'iYbl+ and ...ltdl, and m ate on h l . If 2 9 lt c l lt d S 30 g4 (otherwi se m ate o n g 2 ) 30 .. .fxg 3 31 cxdS gxf2+, soon checkm ating. 50 White definitely h as to pl ay carefully. The other option is an immediate capture with 24 'iYxa7, and thi s looks good, even though White h as lost time. 24 ... 'iYe2 2 5 ltcd3 'iYxb2 2 6 'iYe7 ! h as drawn Black slightly out of pl ay with his pieces. Then 26 ... 'iYg 7 27 ltd8 ltg 8 28 ltxg 8+ (not, of course, 2 8 ltxc8?? 'iYxg 2 m ate) 28 ... �xg 8 29 'iYb7 is decep­ tive. It might look as though the pawn on g2 is seriously weak, but as long as Bl ack can do nothing to add extra pressure (for example, 29 ... 'iYg 6 when 30 ltd7 e S 31 lte7 prevents ... e4), he can do nothing with his pieces, and White can gradually improve. After the move actually played, the computer suggests several option s for Black, including keeping the a-pawn safe with 24 ... a6 or 24 ... aS. The aggres­ sive counterplay, with 24 ltce8 2 5 'iYh4 'iVf5, looked t o me far more ap­ pealing, not least because my king was now safe. I felt that m aybe I was better, and that it was up to him to prove oth­ erwise. ...

H i s next move 26 b4?!, was unex­ pected, a quiet move on the queenside, before the king side storm s. 2 6 ltxa7 is the more direct way of setting up a queen side pawn advan­ tage. Then after 2 6 ... ltg 8 2 7 ltf3 'iYbl+ 28 �h 2 ltxg 2+ 29 �xg 2 ltg8+ 30 ltg 3 'iVe4+ 3 1 �g l ( 3 1 �fl 'iYb1+ al so leads to perpetual) 3 1...ltxg 3+ 32 fxg 3 'iVel+, with a draw by perpetual . Here the computer suggestion, 2 7 �h 2?, is deeply un convincing after 2 7 ... ltg S ! 28 ltd7 lth S 29 g4 fxg 3+ 30 'iVxg 3 ltg8. I carried on with the attack with 26 ltg8, when probably the only sen­ sible reply for White is 2 7 g4! fxg 3 28 �xg 3, clearly with at least a draw for Black after, for example, 28 ... ltxg 3+ 2 9 fxg 3 'iYbl+, but with no obvious win­ ning plan . In stead, 2 7 ltxa7 ? 'iYb1+ 28 �h 2 'iVfl 29 'iVf6+ ltg 7 30 g4 'iVxf2+ 3 1 �h l �e1+ 3 2 �g 2 �d2+ 3 3 �fl ltd8 i s not safe. White's king i s under pres­ sure, and Black h as no need to repeat. White missed a tactic h ere with the move 27 �h2?, and then I missed a good reply. ...

13 1

Why we L o s e a t C h e s s ing the tactic, White kept an edge with 2S lIcd 3 lIe6 29 lIdS lIeg6 30 lIxgS+ 'it>xgS 31 'iVdS+ 'it>g7, and should h ave carried on with 3 2 'iVd5 ! in stead of 32 lId6?!, equal, but l ater a loss. Yes, it has to be admitted that the play in this g ame has been for the most part ex­ tremely unexciting, with too many mistakes on both sides. Unfortunately it doesn 't get better, as we'll see in Test 15.2. 27 ...e S ? (A) is solid enough, but as Keith Arkell pointed out, 27..JJ.g5! (B) i s very strong, with a winning queen sac­ rifice after 28 1:txf7 J::t e g 8 ! ! 29 J::t xf5 lIxg 2+ 30 �h 1 lIg1+ 3 1 �h 2 J::t 8 g2 mate. Sometimes the spectator has the best view. And I think th at neither player was seeing the g ame well any­ way. White can, of course, avoid the im­ mediate checkmate, playing defen­ sively with 2 8 f3 lIeg 8 29 'iYf2 lIxg 2+ 30 'iVxg 2 lIxg 2+ 3 1 �xg 2. Then 3 1 ...'ilVb 1 ! seem s the best way t o prevent White's rooks from coordinating properly, and White still has defen sive problems. One would not expect White to be able to h old the draw after 28 g4 fxg 3+ 29 fxg 3 'iVf2+ 30 �h 1 1:teg 8, but h e can struggle on in a rook and pawn ending, with 3 1 'iVd4+ 'iVxd4 3 2 J::t x d4 1:txg 3 33 J::t x g 3 J::t x g 3 34 J::t d 7 J::t xh 3+ 35 �g2 J::t x a3 3 6 1:txf7. White i s going to be a pawn short of safety, though, once the queen side pawns are eventually ex­ ch anged. After my less accurate move, miss-

13 2

Test 11.4

N.Pert-C.Crouch

B u ry St Ed m u nd s

2006

From the end of Test 8.2, White h as a 5-4 pawn advantage on the kin g side, while Black h as an i solated pawn on the queen side. The danger is, of course, th at Black is going to lose the a-pawn . N aturally Black will not be able to win, but h e should h ave good ch ances of h olding the g ame. White's extra pawn on the king side is a doubled pawn, and sometimes the pawn s can get in the

Te s t E l e v e n way of each other. Visualize a position where White takes the a-pawn with his rook, Black pl ays ... 'iVdl+, then �h 2 � s+ and a perpetual , and the point i s easily m ade. I played very poorly over the next few moves, quite simply losing concen­ tration . 27 ... h S ? (A) was bad. I noted at the time th at " I see from my scoresheet that I pl ayed thi s more or less immedi­ ately, with 29 minutes to go until the time control on move 36. No time shortage therefore, but I was con scious th at there was going to be a quickpl ay finish coming up. I should h ave thought much, much h arder, and m ade White work h ard to try to win the pawn ." U sually I am reason ably good at time pressure, m aking it a rule for my­ self that I try to keep at least five min­ utes for the last few m oves, to en sure that I h ave some time, if necessary, to think about the last couple of moves. Sometimes, as here, I speed up too early before the time control . There is no general rule to deal with thi s type of question, as different people genuinely thin k differently, and there are many players who can think superbly with the fl ag about to fall, while others pre­ fer to take their time, and can get flus­ tered by h aving to move in stantly. On thi s move, I got 'pre-flustered', just m aking a move, hoping to set up some sort attack against White's king­ side, and vaguely thinking th at at least I am opening up a flight square for my

king . I t loses a tempo, and m y oppo­ nent m akes much better use of it. I should h ave thought and planned at thi s moment, deciding h ow to h an­ dle the rest of the g ame. My pawn moves are almost irrelevant. I cannot improve anything with my pawns, so I need to concentrate on my pieces. Clearly my king and rook are stuck to defence, so that leaves moving my queen and knight into better pl ay, and preferably m aking them work together rather than separately. Thi s immediately suggests 27 'iVC2! (A), •••

so that if 28 ttxas??, Bl ack win s im­ mediately with 28 ... 4:Je4 ! . The queen on c2 is superbly placed, putting pressure on f2, and the second rank as a whole, and reminding White that he needs to cover his back rank. Once Black has seen thi s, all th at White can h ope for is to keep equality. In stead, 28 4:Jd7 ? ! 4:Jxd7 29 'iVxd7 'iVC3 keeps Bl ack's a-pawn safe, and it would be up to White to hold the posi­ tion.

13 3

Why we L o s e a t Ch e s s The correct 2 8 'iVC5 i s safe, and then 28 ... 'ivxc 5 29 dXc5 tra8 30 tiJC4 Wf8 3 1 Itxa5 (m aybe 3 1 tiJxa5 ! ?, keeping the rooks, but Black should still hold) 3 1...trxa5 3 2 tiJxa5 �e8 3 3 f3 �d7 34 �f2 �C7 35 tiJC4 tiJd7 3 6 tiJd6 f5 37 tiJe8+ c,t>d8 3 8 tiJxg 7 tiJxC5.

Black i s a pawn down, but it i s diffi­ cult to envisage any realistic winning chances for White, given th at Black's pieces and pawn s are well coordinated, and th at White's doubled pawns give littl e chance of setting up a passed pawn . Almost certainly I would h ave been suspicious of allowing the queen ex­ ch ange on c5, but small-scale m anoeu­ vring with the king and rook on the same part of the board m akes it diffi­ cult for White to create any significant advantage, even with the extra pawn . The simplest plan, 2 7 tiJg4?! (C) 2 8 tiJxg4 'iVxg4, is a n altern ative possibil­ ity, and then 2 9 'iVa7 Itc8 30 iYxa5 h 6 . Bl ack would h ave good chances o f h old­ ing the position if the queens are off the board, but if White were to keep

the queen in active play, the chances are that he would be able to find a win. 2 7 tiJds (C) •••

is another try, trying to keep the queen and knight working together, but 28 'ih> 7 ! keeps White's pieces the more active. With my own misguided move, I was thinking in term s of exch anging pawn s with ... h4, with a 4-3 defence. However, exchanging a good pawn of my own, for a doubled pawn, is not a good option, especi ally as I was losing time.

•••

13 4

Play continued 28 'iVC7! (stopping ...'iVc2) 28 ... h4 29 gxh4 'ii'x h4 30 tiJf3 'iVe4

Te s t E l e v e n 31 �xa 5 lbd5 32 �a7. My play continued to fall apart after 32 ... g6?! (weakening my pawn structure ) 33 �d7 'it>g7?! (in­ credibly, I was thinking of opening up the h-file for queen and rook ) 34 �d6 lbf6 35 �e5, and when I started to recognize that the queen exchange would lose, if not necessarily quickly, I retreated with the queen, without much hope, tryi ng

3 5 ...�C2?!. Then after 3 6 lbg5 �f5, the last move before the time control, I de­ cided that I had to undo the retreat, as otherwise the pin on the knight on f6 would lose quickly. A disastrous finish before the time control, and really I would expect to lose this quickly. There is, h owever, a sequel, as we will see in Test 1 5 .4.

13 5

Test Twe l ve 12. 1 White to play

12.3 Black to play

A) 28 'iWf4 8) 2 8 'iWc2 C) Something el se?

A) 30 .l::!. a c8 8) 30 .. .'!i'lc6 C) Something el se?

12.2 White to play

12.4 White to play

A) 28 fS 8) 28 'iWxC7 C) Something else?

A) 3 1 b7 8) 3 1 ct'lb3 C) Something el se?

13 6

...

Te s t Twe l ve

Test 12.1

C.Crouch-J.Cox

London League

2006

Continuing from Test 9.1, I played 28 'iYf4? (A), pointl ess, as there will be no checkmate on 0. Indeed, Bl ack was able to obtain an edge. Somehow I need to create some sort of edge on the queenside, and with files rather than diagon als. A few month s after thi s g ame, I gave a doubl e exclamation m ark for 28 'iYc2! (B), mainly because of the paradoxical ef­ fect of playing '1Wc2-f2 and then unex­ pectedly 'iWf2-C2. Maybe one exclama­ tion m ark. The critical line would seem to be 28 .. .'�jxg 3 29 hxg 3 4Jf6 30 a4! (White needs to attack quickly) 30 ... .l:!.e 3 31 as 4Jxds ! ? ( 3 1 ... 4Je4 32 axb6 .l::i. x g 3+ 3 3 �h 2 axb6 34 "iYb2 favours White; Black cannot find a way to shift the queen from a6 to the h-file) 32 l:txfs ! ( 3 2 'iWg 2 ? ! allows Bl ack t o wriggle out with advantage after 32 ... 4J0 33 axb6 axb6

34 .l:!.b3 'iWxC4 3 5 .l:!.xb6 'iWd3 ; note Black's central control in thi s end posi­ tion) 32 ... gxf6 33 'iWxfs+ 'iitb 8 34 'iYxds, and White holds the draw. There i s the obvious practical objection th at no player could be able to calculate this over the board. Maybe so, but it is use­ ful for players to sh arpen their posi­ tional in stincts in complicated posi­ tion s. Who knows, if you can find the best move, the rest might follow. I was momentarily impressed with 28 'iWb2?! (C), but 28 ... 'iWa4! allows the queen to squeeze out, or rather to ex­ ch ange. After 29 "iYb3 (otherwi se ... 'iWxc4) 29 ... 'iWxb3 30 .l::i. xb3 (30 axb 3 ? 4JX(3) 30 ... 4Jd2 the knight forks two rooks, and al so an important pawn ( 3 1 l:t a 1 4Jxc4! ?). So White does not want to allow Black to invade the a4-square. In the game, after 28 'iWf4 ...

1 was critical in my notes about 28 4Jxg3: "Not a bad move, but there is a better one", but this seem s harsh. If your flag is hanging, you want to play good moves quickly, and not waste time searching for exotic geometry. That said, ...

...

13 7

Why we L o s e a t C h e s s 28 ... ttJes ! 29 l:I.e1 ttJxg 3 30 l:I.xes ttJe2+! (seen after the game!) 3 1 l:txe2 l:I.xe2 3 2 'iYd6 'iYxa2 is good for Black, i f you have time to think about it.

It would take a lot of time for a pl ayer to convince him self that Bl ack could demon strate th at he eventually can win after, for example, 33 'ilVc6+ 'it>d8 34 'ilVd6+ 'it>e8 3S 'ilVc6+ 'it>f8, and meanwhile the tim e is ticking. Re­ member too th at this is five moves down a sh arp line. In practical term s, there was no reason for Cox to want to touch thi s line. In the g ame, after Cox exch anged, I recaptured with 29 'ilVxg3, covering the squares next to the king. 29 hxg 3 'ilVxa2 30 l:I.d1 l:1.e2 31 'ilVf1 ttJes would in stead put White under pressure. Bl ack collected his second pawn, with 29 'ilVxa2, with every chance of winning a third pawn . I decided th at my only chance was to play actively, with 30 'ilVd6. I felt th at my queen was m oving through the trap door, but what else could I h ave done? The trap was closed with 30 ttJe5. ...

