EXCELLING CHESS CALCULATION CAPITALISING ON TACTICAL CHANCES JACOB AAGAARD EVERYMAN CHESSEXCELLING CHESS CALCULATION JACOB AAGAARD BEES Gloucester Publishers plc www.everymanchess.comFirst published in 2004 by Gloucester Publishers ple (formerly Everyman Publishers plo), Northburgh House, 10 Northburgh Street, London ECTV OAT Copyright © 2004 Jacob Aagaard. The right of Jacob Aagaard to be identified as the author of this work has been as- serted in accordance with the Copyrights, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN 1 85744 360 8 Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 06437-0480. All other sales enquiries should be directed to Everyman Chess, Northburgh House, 10 Northburgh Street, London EC1V 0AT tel: 020 7253 7887 fax: 020 7490 3708 email: [email protected] website: www.everymanchess.com Everyman is the registered trade mark of Random House Inc. and is used in this work under license from Random House Inc. EVERYMAN CHESS SERIES (formerly Cadogan Chess) Chief advisor: Garry Kasparov Commissioning editor: Byron Jacobs ‘Typeset and edited by First Rank Publishing, Brighton. Cover design by Horatio Monteverde. Production by Navigator Guides. Printed and bound in the US by Versa Press.CONTENTS wR wWN on 10 Bibliography Introduction Before you can think, you need to learn how to see Candidate Moves When is the right time to Calculate? Important Thinking Techniques Visualisation and Stepping Stones When it is time to Calculate Creativity and Combinational Vision How to Train Calculation Exercises Solutions 10 18 30 49 73 93 104 134 152BIBLIOGRAPHY Books Secrets of Grandmaster Chess, John Nunn (Batsford 1997) Think like a Grandmaster, Alexander Kotov (New and improved version Batsford 1994) Improve Your Chess Now!, Jonathan Tisdall (Everyman Chess 1997) The Inner Game of Chess, Andrew Soltis (McKay 1995) School of Chess Excellence 2~ Tactical Play, Mark Dvoretsky (Edition Olms, 2002) Attack and Defence, Mark Dvoretsky & Artur Yusupov (Batsford 1998) Endgame Challenge, John Nunn (Gambit 2002) Art of Attack in Chess, Vladimir Vukovic (New improved version, Everyman 1998) Dvoretsky Endgame Manual, Mark Dvoretsky (Russell 2003) Learn Chess Tactics, John Nunn (Gambit 2003) Creative Chess, Amatzia Avni (Cadogan 1997) Other Mega Database 2004 Endgame Study Database 2000, Harold van der Heijden (editor) Chess InformantINTRODUCTION These days a lot of chess books are published on what I like to call ‘general themes’. In most of them the author starts out by claiming that there exist virtually no books on... How to think in chess; endgame strategy; this or that * positional theme; basically whatever... (see most recently in Paata Gaptindashvil’s fabu- lous Imagination in Chess, Batsford 2004). Of course this is no longer quite true (though it might have been when the writers were read- ing chess books themselves). I will not try to claim here, for instance, that there are no books on calculation. There are in fact several and they are all worth reading; I mention them| in the bibliography and more than once else- where. If, after reading this book with the TV. turned on, you cannot remember the titles, then perhaps you should turn off the TV and focus a bit more on chess! In this book J discuss the subject of how to improve one’s ability at chess tactics and calculation. To me, this involves opening the mind to certain ways of thinking, and then training. For that reason T have added a tacti- cal test to this book (to my publisher's regret, as this made the book longer than expected and delayed its delivery until right before the planned publication date). I hope that this test will encourage players to work on improving their tactical ability. At the same time as writing this book I was also working, together with FM Esben Lund, ona CD for ChetsBase, which will probably be titled Train Your Decision Making In Chess. If the current book inspires you to do more work on calculation and decision making, then the CD would be a good place to continue. Calculation is one facet of chess