...

13 8

I played 3 1 l:I.bb1!. Thi s i s a critical defen sive plan for White, win, draw or lose. It was also difficult to find over the board, although it looks obvious afterwards. Shifting the rook to a1 is the only chance. The desperate sacrifice with 3 1 ktfb1 'ilVxC4 2 2 l:1.xb6 axb6 is not worth thinking about. After either re­ capture on b6, Bl ack sets up a swarm of winning checks with ... 'ilVg4+. Then 31 'ilvxc4 was natural and good. In stead, 31 ... ttJf3+? 32 'it>h 1 l:I.e2 33 'ilVc6+ 'it>d8 34 'ilVa8+ 'it>d7 (34 ... 'it>e7 ?? 3S d6+ allows White a win) 3S 'iYc6+ is only a draw. 3 2 l:1.a1! is again the only chance. Of course, Black could give a perpetual immediately, and, of course, he could find a possible improvement to such a sequence, but for White, a degree of discomfort i s preferable to an immedi­ ate loss. There is a simple trap th at Black could fall for, with 3 2 ... 'it>b7 ?? 3 3 l:I.xa7+ 'it>xa7 34 'ilVC7+ 'it> a 6 3 S kta1+ 'it>b s 36 l:I.b1+ �a4 37 'ilVa7+, but Cox would h ave seen that without any dif­ ficulty. •..

Tes t Twe l v e In stead, h e started with checks, with 32 ... �g4+ 33 �hl �e4+ 34 �gl, and quietly improved his pieces with 34 ... �b7. Thi s time, White's sacrifice gives nothing, after 35 .l::[ x a7+ �xa7 36 �C7+ �a6 37 .l::[ a l+ �b 5, and the king escapes to c4. I fought on with 35 I:tfel, and the casual 3 5 ... ct:Jf3+? 36 �f2 even gives White an edge after 36 Wf2 �C2+ 3 7 �xf3 �xc3+ 3 8 �f2 "iVd4+ 39 �f1. So 3 5 ...�g4+ 3 6 Whl "iVf3+ 3 7 Wgl, and to my con siderable surprise, h e offered a draw.

Here 37 ... �xc3 would have been a win for Black, after the adjournment, to be played at some later date. Few play­ ers enjoy a second session in evening games, with time being precious, al­ though there are others who like a quickplay finish even less, with every­ thing decided in a time scramble close to 10pm after a long working day. For myself, I felt I had to avoid a quickplay finish, which was my right in the league, not least through respect to my team colleagues. I did not want to lose for the

team in what could have been a critical match, as a result of tiredness and diz­ ziness. My opponent clearly did not want to have a second session, and I apologized before the start that I could not play a quickplay finish. At the end of this particular session, he decided he wanted to force a draw on the evening. Had he sealed a move, I felt sure that I would have gone home and analysed the position, and decided that it would not h ave been worth playing on, and would have resigned. An expl anation then for an unusual result: an unusual situation . I do not know wh at i s h appening in other coun­ tries in evening league matches. In England there i s often con siderable debate about quickplay versus slow­ play. The late Lev Polugaevsky noted, well before computers were good at chess analysis, th at adjournments in interna­ tional events would at some stage be­ come obsolete, because players with a good database or engine would h ave an unfair advantage compared to the player with less computer help. It is now generally agreed that games should be finished in one session, for morning or afternoon events. In evening games, computer analy­ sis often cuts down second sessions, because one of the players would see th at it is not worth playing one, either because it is a clear draw, or because one of the players is losing . The few adjournments th at do carry on,

13 9

Why we L o s e a t C h e s s though, are often of great interest, and enjoyable for the pl ayers.

Test 12.2 C Crouch P Roberson .

-

.

B ritish League (4NCL)

2007

From a Modern Benoni. White has clearly been able to push his central pawn s safely, and can create a passed pawn, but his position i s not quite as overwhelming as it looks. He still h as to think; h e cannot pl ay fully autom ati ­ cally. After some thought, I decided to ex­ change the pawn s on d6, keeping an isolated but advanced passed pawn on dS, helped by rook and bishop. I played 24 tiJe4. 24 e6 'fie7 i s the obvious altern a­ tive, and who could resist a well­ protected passed pawn on the sixth ? It i s unclear, though, how White can m ake progress in such a blocked posi­ tion . After, for example, 25 'fid2 tiJg 7 26 i.. a 2 tiJfs 2 7 i..f2 b S 28 �c1 c4 29 b4

140

l:tac8 30 axb s axb s 3 1 i..b 1 tiJg 7 3 2 i.. e 3 �fe8 3 3 i.. c 2 l:ta8, i t is still pleas­ ant for White, but what next? Black too has a protected passed pawn . In the game, Black defended on f6, with 24 i.. g 7. 24 ... i.. xf4 2 5 i.. xf6 looks much too dangerous for his king, so he must try to hold his position . ...

I now continued my planned pawn exch ange, 25 exd6. It is again possible, but slightly illogical, to try 25 e6. Black can sn atch a pawn with 2 S ... 'fixa4, now th at White's knight h as moved away. 26 tiJC3 'iYb4 27 tiJa2 'fia4 is a drawing possibility, although Black could try for more with 2 7 ... 'fias . In stead, 2 5 as ! is seriously t o be con sidered, and might well be best, but leads to complication s and counter­ complication s. I would h ave con sidered it, but would probably h ave decided that the move I played was simpler, and gives White a clear edge. I would h ave been h appy with my move avoid­ ing any possible sh arp play. After 25 as fxe s 26 fxe s bS 27 exd6 cxd4 28 tiJxcS 'fibs 29 tiJxb7 'fixb7 30

Te s t Twe l v e dxC7 �xc7 White has a comfortable extra passed pawn . It i s possible that I would h ave been concerned about whether I would h ave a genuine ad­ vantage after the more natural 27 e6 �c8 2 8 �a2 c4, and decided in retro­ spect that m aybe 29 iLe7 l:tfS 30 ttJxd6 ttJxd6 3 1 �xd6 ttJxdS did not give White much of an edge. In the end, I went for the simpler option . This was by no m ean s bad, but was still sub-optim al, and it turn s out th at I soon m ade a mi stake even in the 'simple' line. Play continued with 2 5 ...ttJxd6 26 ttJxd6 �xd6 2 7 �e7. Here Black could try 2 7 ... �xf4 28 .l:i.e4 �fS, but White is doing well after 2 9 d6+ 'iith 8 30 .l:i.e3 'ii'f4 3 1 dXc7 'ii'x C4 3 2 .l:i.d7 �g 8 3 3 'ii' d 6. However, 29 g4? ! would be more speculative, and after 29 ... �xf3 30 .l:i.d3 �xd3 31 �xd3 l:tf7 32 �d6 ttJxdS Black is better after, for example, 3 3 �C4 iLf8 34 'ii' e 6 .l:i.d8 3 5 �xf6 ? .l:i.d6. Bl ack in stead pl ayed 27 ... .l:i.ad8.

Now l over-egged my pl ay with 28 f5? (A), hoping to g ain a tempo by

bringing the bishop quickly into play with iLg 3 . Instead, 2 8 WiXC7 (B) 28 . . .�xC7 29 d6+ would h ave been simpler and more natural. Then 29 ...Wif7 30 �xf7+ (30 ttJes ? fxes 3 1 �xf7+ 'iitxf7 3 2 �xd8 l:txd8 33 fxes is over-elaborate) 30 ... l:txf7 3 1 ttJes ! . Black cannot take the knight, as the pawn is pinned, 31 .. .fxes ? 3 2 �xd8. 5 0 after 3 1 . . .l:tff8 3 2 ttJC4 bS 3 3 axbs axbs 3 4 ttJas White has finally ob­ tained that safe positional edge that he h as been trying to achieve for a long time. The next stage of development is to bring the bishop to f2, and then aim for the queening square on d8, via b6. In stead, I lost a tempo after 28 .. :iVxe7 29 .l:i.xe7 (29 d6+? Wif7 30 �xf7+ .l:i.xf7 3 1 dXc7 .l:i.xd1 3 2 .l:i.xd1 .l:i.xc7 leaves Bl ack a pawn up) 29 ... �xd 5 .

I no longer had time for .l:!.xC7. I n ­ deed, I was a pawn down, and h ad to readjust myself to staying a pawn down, but with good bishop activity, after 30 �e2 .l:i.fe8 31 .l:i.xe8+ .l:i.xe8 3 2 'iitf2 gxf5 3 3 � g 3 . I even later h ad

14 1

Why we L o s e a t C h e s s slightly the better of a draw in the endgame, but after a mistake by either side just before the time control, I missed a chance of winning. We re­ sume the position later in Test 14. 3 . Much earlier, m y superficial original intention had been 28 'iixd6 (C) 28 ... �xd6 29 �e7, but after 29 ... �f7 Black holds, and may be better. My tac­ tical vision was clearly not very good yet.

Test 12.3 M.Cutmore-C.Crouch

Kid l i ngton

2007

We TUsh through the early part of the game, and zoom to my blunder. As readers will h ave appreciated by now, my tourn ament at Kidlin gton was a treasure-house of my own mistakes. Two years after my stroke, I was con­ centrating on trying to regain my fit­ ness as quickly as possible, doing lots of walking, and playing lots of chess. I was playing confidently, but unfortun ately at that stage my brain was not working properly, and I could not quickly an a­ lyse the difference between good moves and mistakes. I g ave my oppo­ nents plenty of opportuniti es. The game was a Nimzo Indian, 1 d4 l2Jf6 2 C4 e6 3 l2JC3 �b4 4 f3 d 5 5 a3 �xc3+ 6 bxc3 0-0 7 e3 �e8 8 cxd 5 exd 5 9 �d3 c5 10 l2Je2 b6 11 0-0 �a6 12 l2Jg3 'iic 8 (all known so far) 13 �f5 !? 'iiC 7 14 �e1 l2Jc6 15 �b2 l2Ja 5 16 e4 l2JC4 17 �C1 g6 18 �h3 Jtc8 19 �g5 �xh 3 20

14 2

�xf6 'iif4 21 e5.

Play is level, in the sense th at it is unlikely that either player will h ave a clear advantage. The position is, how­ ever, dynamic. A slip by either side can end in a quick collapse. My biggest problem is that if White can get his queen to h6, then that i s the end. My hope for salvation i s that White h as several weaknesses in the centre. I retreated my bishop, 21 ... Jte6. The computer initially suggests that this was a mistake, and that 2 1 ... l2Je3 would h ave been good for Black, but retracts when it becomes clear that 22 l2Je2 i¥h 6 23 'iic 1 would h ave been good for White, as if 2 3 ... �xg 2 24 l2Jg 3, winning a piece. I feel sure th at I would h ave seen this. Pl ay continued with 2 2 �a2 �ec8 2 3 l2Je2 'iie 3+ 24 �h1. Both sides now h ave apparently set up some dangerous threats, but noth ­ ing i s going on at the moment. If either side breaks open the bal ance, it is the aspiring attacker who loses the back­ fire. E ssentially the only forcing m ove is

Te s t Twe l v e �Cl for White, with an exchange of queens, but thi s is a simplification rather than a m ating attack.

24 ... h5 was played, gIVing a flight square for the king, but al so perh aps weakening the pawn structure. There are several alternatives over the next few moves. Then came 2 5 h4 cxd4 26 cxd4. If Bl ack were to try for checkm ate after 26 ... �f2 ? 27 �Cl ct:Je3 (27 ... i.h 3 28 ct:Jc3 .txg 2+ 29 'it'h 2 win s a piece for White) 2 8 ct:JC3 �xa2 29 �xe 3 ! , then it is actually White who gives mate. So it was still time for quiet ma­ noeuvring with 28 ... ct:Ja 5 27 ct:Jg3 'iVh6 28 .tg5 �f8 29 �d 3 .i:!.c4 30 f4.

Here, short o f time, I tried to squeeze too much with 30 ... .i:!.ac8? (A), threatening a skewer with ... .l:i.C3. The more thematic move, 30 ... ct:Jc6! (B), gives Black a substanti al edge. H e does not lose time with a n unnecessary developing move with the rook, but immediately puts pressure on the cen­ tre. In the game, Cutmore tried to de­ fend the third row with 31 .i:!.e3, but Black h ad won time with 31 ... ct:Jc6. Of course, both pl ayers were by now short of time, and White h ad clear counter­ play with 32 f5, as we'll see in Test 14.2.

Test 12.4 C.Crouch-J.McKenna

London Open

2006

Continuing from Test 11.1, I pushed the passed pawn, with 31 b7? (A), but, of course, the bish op gets stuck in the corner. I asked myself, rhetorically, "when was the l ast game I played this badly?"

14 3

Why we L o s e a t C h e s s Certainly like all of us, I have lost some games incredibly quickly, but, beyond beginner and minor level, thi s would generally be the case of one ridicu­ lously bad move, completely misunder­ standing wh at was going on, and per­ h aps a couple of slight slips, based on a misunderstanding of the big error. Alas, after my illness I m anaged to lose quite a few of these miniatures, some of which h ave been published in this book. This game, though, i s different, and arguably worse. There must h ave been well over h alf a dozen serious mistakes in thi s game, and I am only giving a few sample exercises, to avoid m aking the reader feel bored. "Yes, I see the point. Now give up the pawn on b7, and your bishop will escape. You do not h ave to make a meal out of this." Statistically, there are likely to be more blunders in a poor game of 50 moves th an in a g ame of 1 5 moves. I was stunned by my previous blun­ der in this g ame, giving away an ex­ ch ange for nothing when I h ad a win ­ n i n g position. I was disorientated, but then felt a sense of relief that I still h ad the two united passed pawn s. If I h ad thought more closely, I would h ave seen that the pawn s were useless, and that it is time to simplify and aim for equality. Indeed, 31 tiJb3! (B) was best. If 3 1 ... c2, there would h ave been a temp­ tation to bring White's bishop into proper pl ay, with 32 3l.c6 ? ! , but as

144

Emms pointed out much later, 3 2 ... tiJe4! h appily picks up the extra f­ pawn . If, for example, 3 3 b7 3l.xf2+ 34 Wfl .i.g 3 35 tiJcl tiJxC5 36 .i.xd5 tiJd7, followed by ... .i.f4. Surprisingly, my po­ sition is not so dreadful after 32 Wfl, and in some lines Bl ack has to be care­ ful not to be worse. In stead, after the dreadful 31 b7?? .i.e5, my bishop was stuck in the cor­ ner, my knight was tied to the oppo­ nent's passed pawn, and my two proud passed pawn s could not break the blockade on c7 and d6. A mess.

3 2 'it>f1 c2 3 3 tiJb3 tiJd7 is h eading for a win for Bl ack. 33 ...tiJe4 is al so good, and if 34 c6? tiJd2+ win s a queen . In stead, 34 'it>f2 h olds the balance to some extent, it not even being impor­ tant whether White's c-pawn stays or drops. Then 34 c6 tiJC5 35 tiJc1 set up passed pawn s for each pl ayer, and de­ fended again st queening threats. We continue the discussion next tim e in Test 14.4.

T e s t T h i rt e e n 13. 1 Black to play

13.3 White to play

A) 3 1 ... 'iVxds 8) 3 1 ... �f7 C) Something el se?

A) 34 e4 8) 34 ttJel C) Something el se?

13.2 Black to play

13.4 Black to play

A) 3 2 .. .fxg 3 8) 3 2 .. .f3 C) Something el se?

A) 34 ... �h 8 8) 34 .. JlxdS C) Something el se?

14 5

Why we L o s e a t Ch e s s

Test 13.1 R.Randall-C.Crouch

London League 2006

We h ave seen thi s position before in Test 9.4. Now there was a whole cluster of blunders. I was feeling groggy, and my opponent was no doubt surprised and nervous when suddenly reaching a clear and good position after a blunder by his 1 M opponent. In time troubl e, both players made errors, and mine was the most ludicrous of them . I tried to regain the initiative with 26 ... f6 27 i.e3 'iYe4,

146

understandably enough, since I am a piece down, and if I pl ay quietly, I will lose. Even so, m aybe 27 .. .fS would h ave been better, keeping my pawn s active. He replied 28 ct:JC3 ! 'iYxb4 29 ct:Jd S, making my pawn sn atch look ex­ tremely inconsequential . His knight is now on the best square on the board.

Ali i could do was to set up a pin on the long diagonal, with 29 . . 'iYb7, with, one would expect, my best chance be­ ing th at h e would lose on time by thinking too h ard again st very minoT complications. H ere the simplest way is 30 i.g 2, g aining a tempo, in view of the threat­ ened discovered check with ct:Jxf6+. If 30 ... e4, the straightforward plan is 3 1 l:txc2 bxc2 3 2 'iyxc2, wh en White i s slightly ahead o n m aterial (bishop and knight versus rook and pawn), and his position i s totally secure, whereas Black's pieces are insecure. If 32 ... �xdS ? 33 'iYc8+, and White will be a full piece ahead. It is difficult to know wh at was go­ ing on in his mind some years later, but .

Tes t Th irte e n wh atever h appened, the line h e tried was not a m assive blunder, contrary to wh at I thought afterwards. He played: 30 1:[xc2 !? 30 bxc2 31 �xC2!? ( 3 1 .ig2 o r 3 1 iLh 3 also gives a n edge).

now played extremely quickly in time trouble, and a few seconds later, I noticed that I could h ave taken the knight: 31. ... �xd5!? (A) could h ave given a m assive shock to my opponent. Or, on the oth er h and, White might h ave analysed thi s right through, but h ad missed that after 32 �c8+ ct1f7 3 3 �xa6 Black was no longer i n check, and so h ad tim e to win the rook, staying a bishop up, with 3 3 ... �xh 1+. It was years before it was pointed out to me that, in fact, my opponent could h ave forced a win with 32 �xg 6+ �f7 33 .id7+ ct1f8 34 .ih 6+ 1:[xh 6 3 5 �xh 6+ �g 7 3 6 �C1, and White h as a winning attack after 3 6 .. J::t a 8 3 7 1:[g l. Or, since everyone loves a queen sacrifice, we h ave 32 ... ct1d8 33 1:[c1 �7 34 �e4! �a8 (34 ... �xe4 35 1:[c8 m ate) 35 1:[c8+ li'xc8 36 .ixc8 ct1xc8 37 �C4+ ct1b7 38 �5+, winning the rook. Here

3 3 . . .li'a8 sets up a counterattack with the threat of ... 1:[xa3+, but White de­ flects the rook with 34 .ib6+! I:txb6 3 5 li'f5, with a catastrophe on c 8 or d 7 . After the most spirited play, White h as to find a coupl e of little combina­ tion s, and a player might miss these in time trouble. The computer did not find the win in stantly, although it did not think for all that lon g . The clear implication is th at I was so shocked that I h ad earlier given away a piece for nothing, th at even a couple of years later I do not know how deeply my op­ ponent h ad examined all this. It would h ave been more entertaining for the spectators if I h ad seen th at I could 'safely' take the knight. And, of course, there would h ave been a slight chance for me to escape. Play finished with 31 .. .'>itf7 (B) 32 1:[gl f5 33 .ixf5 gxf5 34 !iJxe7 li'e4 3 5 li'C7, wh en I resigned (1-0), perh aps a touch prematurely ( 3 5 ... ct1e6 forces him to show the win), but ultimately fair enough. For the record, 32 ... g5 would h ave been best, so th at White cannot open the g-file, but Black would be unlikely to h old for too long.

Test 13.2

S.Gregory-C.Crouch B u ry St Ed m u nd s 2006 Not a particularly h appy day for a significant birthday celebration . Pl ay started steadily, with 1 e4 c5 2

14 7

Why we L o s e a t C h e s s tiJf3 e 6 3 d3 d S 4 tiJbd2 tiJc6 S g 3 .id6 6 .ig2 tiJge7 7 0-0 0-0 8 l:te1 .iC7 9 'ii'e 2 b6 10 tiJf1 a s 11 .id 2 .ia6, equal, with perh aps Black h aving slightly the better chances. White's play is not very ag­ gressive. Black's idea in the early part of the opening was, with ... .id6 and ... tiJge7, to prevent White from gaining a tempo with es, by avoiding ... tiJf6. White readjusts, playing the stan ­ dard king side attacking plan with h4h s , and tiJf3-h2-g4, but without es. He pl ayed 12 exd s exd s 13 h4 h6 14 tiJ1h2 'ii'd 7 lS h S , and Black might well start to think th at White h as overpressed. Then ls ...l:tae8 16 11ff1.

I played 16 ... fS, which I am not sure th at I like now, and judging by my ear­ lier notes, I did not really like at the time. I am gaining space, with pawns and pieces, but the pawn push al so creates weaknesses for my opponent to exploit, and these are arguably unnec­ essary. I would prefer nowadays 16 ... .ic8 ! . keeping the pieces highly coordinated, and stopping White even thinking about .ih 3 .

148

My idea in the game was th at after 17 .ih3?! d4, as played, White has been lured into a now redundant .ih 3, and Black's ... d4 push i s all the better, now th at the bishop i s not on g 2 . In stead. 17 lie2 ! , doubling o n the e­ file, looks uncomfortable, with Black h aving pressure on the a6-fl diagonal. No doubt it i s on thi s basis th at neither pl ayer would h ave examined this closely. Th ere is, however, the tacti cal point that after 17 ... c4 18 l:.ael cxd3 19 cxd3 .ixd 3 ? (19 ... tiJc8 i s equal) 20 l:.xe7 White is suddenly ahead on m aterial . 17 .. .f4 18 g4! ag ain might look ugly for White, but h e i s still slightly better. Back to the gam e, with 18 tiJh4. Now 18 ... 'iVds, trying to win White's h­ pawn, m ay be slightly prem ature after 19 .ig2 'iff7 20 .ixc 6 ! tiJxc6 21 tiJg 6 l:.xel 2 2 l:txel .l:tb8. and White h as kept equality. 50 I brought the bishop into play with 18 ... .ieS ! ? 19 tiJ2f3 .if6 20 .if4, and now it was time to try to win the h -pawn with 20 ...'ii'd s 2 1 .ig2 (if 2 1 .ie?, one possibility is t o call the oppo­ nent's bluff with 21 ... l:tc8 ! ?) 21 ... 1Iff7 !.

Tes t Th irte e n Thi s win s the pawn, but as so often the h ard work is yet to come, as the opponent has ch ances of gaining time. H e tried 22 ct:\g6 ct:\xg6 2 3 hxg6 'Yi'xg6, keeping his pieces better coordin ated. Then 24 .id6 .ie7 2S .iC7.

When pl aying through this position much earlier, I was highly critical of my next move, 2S ... bS, as the game ended up much l ater on with a weakened queenside pawn structure. I preferred in stead 2 S ... .id8, trying to keep the pawn structure tight. My comments at the time now seem over-elaborate, and the move I played was fine. It is only later that I started to m ake mistakes. 2S ... .id8 is playabl e, but after 26 .ixd8 l:txd8 n ct:\h4 'Yi'd6 28 'Yi'e2 White has well-placed pieces, and it is far from easy for Black to break down White's pawn structure. Bl ack's miss­ ing pawn i s only an h -pawn on the edge, and it cannot contribute directly to the central struggle, nor is it about to help start a pawn advance. Black can try .. .f4 in some lines, but g4 would be a standard respon se, taking control of

the king side light squares. I was also critical of my next few moves, after 26 l:te2 .id8 27 .ixd8 l:txd8 28 l:tae1 l:td6, but I still keep an edge, and surely there i s nothing wrong with that. It is only if I am losing, equal, or with a substantially smaller edge th an before th at I h ave gone wrong . Here I should be reasonably h appy. White still h as active pieces, but he is not breaking through, and Bl ack still has the extra pawn . It was later that I started to make mistakes. White started by keeping control on the e-file, with a knight exchange, 29 ct:\es ct:\xes 30 l:txes.

Looking at thi s diagram position, it can become understandable why I be­ came highly critical of my ... bS move, and indeed if White were still to have a pawn on h 2, he would have a clear edge. But the position is as it is. White has weakened his kingside, and so 30 ...f4 seem s highly sensible. Then 3 1 .ie4 'Yi'g4 (3 1...'Yi'f7 ! ? is al so worth con ­ sidering) 3 2 �g2, and we h ave reached the test position .

14 9

Why we L o s e a t C h e s s passed pawn, not yet immediately im­ pressive, but a threat later on . Al so White h as to try to keep his isolated a­ pawn safe.

3 2 ...f3? ( B ) was poorly judged, and, as I admitted at the time, this was "really the start of where I was about to go wrong". In other words, other slips are at best minor, possibly nonexistent. " For som e reason ", I noted, "I felt I could just batten down the f-file, but now of course Bl ack has no real king­ side advantage." I h ad blocked up the f­ file, when in reality it was essential to keep it open . Then 33 �hl covered all Black's threats, and now White h ad lines for his own pieces, as we'll see further in Test 1 3 .4. 32 ...jxg3 (A) 33 fxg 3 �c8 ! ? in stead leads to ten se pl ay. I am not convinced, though, about the computer's sugges­ tion that Black has a clear advantage.

Test 13.3

C.Crouch-R.Granat

B ritis h League (4NCL)

2007

As we saw in Test 9.3, Black h as, very reason ably, just declined a draw by repetition . He h as a supported

150

I decided it was about time to make use of the long diagonal , with 29 .i.f3 . I was surprised by his reply, 29 .. .fS?!, and indeed with his ... g5 pawn push later. There are certainly legitimate minority attacks in the late middle­ game, usually to create a weakness in the m ajority pawn s, but thi s i s not one of them . Before too long, White has a good and safe passed pawn, although, of course, Bl ack keeps his passed pawn as well. Maybe equal ? I h ad expected in stead 29 ... c5, followed by .. .'�jf6, and I would feel slightly uncomfortable as White. At the very least, it should be White, not Black, who h as to work h ard, and spend time, to m ake use of his king side pawn majority. White continues his piece develop­ ment, and Bl ack advances his pawn s, with 30 �b2 gS. I cannot keep the h alf­ open files with 3 1 �a3 ?, in view of 3 1 ... t2Jxe3 3 2 fxe 3 .l::t x e3 3 3 l::!. c 3 g4.

Te s t Th i r t e e n Therefore I needed to add an extra pawn defence on the kin g side, with 3 1 h3, and then came a counterattack with 31 J:td6 32 g4! fxg4 33 hxg4. ..