ability about which most people have an opinion. Amateurs, for instance, often think that grandmasters calculate many moves ahead. Grandmasters do have this ability, but it is not the primary difference between them and the average club player. Rather, there is a degree of accuracy, visualisation and organisation in a grandmaster’s calculation, which together with his overall greater knowledge of chess makes him superior to the amateur. One aspect of this is greater imagination, to which I have devoted quite a bit of space in this book. Hete is one of my favourite moves of all time. No amateur (without some kind of training), I think, could ever have decided upon this move in a tournament game. Krajina-Kozul Vinkovei 1989 Sicilian Defences 1 e4 cB 2 Af3 dé 3 d4 cxd4 4 Axd4Excelling at Chess Calculation 216 5 Dc3 Ac6 6 £4 g6 7 Axc6 bxc6 8 e5 Dg8 9 WE3 d5 10 fe3 hS 11 h3 £5 12 0-0-0 e6 13 2a6 With his last move White threatened both, g2-e4 and £b7. 13...8a3!! What makes this move fantastic is that there is no immediate threat. Black simply undermines White’s queenside defence. 14 bxa3 Stronger perhaps was 14 We2 Wa5 15 Db5! with unclear play. 14...Wa5 15 g4 hxg4 16 hxg4 Exh1 17 Exh1 2xg4 18 Wxg4 Wxc3 19 &g1 19 Wh3! would have kept White alive. 19...Wxa3+ 20 Sd1 Wxa6 21 Eh8 0-0-0 22 Wh7 d4 23 Ext? We4 24 Bxa7 vbs 25 Ea3 f\e7 26 Wh3 2d5 0-1 Imagination (or vision) is not something people always associate with calculation. But the most important ability in chess is not seeing ten moves ahead, but seeing the posi- tion which is right in front of you. This kind of awareness has increased in the last ten years with the emergence of strong chess playing programs, whose short-term vision is their greatest advantage over humans — be- cause they are set up to examine everything. Thus they show us our small tactical mistakes and encourage us to improve out calculating, abilities. The following is one of my favourite examples: Resika-E.Lund Budapest 2002 Zee Ose ae A an wee ae Ws a a. 7 z 2 White to play and win. 1 Gs! A nice tactic that wins the exchange as 1..Axd6 2 2x6 &xg6 3 Bett @h7 4 Eh4 is mate. Black could resign, but instead tried: 1...Ef6 2 Hg1+? This is sufficient to win, but the computer comes up with 2 Hp4H which wins a lot more than a rook. 2...82h7 3 Exf6 Axd6 4 Bfgé Exg6 5 Axg6+ Gh6 6 £3 AS 7 217 Ad6 8 &g8 45 9 ve2 DAh4 10 2F7 AG 11 Bgs+ @h7 12 gi Ddé 13 £98+ &he 14 Eh1+ dg7 15 Bh7+ kxg8 16 Bxd7 Abs 17 &d2 Oxa3 18 Kxa7 Db1+ 19 ve3 @xc3 20 He7 b5 21 Bxe6 b4 22 Bb6 Da2 23 &d2 wt7 24 Hab b3 6Introduction 25 cxb3?? 25 Ha7H! wins at once and 25 c3 would also win in the long run. Now it goes wrong, 25...c3+! What White had imagined would happen here is hard to guess, but this game was played with FIDE-time (where both players get an extra 30 seconds after each move) and White still had about a minute left, so should have smelt the trap. Possibly the stress of prolonged time trouble was getting to him. 26 &d3?? 26 Be3 c2 27 BoG cl 28 Excl Axct 29 b4 would still draw since the knight cannot escape the attentions of the white king, eg. 29,..22, (or 29...Nb3 30 d3) 30 bS ke7 (30...c3?? 31 b6 wins) 31 #d2 ete. 26...Db4+ 27 &xc3 Axa 28 b4 2\c7 29 &d3 &f6 30 %e3 HFS 0-1 But really, how much deep 10-move calculation is needed to play good chess? It used to be said about Petrosian that he did not calculate more than one move ahead, but that he saw everything that was worth seeing, Even in grandmaster games it is limited how much deep calculation is needed. The follow- ing queen sacrifice probably did not take a lot of investigation from Nijboer. Nijboer-Acs Wijk aan Zee 2003 a a) “2 L oo ioe FG SO nwre White to play and win. 22 Wxe4!? Wxed 23 2b5+ Wicd Now White obtains a better endgame. If instead 23.987 24 AgS+ wins, or 23...Hc6 24 Dd4! F7 (24...0-0 25 Bxc6 d5 26 AS wins) 25 &xc6 We5 26 Bfel Wh5 27 213! with a strong, probably winning attack. 