White should now be at least equal, but, of course, Bl ack h ad no need to allow White a passed e-pawn . We were now reaching time pres­ sure, and few players can then play with complete accuracy in complicated mid­ dlegame position s. It is easy to see that some of the next few moves are slightly inaccurate, but it is far more difficult to find the very best moves over the board. Both players have to assess whether to play a move instantly, or to think for 30 seconds, or whether to decide to think longer at a critical position, at the risk of hurrying later. It is much easier, of course, to determine the best move if the reader or annotator h as a computer, a fresh cup of coffee, and, above all, some time to think. 33 Wb7?! is recognisably a time pressure move. Black sees a couple of ideas, then argues that he will have no time to assess or an alyse the next few

moves, and s o h e quickly m akes an un ­ obtrusive move, which he h opes h olds the bal an ce, or m aybe more if the op­ ponent does nothing more than hold the bal ance. 33 ... 4Jf6 ! i s a more constructive move. If White tries to attack, with 34 4Jb4?, a natural enough move, Black hits back h ard with 34 ... ttxdl 3 5 ttxdl 4JC4+ 36 Wb3 4Je 5 . White h as t o improve earlier. The computer sugg ests 34 l::t c 2 l::t e d8 3 5 Wc3, not easy t o think of while being short of time, as Black's last couple of moves look almost random . The com ­ puter implies that the position is close to deadlocked, and equal, since White cannot break Black's pressure on the d­ file, while Bl ack cannot create extra pressure, and he will not want to re­ lease the pressure him self. Black might still create a slight edge, though, with 3 5 ... c5 36 l:tcdl 4Jc6 ! , and the knight is back in play. If 3 7 l:tc2, hoping for a repetition, Black keeps an edge with 3 7 ... .l::i. e 8. Back to the game; it i s White's turn .

...

15 1

Why we L o s e a t Ch e s s I sensed that Gran at's play was in­ accurate, and felt I could even play for a win. I played perh aps too h astily to take the initiative, overlooking my op­ ponent's reply. 34 e4 (A) looks good, but thi s proved to be deceptive, as after 34 ... .l::i.f6! it was Black who was starting to create pressure. This I found to be an unexpectedly difficult middlegame to an alyse, and therefore an interesting test position, I hope, for the reader. Of course, the test is one of positional sense, rather th an tactics. To cut down a couple of pages of analysis, I give only a brief outline. 34 e4 was not a positional mistake. It is equal . My mistake was only on move 40, just before the time control . In stead, 34 0.b4!? (C) 34 ... .l::i.f6 ! 3 5 i.xd5 J::t xf2+ 3 6 J::t c 2 i s equal . And 34 0.el!? (B) 34 .. J:tf6 ! ? 3 5 J::t c 2 0.b4 3 6 110 0.d5 3 7 .l::i. c 2 0.b4 is a repe­ tition . I was fascinated by the thought that here White could h ave tried 3 7 l1Cl ! ?, a rem arkable zugzwang idea.

We h ave seen thi s position before,

15 2

but with Black to move, rather th an White. The computer suggests that it is better for White if it were Black to play here rather th an White to play. If so, this would h ave been a rem arkably delicate zugzwang, with nine pieces and pawn s on either side. Every one of Bl ack's moves, except only for 37 ... l1ef8 ! , leads to a worsening of Black's play. The fin al result should end up as a draw, after 3 8 i.xd3 ttxf2+ 39 'it>a3 cxd5 40 J::t x d5 .l::i. e 8 41 .l::i. d 7+ 'it>b8 42 ttC3, and if 42 ... .l::i. x e3 43 .l::i. d 8+ 'it>b7 44 J::t d 7+ �b8, with a draw by repetition. 44 ... 'it>a6? would h ave been far too ambitious, as White has a big m ate threat after 45 .l::i. x e3 0.C4+ 46 'it>b3 0.xe3 47 0.d3 . But h ow m any players would h ave seen all thi s in ad­ vance? Back to the g am e, and I delete a few earlier criticisms on my next few moves, with 3 5 i.g2 0.f4 3 6 0.xf4 .l::i. xf4 3 7 f3 'it>b7 38 'it>C3 0.b7.

Sensible so far, but the next move was not good, as we'll soon see in Test 15.1.

Te s t Th i r t e e n

Test 13.4

S.Gregory-C.Crouch

B u ry St Ed m u nd s 2006

In Test 1 3 . 2 we h ave seen a psycho­ logically dangerous part of the game, when a player feels h e has been in con ­ trol for most of the time, and then something unexpected h appen s, close to the time control. There is little chance to readjust emotionally, no time to wander around or take a tea, no time just to chill out for five min­ utes. The pl ayer has been h appily play­ ing for a win, then a few seconds l ater, he does not know whether his position is better or worse, or whether he should attack or look for a draw. Quite often, a mistake in time trouble can easily lead to a second, m aybe worse, slip the next move or the second move. This h appen s shortly in thi s game. 3 3 ... c4 was reason able. I can do nothing immediately threatening, and so I bring my more active pawns into pl ay, and hope to open something up

for the bishop. Then he gave a check, 34 .i.d S + , with three reason able replies. There are two opposing weaknesses when playing in time shortage. Either the pl ayer some­ times plays like a jelly, just dribbling through to an endgame, and hoping to regain composure after the time con ­ trol . Or, alternatively, it i s possible to play the sharpest move, with the hope that it i s best, and in any case one can h ope that the opponent will m ake a mistake under pressure.

But h ere, which plan is better? There are two choices, sacrificing the exch ange, and quietly moving the king . This takes time t o reflect, and I got it wrong . Almost without thinking, I sacri­ ficed the exch ange with 34 ...l:txd S?? (B). I felt sure that there was something there for me, but there wasn't. After 3S l:txd S cxd3 36 cxd 3, if 3 6 ....i.b7, then White h as 37 'YWh S ! , easy enough to miss, and he win s. The desperate sacri­ fice 3 7 .. :�xg 3+ 3 8 fxg 3 f2+ 39 �f1 does not work.

153

Why we L o s e a t C h e s s I struggled o n with 36 ... b 4 3 7 �h4 'iVxh4 38 gxh4 .i.xd3 39 l:txd4 .i.e2 40 l:tc1 l:tf6 41 l:tg4 g6 42 l:tc5, and played a few more moves, but he did not lose on time, and so he won . Instead, 34 . .'JithB (A) should be a draw, although perh aps slightly the worse of the draw. For example, 35 l:te8 l:td8 36 l:txd8 l:txd8 37 .i.xf3 �g s 3 8 'ifh s 'iVxh s 39 .i.xh s cxd3 4 0 cxd3 l:tc8 41 .i.g6 .i.b7, a small edge to White as the bishop on g6 i s uncomfortable. .

154

Alternatively, 35 l:t1e4 'iVg 6 36 'iVxf3 l:txf3 3 7 l:te8+ �h 7 3 8 .i.g8+ Wh 8 39 .i.f7+ with perpetual . Or in this line, 3 s ... �c8 3 6 dxc4 bxc4 3 7 l:txd4 c3 3 8 b3 IUS 39 l:txfs �xfs 40 .i.e4 l:txd4 41 .i.xfs l:td1+ 42 'it>h 2 l:txh 1+ 43 'it>xh 1 gs, and the position i s blocked after, for example, 44 .i.e4 .i.e2, probably ending up as an unusual draw. There is still some life in the game, so even the quiet and uninspiring move can blossom into interesting play.

Te st F o u rt e e n 14. 1 Black to play

143 White to play

A) 3 S ... .l:i.g8 8) 3s .. .'!iJe4 C) Something el se?

A) 3 7 tLJg s 8) 3 7 l:txdS C} Something el se?

14.2 Black to play

14.4 White to play

A) 3 S .. J:tg4 8) 3 S .l:i.dl+ C} Something el se?

A) 3 8 'it'xc2 8) 3 8 b8'i\i' C} Something else?

...

155

Why we L o s e a t C h e s s was fortunate th at I had an opportu­ nity to keep some counterplay.

Test 14.1 C.Morris-C.Crouch

British League (4NCL)

2006

As we saw in Test 10.1, I am in trou­ ble h ere, with several pawn weak­ nesses, and some holes for my oppo­ nent's pieces after my pawns h ave been forced to advance. After some thought, I found a way to keep my po­ sition alive, giving away a couple of pawn s, while creating some much needed pawn activity. I tried 30 ... a4 31 tLla s ( 3 1 tLlcS is solid enough, and keeps an edge, but does not attack any of the pawn s), and now the pawn s went over the top with 3 1 .. . CS!. It is an important defensive princi­ pl e th at if your pawn s are in poor health anyway, and some will drop be­ fore too long, it is a good idea to en sure th at when the opponent starts picking up the pawns, you can create piece complication s while the opponent spends time chewing up the pawns. I

156

Obviously 32 dxcS ttxes allows Black to recover the pawn, and Black's pieces are now active, with the rook being an annoyance. Therefore my op­ ponent took the other way, with 32 ttxcS?! tLle6, but now Black suddenly h as tactical resources.

'Therefore' is a dangerous word in chess, and White should probably h ave played 3 2 tLlc6 ! tte6 3 3 ttxcs instead. Black no longer has the ... tLle6 option. After 33 ... tLla6 34 .i:f.cl .ll d 7 Black is a pawn down, while White's pieces are solid. Perhaps there is no quick win for

Te s t Fo u rt e e n White, but he should eventually prevail. In stead, after the text move, the po­ sition i s still alive.

If now 33 l:txb S ? tLlxd4, forking the two rooks, and Black is ahead. There is, it h as to be admitted, an element of bluff in attempting to es­ cape from such defen sive break-outs, and to some extent I was relying on the thought th at 33 l:txdS looks risky for White, when previously he h ad such a safe edge. Who wants to take risks when you are already ahead? I could not try to pretend th at I h ad calcul ated through to a satisfactory conclusion, but then there i s a reasonable probabil­ ity that most players would be un able to find a totally clear advantage either. Here 33 lIxdS lIC7 (also maybe 33 ... lIa7) 34 tLlxfs lIcl+ 3 5 Wf2 lIc2+ 3 6 We3 lIxb2 would undoubtedly add t o a touch of fear for White to think about, and if 37 lIh 3 tLlC7 38 lIxh6+ �g8 39 lIcS i.xfs 40 l:txc7 l:txa2, when White, although maybe better, has to deal with the implications of two connected passed pawns for the opponent. In prac-

tical terms, considering Black's earlier dreadful position, this is a success. So White tried instead 33 tLlxfS lIa7 34 lIxbS tLlgs, when arguably White's position i s even more critical .

White has two extra pawn s, cer­ tainly, but both his rooks and his knights are under attack, or potentially under attack. I was in with a chance, whereas a few moves previously, I h ad felt my position was close to resign­ able. Then in the h otel, the alarm bell went. Thi s for m any years has been almost a tradition al entertainment on the l ast days of various tournaments in England, m aybe through m alice. There h ave been several other occasion s in recent years in the 4NCL, but these, as far as I know, h ave h ad innocent expl a­ nation s. In a big hotel, with so m any people wandering around, especially after most of the g ames are finish ed, accidents h appen. The story on this occasion i s th at one of the players propped his elbow on the bar on one of the counter ledges, which h ad been

157

Why we L o s e a t C h e s s shut, and the alarm was set off. I n other words, a complete accident. There was not much chance to con­ centrate with a noisy fire alarm. When all the players had to escape from a crowded room I am sure that many players would have been extremely dis­ rupted during the last few minutes of the time control. But what can be done? Upon resumption, I completely lost my thread of thought, and lost even before move 40, starting with the blunder 3 5 1'!.f4 ttJe4?? (B), overlooking that h e could take the central pawn after 36 ttJd6 (or 3 6 1'!.xdS immediately) 36 ... �a6? (a quick loss, but 36 ... ttJxd6 37 exd6 ..te6 38 ttJb7 only slightly de­ l ays the result) 37 1'!.xd 5 ..td3?? (another blunder) 38 ttJxe4 1-0. This was sickening. I could h ave h ad no complaints if I h ad lost as a result of my poor earlier play, but losing as a result of a fire alarm, when I was back in the game after much effort ... Well, this was tough. After the bell, 35.. Jlg8! (A) was, of course, much better.

Then 3 6 ttJg 3 (Black was threaten­ ing ... ttJh 3+) 36 ... ttJe6 37 .l:th4 .l:tC7, and I am reasonably certain that I h ad con­ sidered this position in advance, but the chaos of evacuation disrupted me, and my confidence evaporated. A pos­ sible and likely drawing line would be 38 .l:txdS .l:tc1+ 39 'it>f2 .l:tc2+ 40 'it>e3 1'!.xb2 41 ttJfS 1'!.gxg 2 42 1'!.xh 6+ '.t>g 8 43 ttJe7+ 'it>g7 44 ttJfS+, repeatin g . The computer suggests 43 ttJg 3, and if 43 ... 1'!.xa2, White h as a good escape route with 44 1'!.d6 ! , then 'it>e4-ds, but 43 ... .l:!.cc2 ! leads to a draw. There is much to be explored in analysis. In practical term s, White would h ave h ad to find good moves in a complicated position up to move 40.

Test 14.2 M.Cutmore-C.Crouch

Kid l i ngton

2007

From Test 12.3, clearly the position is about to open up, on both sides. Play continued 3 2 ... 1'!.xd4 3 3 �bl �xf5 34

158

Te s t Fo u rt e e n 40 .l:i.e4 .l:i.h 1+ 4 1 'it'g 2 ltcg 1+ i s a draw. After my panic, Cutmore played the simple and natural 36 .i.f6!, and I was in trouble. 36 ... .l:i.g6 37 e6 was unentic­ ing for Black: for instance, 37 ... tiJe7 3 8 exf7+ �xf7 39 .l:i.xe7 ltc1+ 4 0 �h 2 �xf6 41 �xdS+ 'it'h 8 42 Vixh S+ 'it'g 8 43 VidS+ 'it'h 8 44 Vid8+ l:tg8 4S l:th 7+, picking up the queen . So, 3 6 tiJd4 3 7 Vixh S .l:i.g7.

tiJxfS gxfs 3 S VixfS.

...

I pl ayed 3 S .l:i.g4?? (A). In my notes, I said that, "as so often, a panicky reac­ tion, short of time, m akes things worse. My last few moves appear to h ave been correct, but my king i s suddenly ex­ posed. Oh dear, h elp, bring another piece closer to the king, and hope for the best - but thi s is not good chess thinking ! I h ave played attacking de­ fence, and should carry on with this, forcing White to go on the defensive. That way I can keep the balance and even more." I should h ave played 35 1:.dl+ (B) 3 6 'it'h 2 tiJd4, g aining a tempo to bring extra pieces into play, the knight, and the rook on c8. It still looks fragile, but White's queen must move, and cannot find a good way to attack the rook on d1. Then 3 7 Vif2 .l:i.cc1 3 8 .l:i.e1 .l:i.xe1 39 .i.xc1 l:txc1 40 Vixd4 Vie7 41 g 3 gives an advantage to Black, but it will be diffi cult to force the extra pawn to a win . Instead, 3 5 tiJe7 ( C ) 3 6 .i.xe7 l:tC1+ 3 7 'it'h 2 Vixe7 3 8 g3 .l:i.dd1 3 9 �xh S d4 ...

•••

• • •

My position was close to resignable, but my opponent made two mistakes in a row, in time trouble, starting with 38 .l:i.g3?, overlooking my reply, or at the very least overlooking my follow-up. 38 .i.xg 7 ! was simple enough, then 38 .. .'iVxg 7 39 'it'h 2 .l:i.c6 40 l:tg 3 .l:i.g6 41 l:txg 6 fxg 6 42 Vig s, keeping an extra exch ange, with White's queen and rook being highly active. Straightforward chess, but unspectacul ar, and my op­ ponent was presum ably wanting a quick checkmate, or a win of the queen . I squirmed on with 38 tiJfS!, some­ h ow holding on with sticking tape. If 39 VixfS .l:i.xg 3 40 'iVh s .l:i.g 7, level. With care, White can still keep a highly sig...

15 9

Why we L o s e a t C h e s s nificant edge, though, with, for exam­ ple, 3 9 .l:i.f3 lbd4 40 .l:i.d3 lbe6 41 ..txg 7 'iYxg 7 42 .l:i.f2 . Instead, however dire my position was, I suddenly had counterplay after 39 .l:i.gS?? .l:i.C1+ 40 'iit h 2 "iVCS.

46 .l:i.g 2 is a draw by perpetual . There is nothing more after 46 ... 'iWf5+? 47 �xf5 lbxf5 48 .l:i.f2.

Test 14.3

C Crouch P Roberson .

-

.

B ritis h League (4NCL)

I h ave at least a draw! N aturally, I could not expect m ore. However, after 41 .l:i.xg7+ lbxg7 42 �gs "iVg1+ 43 'iit h 3 .l:i.c3+, there was the biggest oversight of all, with

44 'iit g 4?? 'iWd1+ 45 'iitf4 �d4 mate (0-1). An embarrassing result, even for the winner. In stead, 44 g3 'iVh 1+ 45 .l:i.h 2 �f1+

160

2007

By the end of play in Test 12.2 Bl ack was two pawn s up, but he h ad isolated doubled pawn s on the f-file, and it looked more than likely th at at least one of his queenside pawn s could fall, with his minor pieces being under threat, and therefore being un able to cover his pawn s. I felt relieved that I h ad recovered from my earlier mistake, and I was hoping that I could try to play for an edge .

Black immediately returned the doubled pawn, with 33 ...f4 34 i.xf4 lbe6. Thi s is not the only line, but seem s to be the most sensible. The computer suggests 3 3 .. J::t x e2+ 34 Wxe2 i.C4+ 35 'iitf2 lbd5, 'equal ', but if White can regroup his pieces to more effec-

Te s t Fo u rt e e n tive squares, Black would h ave to careful after 3 6 as f4 37 �h 2 b S tDd2 ! . I n the m ain line, White played .i.e3. Black's bishop i s now under tack again.

be 38 35 at­

In my contemporary notes, I implied that Black, with his extra pawn, should be comfortably equal, and that h e should h ave chances o f being better. Therefore I g ave 35 J:td8 a query, but defending the bishop is a natural enough idea, with chances of h olding the draw. In stead, any bishop move would allow a quick l::t d 6, with chances of chewing up Black's queenside pawn s. It turns out th at Black i s slightly worse in other lines, except for one, so perh aps .. J:td8 is a ' ? ! ' move rather than a '?' move. The best line for Black i s 3S ... tDC7 ! , and then i f 3 6 .if4 tDe6, repeatin g . White h as choices, for example, 3 6 as bxas 3 7 .ixcS fS ! , but no real advan­ tage. In stead, 3S ... �xf3 ? ! 3 6 gxf3 (best)

3 6 . . . as 3 7 lId6 tDg s 3 8 �C4+ 'it>f8 3 9 �f1 ! gives White a clear edge. The point of the interpolated check is soon seen after 39 .. J:tb8 ? ! 40 l::t x b6 l::t x b6 41 �xcS+, and Black's queenside col­ l apses. Another attempted spirited line is 3S ... .i.b3 36 l::t d 6 tDC7 37 .l::!. x b6 .l::!. x e3 38 .l::!.b 8+ (38 'it>xe 3 ? ? tDdS+) 3 8 ... lIe8 39 l::t x b3, but Black's four isol ated pawn s will be far less effective than White's two connected sets of pawn pairs. In this line, 3 6 ... .i.xa4 3 7 .i.C4 Wf7 3 8 .l::!. x b6 �bS 39 .i.xbs axbs 40 .l::!. xb S fS 41 b3 i s al so difficult t o defend for Black. Back in the g ame, I tried 36 �xa6, h aving noted that 3 6 lhd S ? lhdS 3 7 .i.C4 lId6 3 8 .if4 l:!.c6 3 9 .i.ds f7 3 9 ii.d3 cxb4 40 axb6, with perhaps a slight edge for White. What, however, i s the problem with the other pawn move, in the game? The answer i s that Black has m ade a slight weakness in his pawn structure, the f­ pawn no longer defending the e S - or g s -squares. I h ad looked at ttxds on move 36, a move earlier. I did not at­ tempt to analyse the position a move later. I h ad already spent time analys­ ing it the previous move, and did not want to eat up extra time on the clock. As a result, I missed a chance of win­ ning.

16 2

I h ad the chance of playing 3 7 �d5! (B) 3 7 ...ttxds 3 8 ii.C4 ttd6 ( 3 8 .. .tbc7 39 ii.f4!) 3 9 .tf4 .1:.c6 40 tbg S ! , which i s now winning. Maybe the moral i s th at you should not completely abandon the sacrifice you h ave rejected. Who knows, a little pawn tweak, or a small, quiet piece move, could alter one's perspective quickly. 37 tbg5?? (A) was an awful move anyway. Black h as now found good squares for his pieces, with 37 ... tbd4!. After 3 8 ii.f4 h6 39 tbf3, I suddenly no­ ticed that 3 9 ... tta8 ! was strong for Black. Fortunately, a second or two l ater, another piece l anded, with 39 ... tbxf3? I finally m ade the exch ange sacrifice with 40 ttxd 5 ! ttxd 5 41 ii.C4 tbe5. Then 42 jLxd 5+ 'i.t>f8 43 'i.t>e3, and I h ad h opes of keeping a slight endg am e edge with two bishops versus bishop and knight. It was not to be, and he pl ayed accu­ rately with 43 ... 'i.t>e7 44 b3 'i.t>d6 45 ii.g8 'i.t>e7 46 ii.g3 tbc6 47 ii.e1 'i.t>d6 48 ii.g3+ 'i.t>d7 49 .tf4 tbe5 50 ii.h2 tbc6 5 1 ii.C4

Te s t Fo u rt e e n �d4+ 52 �e2 �e5 5 3 �gl �d4 54 �e3 �xe3 55 �xe3 �e7 56 �g8 �f6 57 �d3 tbe7 58 �c4 �e5 59 �a6 �d6 60 �b7 tbg6 61 �C4 tbf4 62 �b5 �C7 63 �f3 tbe6 64 �C4 �d6 65 �b7 tbf4 66 b4 cxb4 67 �xb4 tbd 3+ 68 �b5 �C7 69 �d5 tbf4 70 �f3 tbe6 71 �a8 tbf4 72 g4 fxg4 7 3 hxg4 tbg6 74 �e4 74 ... tbe5 75 �f5 tbf7 76 �e6 tbe5 77 �b4 �d6 78 �c8 tbd 3+ 79 WC4 tbC5 80 �b5 �e5 81 as bxa 5 82 �xc5 �f4 Yz-Yz. A frightful number of years ago, I won the British Under-18 Ch ampion ­ ship with a string of minor endgames, solely on the basi s that I was able to understand such endgames better than my opponents, who m ade mis­ takes. Nowadays the young defenders seem to pl ay more accurately!

. . .�e7 would tran spose i f he tries . . .�C7 then ... �e6. There was, however, th e more worrying prospect of 3 5 ... d4 fol ­ lowed by ... d3, and a severe danger of being zugzwanged. Therefore I was relieved that h e played quietly, a n d I saw the opportu­ nity of grabbing Bl ack's advanced c­ pawn, with 38 �e2 �f8 37 �d2 �e7. Excellent, I thought, take the pawn, 'job done'.

Test 14.4 C.Crouch-J.McKenna

London Open

2006

Following on from Test 12.4, Black played 35 ... �C7. 3 5 ... �f8 followed by

Except th at the highly natural 38 �XC2?? (A) was a serious mistake. After 38 ... �d6!,

Black is not only sitting next to the passed c-pawn (he cannot, of course,

163

Why w e L o s e a t C h e s s take i t yet, a s White would reply with b8�+), but more significantly, the king i s now protecting the ds-pawn, so that any breakout sacrifice with b8� fol ­ lowed by C7 no longer allows White to win Black's ds-pawn . My last legitimate chance to save the game was 38 b8�+! (B) 38 ... �xb8 39 C7 lLle4+ 40 'iii x c2 �xC7 41 �xdS lLlxf2, and although Black i s a pawn up, his position i s totally unwinnable, with no passed pawn s, no pawn weak­ nesses, and with opposite-col oured bishops, leaving no chan ces of a zugzwang . In the game, after Black played 38 ... �d6, White somehow convinced himself that he was a pawn up, and therefore had no reason to worry about losing. A big mistake. It was time to grovel by sacrificing the passed pawn on the seventh, bringing the bishop back into play, although he still h as to work hard to try to set up a draw.

164

In stead, 39 lLla2? lLla6 40 lLlC3 lLlb8 was by now a clear win for Black, even though White has two advanced pawns.

I tried 41 lLlbS+ �xc6 42 lLlxC7 'iiix C7, with the expectation th at I was going to resign in a few m oves. After all, White's bishop and pawn are immo­ bile, and Black's knight h as no incen­ tive to move, so in effect Bl ack is a pawn up in a king and pawn endg ame. We'll see wh at h appened l ater in Test l S . 3 , with a surprising result.

Te st F i ft e e n

15. 1 White to play

15.3 Black to play

A) 3 9 'it>d4 B) 39 l:.h l C) Something el se?

A) 4B .. .'�JbB B) 4B .. .fxg4 C) Something else?

15.2 Black to play

15.4 Black to play

A) 46 ... e4 B) 46 ... 'it>fB C) Something el se?

A) SO . .. .l:td2+ B) SO .. .lDf2+ C) Something el se?

165

Why we L o s e a t C h e s s

Test 3.5.3.

C.Crouch-R.Granat

Britis h League (4NC L)

2007

Following on from Test 13.3, I blun­ dered h ere, losing a pawn with 39 'it>d4? (A). I was con scious that my other pieces were not very active, and so I wanted to bring my king into play, overlooking th at I would lose a pawn with 39 ... .i:!.a8 40 .i:!.a1 ttJc5. Black threaten s a knight fork on b 3 . Clearly i f I h ad had the chance of thinking before the time control, I would h ave seen this tactic in advance. It would h ave taken much more time to find a wholly convincing way to equal­ ize, though, Black's pieces being better pl aced than White's. 39 1ihl! (B) 39 ... .i:!.f6 40 l:i.h 2 ! i s best, putting the rooks into wh at seem s to be a blind corner on the h-file, but White is al so doubling on the only ef­ fective file. If, for example, 40 ...ttJC5 41 'uch 1 'uee6 42 l:i.a1, and Bl ack can m ake no real progress. Bl ack could try in stead

166

41 ... ttJxa4+ 42 'it>b4 ttJC5 43 'uxh 6 'uxh 6 44 .i:!.xh 6 ttJd3+ 45 'it>C3 ttJf4 46 ..ttl, but White should be able to hold. All this was quite a surprise for me when I resumed my analysis while writ­ ing this book. I h ad assumed that while I knew that my 40th move was a losing mistake, I was in a bad shape already, but it turns out that I was equal . In stead, after the pawn dropped, Bl ack won with 41 'it>e3 ttJxa4 42 ..tf1?! (a lack of confidence, following my er­ ror; the only chance was to attack, with 42 ,Uh 1 .i:!.f6 43 e5 .i:!.e6 44 f4 gxf4+ 45 'it>xf4 'uf8+ 46 'it>g 3 b5 47 'uh 5, when Bl ack should be winning, but he would h ave h ad to work h ard) 42 ... b5 43 ..te2 .i:!.af8 44 .i:!.h1 .i:!.8f6 45 .i:!.h3 .i:!.e6 46 ..td1 ttJb2 47 ..tb3 .i:!.d6 48 'it>e2 .i:!.d 3 49 ..te6 .i:!.xe4+ 50 fxe4 .i:!.xh 3 51 ..tf5 ttJc4 52 .i:!.a7+ 'it>b6 0-1.

Test 1 5.2 D.Buckley-C.Crouch

B ritish League (4NCL)

2006

Te s t Fift e e n White is still better after mediocre play by both sides. I h ad missed a clear win a few moves ago, as we saw in Test 11.3, and now he is on top. He has a clearly better queen and rook ending. It is easy enough for him to create a well ­ guarded passed pawn o n the queen­ side, whereas Black's extra pawn on the kingside m ay be vaguely useful, but is unlikely to do anything con structive. My opponent immediately threw away h alf his advantage, by offering the exchange of rooks, with 3 2 1:td6?, in stead of trying to strang Ie the queen­ side with 3 2 'iWdS ! . Indeed, White soon lost the whole of his advantage, as Black now fully equalized with 32 ...f3 33 1:txg6+ hxg6 34 'iWd S.

We were approaching the time con­ trol, and I sensed th at the psychological initiative was now slightly in my fa­ vour, and so tried 34 ... 'iWf4+, instead of the simple drawing line 34 .. .fxg 2 3 S 'it>xg 2 'iWg S+ 3 6 'it>f1 'iWC1+. Then came 35 g3 'iWd4, on which I commented: "The best move, but at the

time I thought i t was a blunder... 3 6 'ii'xf3 ... a s I h ad missed this." Still . it wasn't too bad as I could immediately regain the pawn with 36 .. :i!VXC4. It was disconcerting, all the same, th at I h ad not seen his simple capture.

The position i s now equal, but I was still h oping to squeeze an edge, out of sheer pride. Or perh aps punch-drunk chess. We quickly reached the time control with 3 7 'it>g2 a s 38 bxa s bxa s 39 'iWe3 'iWd S+ 40 'it>h2 'it>f6. As I noted earlier, "A wrong change of direction, ... and eventually I h ad to find a way to return the king to g 7 . I h ad vaguely thought that I could pressurize White by trying to exch ange the queens, but White's queen was too active to allow Black to attack." There i s the further danger that White has the possible threat of the outside passed pawn, if given the chance of h4, g4, h S , and then Black's king would be seriously displaced if White could create a passed pawn . Pl ay continued 41 'iWb6+ 'it>e7?! (still the wrong direction) 42 h4.

167

Why we L o s e a t C h e s s king stays in the centre a little longer) 46 g4, and I still h ave to play accurately.

There is still some life in the g ame. If 42 .. .'i!t'd4 43 'it'xd4 exd4 44 'iit> g 2 'iit> e 6 45 g4 f5, and there are several ways, as far as I could see, to draw, one being 46 g5 f4 47 f3 (but not 47 'iit>f3 ? ? d3, and White is in zugzwang) 47 .. .'it>d5 48 �f2 a4 49 �e2 'it>e6 50 �d3 �d5 5 1 �e2, when neither side can m ake progress. I also felt vaguely uncomfortable af­ ter 42 ... 'iit>f8 43 'it'f6, but maybe this was a safe level draw. My 42 ... �d2 was al so possible,

but now it was his turn to think about playing for m ore, with 43 'it'b7+ 'iit> e 6 44 �c6+ 'it>e7 45 'iit> g 2 'it'el (45 ... 'it>f8 ? 46 iVC5+ loses a pawn, so the

168

The position is, of course, equal. The only likely way of losing such a posi­ tion, with either colour, would either be to overpress, trying for a win, or by a blunder. Somehow, though, both play­ ers managed to go wrong . I played t h e unnecessary 4 6. . .e4? (A), which vaguely m akes possible some dangerous checks on f3, but also m akes weaknesses on Black's own squares. It was much more precise to pl ay 46 ... �f8 (B), covering the opponent's potential passed pawn . If 47 hS g xh s 48 gxh s 'iit> g 7 49 h6+ 'iit>h 7 S O 'it'f6, and the position might look scary, but Black is fine after 50 .. :�e4+ 51 �f1, and then a perpetual after 51 ... 'iVb1+. Here 5 1 �h 2 keeps the play going, but the pawn ending is a draw after 51 .. .'�Vf4+ 5 2 'it'xf4 exf4 53 'iit>h 3 . Buckley's queen move, 4 7 iVC4?!, was a square too far. He missed his big chance to create some pressure, with 47 "1i"C5+!, which forces the king to an un-

Te s t Fift e e n desirable square. There would be some uncomfortable zugzwangs later on. Black's queen looks good at first, and I was relying on this, but while the queen is on an advanced square, it h as little impact further behind. There is only one square, on e1, to protect both the e4and a5-pawns, and this is the reason why there could be zugzwang problems.

Here 47 ... �e8 ? ! 48 g5 �d7 ? ! 49 "iVf8 �e6 50 'ilVe8+ �f5 5 1 'iVxf7+ �g4 5 2 "iVxg 6 'iVe2 5 3 "iVe6+ �xh4 54 'iVf5 will sooner or later queen a pawn for White. Therefore Black cannot move the king again h ere, and so has to rely on a queen m ove, 48 .. :iVe2 ! . Then 49 'iVxa5 'ilVg4+ 50 �f1 'iVxh4. Maybe it shoul d eventually end up as a draw (though there is an element of doubt), but Black h as to work h ard. There would be much m anoeuvring by White to try to squeeze a slight edge after 5 1 'iVe 5+ �d7 5 2 'ilVf6 � 3 + 5 3 � e 1 'i¥b3 54 'iVd4+ �c7 5 5 a5 'iVc2. Of course, if I h ad played precisely on move 46, it would h ave been a very simple draw. Could Black play better with

4 7. . .�f6, with a draw i f White allows the king to the corner? I was worried about 48 'iVf8 'iVe2 ! 49 g 5+ �e6 50 'iVe8+ �d6 51 'iVd8+ �c6 ( 5 1 ... �e6 ?? 52 "iVf6+, winning a pawn with g ain of tempo) 52 'iVxa5 'iVg4+ 53 �f1 'iVxh4, but it is likely still to be a draw. In the game, I played 47 ... �f8 with a sigh of relief. He tried 48 'iVC5+, a move too late. Then 48 ... �g8 49 'ilVe3 .

I n o w saw his idea, th at with the obvious queen exch ange, White ends up winning after 49 .. :i:Yxe 3 ? ? 50 fxe3 f6 5 1 g 5 fxg 5 5 2 hxg 5 �f7 5 3 �g 3 �e6 54 �f4 �d5 5 5 a4. But one of the most basic position al rules of the endgame is never to be absolutely certain of what is going on in a pawn endgame, as there i s so often such a small gap be­ tween winning, drawing and losing. There i s more latitude in piece end­ g ames, and so I kept the queen s on, 49 . "iiYa 1. I doubt whether I could h ave concealed th at I was starting to feel tired, and so instead of offering a draw, he made a last winning try with 50 h 5 'iVe5. ..

16 9

Why we L o s e a t C h e s s His sister, Melanie, said th at David ever so often completely forgot about his time, even in the last few minutes.

Test 1 5.3 C.Crouch-J.McKenna

London Open

And now he was the one t o blunder, with 51 h6?? ( 5 1 hxg 6 fxg 6 is, of course, a draw) 51 ... g5, and Bl ack soon pock­ eted a pawn . After 5 2 f3 exf3+ 53 �xf3 �e6 54 �h3 �h7, there was an even bigger surpri se. I was about to be a pawn up, but it was not clear th at I genuinely h ad a winning advantage. My opponent was thinking hard, while I was trying to think of ways of making progress, and gearing up to h aving to play ag ainst quick moves from my op­ ponent, short of time. But then he just thought and thought, and lost on time. I was confused. H ad h e m ade move 60 or not? Had I got the time control wrong ? H ad I misread the digital clock? Quite possible when I could see with only h alf an eye. Or, most unlikely of all, did he not notice that he h ad lost on time? After a brief gap, I pointed to the loss on time, still slightly confused. Here, the traditional fl ag-fall would h ave been clearer, but with a digital clock, I h ad so m any bits and pieces to look at. To my relief, David agreed that this was, indeed, a loss on time, so 0-1.

170

2006

Following on from Test 14.4, all I could do was to move my king, plus a few weakening pawn moves. It i s a simple win for Bl ack, the sort of exer­ cise th at you would use to teach chil­ dren . All that Black needs to do is to concentrate on the king and passed pawn, and White's king will gradually h ave to give way, because of zugzwang . Then after a few more zugzwan g s, the d-pawn will be promoted, with a win for Black. I played a couple of token moves, waiting for the opponent to demon­ strate th at he showed h e knew h ow to do it. I did not expect this to last for lon g . I centralized with 43 �d3 �d6 44 �d4. There was still a glimmer of hope

Tes t Fift e e n when he pl ayed 44 ... tt'lc6+?!, still win ­ n i n g , but not so direct, since Black will h ave to play the knight back to b8 at some stage. The most direct plan for Black would h ave been 44 .. .f6. The computer still initially suggests that White is better, but thi s is clearly mis­ taken. After, for example, 45 'It>d3 h5 46 'It>d4 'It>c6 47 f4 h4 48 'It>e3 'It>C5 49 'It>d3 d4 50 'It>e4 'It>C4 51 'it>f3 'It>c3 it i s easy to see that the pawn will queen . After 45 'It>e3 f5, I could h ave con­ tinued pl aying rather longer with 46 b8�+ ! ? tt'lxb8 47 'It>f4 tt'ld7 48 'It>g 5, al ­ though Bl ack eventually win s with ... 'it>e 5 and ... tt'lf6. I pl ayed the opponent h ere, rather than the board. I reasoned th at if h e wanted to 'activate' the knight, rather than squash the passed pawn, then let him. Amazingly, it worked. After 46 f4 'It>C5 47 g3 'It>C4 48 g4, Bl ack blundered.

( B ) 4 9 hxg4, and White will soon h ave two passed pawn s, with enough coun­ terplay to equalize. Play suddenly fin ­ ished 4 9. . .'lt>c5 5 0 f 5 gxf5 51 gxf5 'It>d6 5 2 f6 'It>e6 5 3 b8� tt'lxb8 54 f7 'It>xf7 5 5 i.. x d 5+ Yz-Yz. There h ave been m any dreadful g ames in thi s collection, and I am not hiding losses, even when I did not h ave the physical strength to be able to find good m oves. My pl ay was dreadful that evening, and if he did not win it, then there is criticism of his own play. If thi s seem s ungenerous t o m y opponent, who achieved a draw again st a much higher-graded opponent, the point is th at he had the chance of beating an 1 M opponent on his pl ate, and he missed it.

Test 1 5 .4 N.Pert-C.Crouch

B u ry St Ed m u nds

Most reasonably sen sible moves win, the most thematic being 48 tt'lb8! (A), in effect a simple king and pawn ending. Astonishingly, he tried 48...fxg4??

2006

•..

We h ave just reached the time con ­ trol, and I had played dreadfully, as we

17 1

Why we L o s e a t C h e s s saw i n Test 11.4. My position i s , of course, losing . At the time, I noted that "maybe I did not quite h ave the strength to aim convincingly for the draw". To which we can add that in a weekend tournament, with five long games, sometimes even six, in two­ and-a-half days, even the young and physically stronger players can become tired, and make mistakes. I used to play m any weekenders, from the late 1970s onwards. Nowadays, it becomes more difficult. What i s interesting i s that for most of the finish, my opponent, a grand­ m aster and the younger player, actu­ ally played worse than me in the next part of the game. Thi s can be proved, and is not a matter of opinion. Just now, he has a clearly winning position, but at a l ater stage I had the chance of forcing a draw. The trouble is that I h ad only a couple of minutes before the flag-fall, and I missed the opportunity. Pl ay continued 3 7 :ta7?! (the imme­ diate 3 7 �xf5 gxf5 3 8 ,U,a7 is perh aps more accurate)

17 2

3 7 ...�g8 ( 3 7 .. .'i¥b1+ 3 8 'it'h 2 'it'g8 39 f3 ! should nevertheless eventually win for White; h ere 39 'iixf6?? '*Ifb8+ is Black's trap) 38 "i\Vxf5 gxf5 39 'it'h2 tLlh7 40 tLlf3 'iit g 7 41 tLle5 tLlg5 42 f3 :th8+ 43 'it'g1 :tf8 (back again).

Now 44 'it'f1?! started to move in the wrong direction. The king moves into open play, and Black h as ch ances of giving a check with the rook. In quickplay finishes, the subtleties of a delicate zugzwang play m ay be missed. Correct was 44 :tb7 ! , and if 44...:ta8, then 45 f4 tLle4 46 'u'xf7+ �g 8 47 :te7 .l:!.a1+ 48 'it'h 2 ! (which explain s why White avoids �f1) 4 8 ....l:!. e 1 49 :txe6 .l::!. x e3 50 d5, and White h as two extra pawn s, while his king is safe. Pert's choice gives Black much better chances of h olding. Then came 44...�g8 45 �e2 :tb8. Now White's king i s exposed, and Black h as drawing chances. He played 46 f4 to my surprise, giving Bl ack's knight an active square, but how else is White to make progress? Maybe starting again with 46 �f2.

Tes t Fift e e n I counterattacked, or at least aimed for a perpetual, with 46 tiJe4 47 I1xf7 I1b2+ 48 'it>el 11bl+ 49 'it>e2 11b2+ 50 �d3. •••

Now I missed a draw, by playing 50 11d2+? (A). There was a draw with 5o tiJf2+!! (B) 5 1 'it>C4 ( 5 1 'it>C3 tiJd1+ 52 'it>d3 tiJf2+) 51 ... tiJg4 52 �e7 tiJxe3+ 53 'it>C5 tiJxg 2 54 �xe6 tiJxf4, which i s quite a mes­ merizing knight jump with less th an ten minutes to play in the rest of the game. At the very worst, Black would be ...

...

able t o sacrifice the knight for White's fin al pawn, and Black can draw the rook and knight against rook endgame with care - except that in 1997 I could not quite h old thi s against Mickey Ad­ ams, as my flag was about to fall in a quickp1ay. So Black still has to be care­ ful, and who knows, I could h ave lost the g ame anyway, with little time left. After my inferior move, Pert played 51 'it>e4 tiJd6+ 52 'it>e3 I1xg2 (or 52 ... tiJe4+ 53 'it>b4 �xg 2 54 �e7 �b2+ 55 �a5 I1b3 5 6 .l:!.xe6 �xe3 5 7 �e8+ 'it>g7 5 8 d5 with broadly similar types of position) 53 l1e7 tiJe4+ 54 'it>e4 11e2+ 5 5 �b5 ':e3, and then there was a time scramble. I cannot remember the finish in detail, but a plausible finish might h ave been something like 5 6 �xe6 .l:!.xe3 57 'it>c6 �h 3 58 �e8+ 'it>g 7 59 l1e7+ �g 8 60 'it>d5 I1h4 61 tiJg6 �g4 6 2 'it> e 5 I1xg 6 6 3 'it>xf5, and White win s, which he certainly did in the game: 1-0.

173

Why we L o s e a t C h e s s

Yo u r M ove

N o attempt i s given to suggest wh at sort of percentage scores should be expected for pl ayers of particular grades or ratings. This would be slightly artificial . By defi­ nition, every move I made was, in the end, a mistake, with a 0% score. I am sure th at any reader would h ave made a better percentage, even a reader who has not played chess before. By sticking a pin on th e board, the pl ayer cannot stick far be­ low 30%, which my own score seem s by contrast puny. The reader, aiming to improve his or her game, should try to think about th e exerci se games, and should try to analyse and learn . A reader who scores more th an 70%, con siderably in advance of the random result, is doing well, but there are possibilities of improvement on even this. After all , even a 70% score would suggest th at on each positi on, there i s a 30% chance of a mi stake on these critical positions, and a score of 30% error would be a worry for most strong players. How to Audit Your Resu lts

Players want to improve their grading scores. An extra point per ten games over the year would mean 10 extra points in term s of the English grading system (this is very easy to calculate ! ) , or approxim ately 7 2 extra points in Elo term s. All this would mean a substanti al increase in grading term s. I h ave, of course, calculated my results of the 200617 season, as analysed in this book, and we shall turn to thi s later. In th e meantime, I h ave gone through the games pl ayed in the few month s since I wrote thi s book, to see whether I h ave im­ proved my accuracy. Between May 2009 and September 2009, I pl ayed 18 tourn ament or match

174

Yo u r M o ve g ames, but not in international events. I wanted to concentrate on this book ! There were 18 games, m ainly but not entirely against con siderably lower rated opposition. My results were: +12 4 -2, which m akes 78%. The scores were inflated by six easy win s in the British Braille Champion ship. My seventh opponent, Chris Ross, was a much stronger opponent, and drew again st me. The interesting question i s whether I could h ave scored con siderably more if I h ad avoided mistakes. Clearly if I h ad avoided my two losses, I would h ave scored at least an extra point. In fact, in one of these games I could h ave won with accu­ rate play. Of the four draws, I could h ave found a win against Graham Morrison . What should h ave h appened then would h ave been : +14 4 -0, which m akes 89%. Thus just avoiding three slips could h ave created an 11% increase in per­ formance. Maybe this is the one part of the g ame, the avoidance of errors, th at m akes the biggest difference in improving one's perform ance. My challenge for the season is to h alve this 11% slippag e. It is perh aps important to recognize that you should avoid double-counting. If you m ake a mistake, and reach what should h ave been a losing position, but later won, then congratulations, but you are lucky. Wh at is more important, in term s of results, is to remember the games you h ave lost, or only drawn after h aving h ad a winning position. =

=

Using the Com puter

In thi s book, I h ave used the computer in analysing my own games, and the rec­ omm endation for the reader is, of course, to take advantage of the computer. Do not overestimate th e computer though, but use it as a tool . I was relatively l ate in making use of the computer, partly being suspicious of computer assessments. I h ave been suspicious of reading computer-based as­ sessments which h ave clearly been wrong, and h ave preferred my own an alysis, with board and pieces. The computer i s important in generating ideas quickly, but even h ere th e hum an reader should treat everything with suspicion. It is still important to take advantage of making brain -power in analysis. After all, in a proper g ame, the player is not allowed to use the computer, and the hu­ man needs to practice in thinking about the game. And Final ly ...

Except, of course, chess is not final . Even after the last melting of the pol ar icecaps,

175

Why we L o s e a t C h e s s there will still b e chess being played, and new Noah s will still b e enthralled by the game. It is, of course, quite possible th at theory has gone so deeply that almost everything has been studied in depth, and that new version s of chess, maybe based on western chess, or eastern chess, could be tried, m aybe a fusion of rules with ideas from western and eastern chess. Thi s would be a long way ahead. For the next h alf-millennium, we still h ave the current chess of Greco and Ruy Lopez. For individual pl ayers though, any chesspl ayer is mortal . A pl ayer might h ave learnt the game, become fascinated, learnt so m any ideas of chess strategy and tactics, maybe becoming a strong pl ayer, m aybe hoping to become an even stronger player. Then perh aps before the final push, at the peak of the pl ayer's strength, ch ess ability starts to decline. It is now reasonably well established that for a professional player, the peak strength will be around the mid-thirties, and will start to decline slowly. For me, I really only seriously started studying chess at the time th at theoretically I should h ave been starting to decline. I h ave played a lot of chess, but did not h ave time to study h ard. Perh aps I can still add to the g ame, but not through my playing strength . I h ave continued to study the g ame, but can contribute more to the theory of the game, rather than through pl aying over-the-board chess. As the old saying goes, I h ave forgotten more about chess th an most players h ave even learnt. A S u m m ary of Games, in Chronological Order

Note: See the list of exercises at the end of the introduction for all the test g ames. This list is used partly as a research tool , to h elp to indicate when and why bad moves h ave been played. There being rel atively few games in thi s book, all pl ayed by myself, that it h ardly seem s worthwhile to add a second alph abetical list. All games were pl ayed in England. September 2006 Buckley-Crouch, 0-1 (win) October 2006 Hebden -Crouch, 1-0 (loss) Crouch -Cox, '/2-V2 (draw) Gregory-Crouch, 1-0 (loss) Crouch -Hutchin son, 1-0 (win) Pert-Crouch, 1-0 (loss)

176

Yo u r M o ve November 2006 Nurmoh amed-Crouch, 0-1 (win) Morris-Crouch, 1-0 (loss) Randall-Crouch, 1-0 (loss) Crouch -G ait, 0-1 (loss) December 2006 Sen-Crouch, 1-0 (loss) McKenn a-Crouch, %-% (draw) Crouch-Oryakh al, 1-0 (win) January 2007 Crouch-Radovanovic, %-V2 (draw) Crouch-Okike, 1-0 (win) Crouch- Roberson, V2-V2 (draw) [see al so Crouch -Povah, V2-V2, November 2006, given in the notes] Lauterbach -Crouch, V2-% (draw) February 2007 Crouch- Peacock, V2-V2 (draw) Cutmore-Crouch, 0-1 (win) Crouch - Rose, 0-1 (loss) Crouch -Lewi s 0-1 (loss) March 2007 Crouch -Gran at, 0-1 (loss) Wall-Crouch, 0-1 (win) Sowray-Crouch, 1-0 (loss) A substanti al m ajority of the remaining games in this period were wins, al ­ though it i s possible th at a small number of games may have been misl aid, a prob­ lem of being partially sighted. 24 g ames h ave been analysed in this collection, of which 7 were wins, 6 were draws, and 11 were losses, a score of 45%. Once you h ave gone through your games, the next stage, if you want to im­ prove your playing strength, i s to go through your statistics. Altogether in the main part of the season up to late March, I h ave recorded 5 6 games. A few of the games might h ave dropped out of the system, sometimes because with poor eye-

177

Why we L o s e a t C h e s s sight, I could not read the scoresheet. Also, I should perh aps h ave added the final game of the season, a London League loss against 1 M John-Paul Wall ace, when I should probably have been better. I beat him the next time around, to level the scores, sealing a match win against the formidable Wood Green team, previously unbeaten for close to ten years. I have no doubt that in the 5 6 games I h ave recorded, there will be unrecorded mistakes, often slight inaccuracies. In 24 of my g ames, I found identifi able in accu­ racies. As an approximation, I won, or occasionally played a solid draw, in h alf my games, and I h ave no reason to an alyse these in much detail. The other h alf is of much more interest. These are the g ames I need to focus on to cut out mistakes, and to increase my percentage scores. So: 5 6 games (100%) were played, of which 3 2 g ames (5 7%) seem to be 'clean ', including two win s against I Ms, and indeed a win against a G M . Many o f m y opponents were o f con siderably weaker strength, and therefore did not force me to h ave to play under pressure, so it would be unrealistic to claim th at I necessarily played well. The rem aining 43% (24 games) is of more interest. Thi s included: 7 win s ( 1 2 . 5 %), 6 draws (10.7%), and 11 losses (19.6%). Clearly at the very minimum, I would h ave gained 10.7 percentage points by avoiding losses, if I had turned losses into draws. The theoretical maximum of turning bad moves into win s would by 3 1. 1 percentage points, but this, of course, would be unachievable, in th at we cannot claim th at my opponent would h ave made mistakes if I h ad pl ayed the best moves. So we need to examine, move by move, wh at the most likely result would h ave been h ad I avoided the mistakes, and found the best m oves. Starting with the Losses

The obvious starting point is cutting down the losses. If there are pl ayers of roughly equal strength, but not players of top grandmaster strength (where draws are more likely as there are few mistakes), a typical score might be +40% = 20% 40%. As an illustration, let us assume that a player improves, and can cut out a quarter of his losses. Thi s might turn out next season with a result of +50% =20% 30%, an end percentage score of 60% as opposed to 50%. For simplicity, we call thi s an increase of 10 English (ECF) grading points. It is, of course, unlikely that a pl ayer will turn all his or her losses into wins. A m ore likely assumption will be th at h alf the reduction of losses will turn into draws, and the other h alf will turn into wins. The previous draws would, after a

178

Yo u r M o ve year of improvement, turn into wins. 10 grading points (an approximate equivalent of 7 2 Elo points) is quite a big jump. Even h ere, the pl ayer is far from cutting out losses, and thi s is far from per­ fect play. It is, however, progress. A fast improving junior, absorbing the ideas of chess rapidly, may often show an increase of 2 5 English grading points, with a score of +65% = 20% - 1 5 %, and perh aps the implication that it would be time to play stronger and more testing opposition . For more established players, it i s difficult to gain 2 5 grading points a year, or even a decade, or ever. All you can do is to chip away at the margins. If your mind is befuddled when trying to calculate complications, then sadly you will not be able to calculate ten moves deep, and, of course, there are other physical limita­ tion s. Therefore one can not h ave unusually high expectations. If you can cut out even a tenth of your losses, thi s is an improvement, and maybe if you start to feel encouraged, you can try again next tim e around. Thi s is perhaps a more pessimis­ tic, but certainly realistic, view th an in the m ain argument of the book. After all, one of the m ain thrusts of the argument i s that if the reader can learn to cut out a few basic and not so esoteric weaknesses, you can improve your chess substan ­ tially. Just cut out a few silly blunders, and there can be a substantial gain. It i s time to con sider the perspective of the Elo rating system, rather than the English rating system , a ladder system rather than one of finding th e average. Two good and useful rating system s, but with highly different perspectives. In the Elo system, we start off with the rating of the two players, and calculate in an individ­ ual game or tourn ament how m any points may be gained or lost in a particular game. If two pl ayers m ay be of roughly similar strength, then we can imagine th at one of the pl ayers might gain 20 Elo points over 20 games without gross blunders, through better understanding of the game. This is a slow progress. Under the cur­ rent system of the Elo international rating system, a player can suddenly drop in a move from a gain of 10 points to a loss of 10 Elo points, a total drop of 20 points. The D u m best Moves

Maybe it is time to consider the worst moves, and kick them out of the way quickly. I should perh aps note th at in the previous few years, the effects of brain dam age h as h ad rel atively little damage to my understanding of the game, but a much greater impact in my speed of thought, both in term s of speed of calcula­ tion, and lack of clarity of eyesight. Al so dizziness and tiredness have tended at times to make it difficult for m e to think clearly, and I have at times been unable to concentrate on pl aying anything other than the most superficial an alysi s.

17 9

Why we L o s e a t C h e s s There have i n fact been much worse periods of pl ay in chess terms, particularly at times when I have been actively working on writing books. A compl aint for many other chess writers, but probably worse for me, since my physical ability over the board has led to a deterioration in play. Just after my stroke, I gloomily wrote that I h ad probably gone down to 1 7 5 (about 2000 Elo) in playing strength, but fortun ately I h ave consistently reached 200 ECF, and stayed well over 2 3 50 in Elo terms. My target ambition is to get back to 2400, but any loss will tend to go down the greasy l adder, so I need to plan exactly which tournaments to play. The general conclusion is that I am still quite likely to m ake serious and unex­ pected mistakes, and my hope i s that I can cut down the blunders to a rating I would be h appier to reach . Time now to look through the most h orrendous blunders; the single moves that with even the smallest of thought I could h ave recovered my score. The Very Worst Games

1. Crouch-McKenna Drawn . I should h ave won easily, but I displ ayed a lack of concentration, then m ade several bad moves, and I should h ave lost. Half a point lost. When I think of all the bad games in th at year, thi s is the one which I remem­ ber the most readily. I feel confident I could h ave won it easily with better health . 2. Randall-Crouch A loss, which should h ave been a win, with a strong attack for me. I h ad the chance of taking a rook, but I worked out that if I h ad taken this, he h ad th e chance of pushing two m ajor pieces to a winning back row check. In fact there was only one check, and I would h ave h ad everything covered. To m ake it worse, I h ad the sim­ ple opportunity of taking a rook, after h e too h ad m ade mistakes, but I rejected it, not noticing that after he won a rook in reply, I could then take the rook with check. A full point lost. Again, with my mind functioning properly, I should h ave won easily. 3. Nurmohamed-Crouch By common consent, the venue at thi s club was not the most attractive place to pl ay. We were playing in a hut, in the winter, with in adequate heating. I played the opening embarrassingly badly, losing concentration after quickly g aining an edge, and before long I h ad a losing position, but recovered to some extent, and even later won after my king somehow wriggled out of an attack.

180

Yo u r M o ve 4. Buckley-Crouch Thi s was one of those nightmare games where both players exchange mistakes, neither player i s able to find a knockout, and eventually my opponent found an auto-blunder, when even I could find an easy way to win a pawn . Even worse, h e ran out o f t i m e after a lot o f thought. It was a dreadful g ame, although i f o n e of the players h ad played even very slightly worse, the opponent would h ave won quickly. No points lostfor me, as I won. 5. Wall-Crouch An awful game, with serious mistakes on both sides. My king move, when I at­ tempted to bring it towards safety, was quite simply bizarre, bringing it into the open . My opponent l ater made a serious miscal culation. One-move Shockers

In each of these cases, I feel I pl ayed well, or reasonably, in the opening, then lost concentration, pl aying a superficial move quickly, not calculating as far as I should h ave done. I h ave played sh arply with good effect, but suddenly when the game became critical, I played limply. 1. Sen-Crouch Good opening play, and with the chance to sn atch a pawn as Bl ack. Then, at a later stage of the opening, I needed to think of the one move for my queen to escape to a good square, admittedly difficult to find a few moves ahead, and instead pl ayed a 'simpler' line, which soon ended in collapse. I lost, and instead should have been better. 2. Hebden-Crouch I trusted him when h e offered a pawn, to keep the initiative. I shouldn 't h ave. It would h ave been better to grab the pawn, with the aim of breaking open the queenside. I lost, and with good play it should have been a reasonably comfortable dra w. 3. Crouch-Rose An experimental opening to avoid m ainline theory. The opening itself looks play­ able, if not especially dynamic, but I missed a big tacti c, which I should have seen in advance. I lost, although the opening was about equal.

18 1

Why we L o s e a t C h e s s 4 . Crouch-Lewis I wanted to prove th at the Crouch-Rose game was not really so bad, and so over­ night I aimed to find improvements. Maybe the opening was not so bad in analyti­ cal term s, but I was under the psychological pressure of proving that my idea was good, and lurking behind is the thought that if anything h ad gone wrong again, I would have been extremely anxious about my pl ay. Thi s was the wrong way to start a game. I thought I was doing reasonably well in the opening, but a ' Petrosian ex­ change sacrifice' knocked me back. Analysis showed that, slightly beyond my game horizon, I could h ave satisfactorily found a draw after taking the exch ange, but the position was scary, and I tried to pl ay around the sacrifice, rather than accept it. f lost, when f had probably a clear dra w. Four losses, then, when by pl aying better moves on four critical moves, my ex­ pected score would h ave been 2'1214. Over- p laying the Opening

The last four games indicate a lack of confidence, and I lost heavily. Quite often, though, there is a tendency in my game for over-confidence in the opening, and there are a few examples to be con sidered. Of course, the best way is to find the correct level, but it is extremely difficult to find it. Of 'my 60 forgettable mistakes' in thi s book, a quarter are uncovered on move 16 or earlier. The next quarter move us only to just beyond move 20. The opening is the most difficult part of the game, because the vast possibilities of chess still remain, and on every m ove the player has to assess not only what sort of position is likely to arise (attack, positional play, complications, or simplifications), but al so wh at possibilities need to be gradually eliminated. It is only several moves later th at it becomes clearer as we move past th e opening. In the next few g am es, I pressed the position much too far, aiming for attack, but with no real chance of finding a way to safety. It is do or die, but before too long I would be highly relieved if I could escape to a draw. When I played at Kidlington early in 2007, I clearly decided that I wanted to play imaginative chess, with the confidence that spring was about to come. There were too many bad games as a result. 1. Crouch-Peacock I played far too imaginatively, and I cannot reali stically claim th at, even after some m odest improvements, I h ad anything more than the advantage of the first move.

18 2

Yo u r M o ve My play went out of control, and I was fortunate th at my opponent went for a perpetual, when he coul d h ave gone for a winning endgame. A lucky draw. 2. Crouch-Gait Without thinking in great length, I developed my bishop with gain of play, by threatening a big check with the queen and bishop. H ad I thought for longer, I woul d h ave appreciated that h e could defend the threat with counterplay, and that a move later I could have been much worse. H e missed it, but my equilibrium was disturbed, and I m ade mistakes l ater. A loss, but I was slightly better out of the opening. 3. Crouch-Hutchinson I could not claim much of an edge, and it should h ave been very close to equal as White. I found a complicated way of pl aying for an advantage, but this was an il­ lusion, and with accurate tactical play by the opponent, I would h ave been much worse. He missed it, and ended up losing. A win, but I overpressed. I should have lost. 4. Crouch-Radovanovic I was still out of touch with my opening theory, and I improvised. My opponent missed a well -known opportunity to give Black a slight edge, and then later I m ade a bad knight versus bishop exch ange on g6, allowing Bl ack a semi-open file lead­ ing towards h 2 . I h ad to work hard for a draw, although there were improvements for him. Afortunate dra w. Sim ple Tactical Slips in the Middlegame

Everything seem s to be going smoothly. I am heading for, I h ope, a winning edge, then suddenly a miscalcul ation, or more probably, a slip in my thought process, wh en I assume something, but tactically it doesn 't work. 1. Gregory-Crouch Thi s seemed smooth enough during the game, but while I won a pawn, it was dif­ ficult to convert it into a serious edge, as his pieces were active. Becoming short of time, I immediately sacrificed the exchange, to eliminate a dangerous bishop, but I did not have the compen sation th at I was expecting. I should in stead have qui­ etly moved the king away. A loss, rather tha n a dra w.

183

Why we L o s e a t C h e s s 2. Crouch-Roberson I was able to keep a stable positional edge, with a passed d-pawn against a Be­ noni. I was then lazy in my calculation s. A dra w, rather tha n a win. 3. Crouch-Okike I had to give up a queen for assorted play, but I did it the wrong way round, and he could have improved. I still won, but I could have played better. 4. Crouch-Cut more A defen sive mistake, rather th an a tactical mistake. I was better, in a sh arp posi­ tion, but I soon m ade the big mi stake of playing too slowly in a sharp position of attack and counterattack. Before too long, I was clearly losing, but then he h an­ dled the attack badly, perhaps surpri sed that he was suddenly in a winning posi­ tion again st an 1M. He allowed m e a sudden checkm ate. I won, but I very m uch de­ served to lose. Getting Gro u nd Down

It h appen s, and players hate it. With very best play, one cannot, of course, lose a g ame from the beginning, but sometimes even the slightest mistake, alm ost in­ visible, will decide the game. This in fact very rarely h appen s except at the highest level - in my previous book, I am thinking, for example, of Kramnik-Leko, Dort­ mund 2006. U sually what will h appen, if the players play well but not totally accu­ rately, is th at there will be a series of minor slips, the defender gradually slipping away, or the attacker losing his advantage. Th e examples I give tend in the end not to h ave been rel atively subtle errors, and there is therefore the chance of being able to learn more from the errors. 1. Lauterbach-Crouch Even the less experienced pl ayer would quickly h ave seen th at my knight was bad on h 8 . After a few slips, I should h ave lost, but she avoided the best m ove, and I was able to hold for a draw. Of course, I did not particularly want to move the knight to such an uncomfortable square, but my mistake came earlier. To kick out White's bishop on g 5, I should h ave hit the bishop with .. .f6, rather than ... h 6 . I needed my pawn on h 7 . A fortunate dra w. .

2. Morris-Crouch I again m anaged to find one of the knights in the corner, this time on a8 rather th an on h 8 . Thi s was a result of placing my pawn s carelessly, creating weaknesses

184

Yo u r M o ve on the queen side. I wanted to advance my pawn to as, to prevent my opponent pushing forward with b4 to attack the knight on c5, but in the end my knight h ad to retreat to a8 much l ater. There was the strange incident of the fire alarm just before the time control, but even so, I deserved to be losing anyway. A loss, even though I seem to have equalized earlier on. Messing up the Endgame

There are surprisingly few examples of losing in the endgame, or throwing away a win . My biggest fear after returning to playing chess was that I would be so tired, and my eyes dizzy, that I would completely lose the thread of things in the end­ g ame. Thi s in fact has not h appened so often, m ainly because if I was not feeling in good health that day, I would probably h ave gone wrong much earlier, in the opening or the middlegame. Nevertheless, in earlier years I often m an aged to lose concentration in the endgame in the fifth hour, m aking ridiculous errors from good endgames, and then losing. Such disasters stick in the mind, but, of course, often it was my opponent who m ade the silly losses, and usually I h ave forgotten about those games. In general, I do not particul arly fear the endgame as such, but I do fear being short of time, and I do fear tiredness. I h ave already noted the poor play again st both Buckley and McKenna. In both cases I ought to h ave been holding the draw, rather than under the illusion of try­ ing to squeeze an endgame advantage, but both my opponents m anaged to play even worse. Also: 1. Pert-Crouch I got tired in round 5 in a critical weekender, and as I drifted from a late middle­ g am e into an ending my position became worse and worse. My opponent did not play particularly accurately, and to my regret I managed to miss a chance of a draw. A loss, but I could have recovered half a point. 2. Crouch-Granat A difficult queenless middlegame and then endgame, where I started with the bishops and h e started with the knights. Neither pl ayer h andled thi s with com ­ plete positional confidence, but in th e end I h andled things worse th an he did. A loss, but with good chances perhaps of a slight edge at various stages.

185

Why we L o s e a t C h e s s Two Difficult Games to I nterpret

Sometimes it is difficult to understand wh at is going on during a game, and al so difficult to understand at the post-mortem, or afterwards with the computer. If the game is played at relatively quick time limits, there are almost certain to be mistakes, quite possibly on both sides. The good news, for both players and spec­ tators, is that at least the games are interesting. In the two London League games I am covering, I m an aged a draw and a loss, but I could easily h ave ended up with two wins, or two losses. In my earlier analy­ sis, I must h ave questioned over half a dozen of my moves, but l ater I h ave decided that most of my moves seem to h ave been okay, and only a relatively small num­ ber eventually seemed questionable. So both players played reasonable chess in difficult positions. 1. Sowray-Crouch I probably h ad a slight edge as Black in a Sicilian, but I wasted a move with ... bS, normal enough, but h ere a waste of time. It i s difficult to believe that Black was worse, and even significantly worse, but almost everything I tried in my l ater analysi s ended up as bad. The Sicilian is so often on a knife-edge for both players. Eventually I found th at the 'natural' check with gain of tempo (with ... .ltf8-a3+ fol­ lowed by castling) was a serious mistake, and that the 'unn atural ' .. .f6, defen ding the e-pawn, appears to equalize. At least a dra w, probably more, from the opening, but I lost my way later. 2. Crouch-Cox Even more complicated, and a draw, but with opportunities for a win for either side. He h ad the more direct winning ch ances just before the time control . I would h ave needed to h ave found much more complicated options, and I very much sus­ pect that I would not h ave found the best option s even if (the standard excuse) my health was better, with a clearer mind. The difficulties for both sides were based very much on the question of finding the right balance between positional advan­ tage and m aterial advantage, with both sides at times h aving to con sider sacrifice and counter-sacrifice. A complicated dra w. And a Little C uriosity

Crouch-Jamshit My concentration l apsed in an unorthodox opening, and I was already worse after my fifth m ove. Before too long, I recovered my edge again. A win.

186

Yo u r M o ve Assessments

My calculation s suggest th at I dropped ten points as a result of serious mistakes, m ainly through dropped h alf-points, rather than losing from a winning position. This necessarily involves some degree of approximation, as it i s difficult to estab­ lish early on what the result would h ave been if I h ad pl ayed the game more accu­ rately. I h ave tried to err to the side of the draw. In term s of statistics, the most meaningful assessment in practical term s would be to say th at with accurate pl ay, I could h ave g ained 10 points out of 5 2 games, purely through active pl ay, and avoiding serious mistakes. Thi s is a score of 19.2% th at I h ave lost. If I h ad been able to play more accurately, my ECF grading would h ave jumped by 19 points. It is difficult to m ake a reliable Elo calcul ation because of the various methods by which Elo ratin g s are calculated, depending on the K-factor and the number of g ames (as well as the average rating) being played. We are, however, dealing with a ball-park figure of around 140 extra Elo points. N aturally no player can cut out their weaknesses in stantly, but even h alf an elimination of clear weaknesses would, on this basis, give about 10 extra ECF points. Some of the errors in this book m ay well seem extremely obscure and complicated to many readers, while others might be reg arded as simple mistakes, which most reasonably stron g pl ayers should be able to avoid much of the time. The basic question for the reader is to think whether you are capable of avoid­ ing at least some of the errors given in the exerci ses, and if so, whether you will be able to improve in your games. Good luck. I am going through my new g ames to see whether I can improve on my play. Maybe at some future date we can compare and contrast our efforts? The computer these days allows the player to indicate where possible im­ provements are to be found. Go through the lines and work out what is going on. Don 't bother so much about trying to learn opening theory. Just learn, paying equal attention to the different parts of the g ame: opening, middlegame and endgame; attack and defence; strategy and tactics. And above all, make fewer mis­ takes.

187