24 Sxc6+ Exc6 25 Aa S17 26 Dds Bg8 27 g3 Hg4 28 b3 axb3 29 cxb3 2d8 30 Hf Uxfa 31 Axa 2b6 32 Ads Rc5 33 Hl de6 34 Axfé d5 35 Axh7 Re7 36 g4 d4 37 g5 d3 38 g6 d2 39 g7 Bet+ 40 Bxct dxctW+ 41 coxct wf7 42 D8 kxg7 43 Ad7 ket7 44 dec2 Hes 45 Dbo &c5 46 Dad 2f2 47 a3 bad 48 Db2 1-0 All very convincing. But in fact White could have done better at the beginning! If he had played 22 &bSH thef Black would not have been able to reach an endgame in which he had some chances of survival. After something like 22...8.c6 23 Bxc6 Hxe6 24 Hd4! Black is completely outplayed and it is not easy to see how he should improve on this. The method which could have helped Nijboer to find this is changing the move order of the combination, as explained on page 61. This is one side of the coin. We need to learn these tactical schemes so that we do not end up on the wrong side of such examples. But there is also another element to calcula- tion: sometimes it is vety difficult and re- quires hard work at the board to figure out the right way to proceed. The following is a flawed study from one of our time’s greatest study composers, the Israeli IM Yochanan Afek, which could very well have arisen in a normal over the board game. (cee following diagram) White is in serious trouble in this endgame and only saves it through a miracle. The fol- lowing analysis was given in the Study Data- base 2000 and probably comes from Afek.Excelling at Chess Calculation Y.Afek Chess in Israel 1999 vy. Aomoe & OY ve _ ome o oe nee anata. a” es White to play and draw 1 b6! Exa7+ Tf 1...8.d8? 2 Dc5+ Hc8 3 Ld7 mate, or 1..R£2 2 Bdt Bxb6+ 3 Axb6 Eaa7+ 4 Kb5 £25 QGH (5 Bc2? Ha2 6 263+ Haz 5..82c7 6 DdS+ Gd6 7 Dc3 and Black can- not win. 2 bxa7 &e1+ 3 Sb f2 4 aBll+! dxas 5 wae! ‘To allow a promotion with check is always attractive. Instead 5 #6? Sa5+! and Black wins. 5...f1H+ 6 2b5 WES “98.2 Vfl lg yyy Ml yyyy Mlle, BO iy bay flr ‘Uy yy bY) Yi, WU oa a! 7 £c6+! Wxc6+ 8 Ab6+ Lb8 stalemate All of this is extremely nice, but there is a small problem in one of the lines: After 2 &d1 &xb6+ 3 Axb6 Black can play 3...f2!, changing the move order, and after 4 e2 (no better is 4 263+ Bxa7 5 Lp? Bos 6 2h3 By5+ and Black wins first the knighe and then the bishop) 4...8xa7+ 5 d2b5 Ha2 6 23+ @a7ll 7 2g? Bb2+ and Black wins. One of the points of the study was meant to be that White could force the black king away from the knight, and thereby be in time to stop the fpawn. After 2 di this is no longer the case. So what would we do if we reached this position in a game? Well, 1 b6 is the only move that docs not lose very quickly. For instance 1 Ac5+ @2xa7 2 b6+ Sb8+ 3 Dabt &b7 4 &b5 Bel+ and Black wins without a fight is easy to calculate. With 1 b6 we would at least have a little counterplay. So after 1 b6 £2! we would again stand at a crosstoads. Re onl 2 a te “a faa ae pee alt i Here we would try to set up our candidates. Besides 2 &d1 we also have 2 &c2 and 2 &b3, all of which send the bishop in the direction of the fl-square. From our calcula- tion of 2 Sd1 (which is the most forcing move and therefore the first one we should analyse) we discovered that the undefended bishop became a problem. Therefore it is better to keep the bishop closet to the king. OF course this is an abstract conclusion which would be impossible to reach without calculation. In such a complex position as this experience will only help so fat, while 8introduction general rules and guidelines will not help not at all. What we need to do is to analyse the moves carefully. Actually, when we get to the bottom of it, there are not that many varia- tions: a) 2 Rd1 &xb6+ 3 Axb6 al) 3..Exa7+ 4 Yb5 £25 QBH (5 Ler? Ha2 6 263+ sha7!) 5.7 6 DAds+ Bdo 7 4e3 and Black cannot win. a2) 3...f2! 4 Re2 (or 4 263+ Bxa7 5 Be? Hp8 6 Bh3 By5t and wins) 4..Bxa7+ 5 &b5 Ha2 6 S£3+ Sa7!! 7 Bg2 Bb2+ and Black wins. b) 2 Bc2 Rxb6+ 3 Axb6 bl) 3..Bxa7t+ 4 b5 £2 (or 4..a2 5 Rett Ba7 6 Ax) 5 Bett Bc7 6 Dd5+ Bd6 7 Dc3 Bed 8 Lp? By7 9 Vfl Bgt 10 Bc5!! Bed 11 cA! and it appears that Black cannot win. 2) 3..f21 4 Bett Sxa7 5 Ld3 Has! 6 Me2 Be8 and Black wins. ©) 2 2b3! cl) 2..&xb6+ 3 Axb6 Hxa7+ (this time 3.2 4 B04 Hxa7+ 5 GbS is a little differ- ent!) 4 @b5 25 Bc4 Bc7 6 Ads+ Ld 7 De3 Hb7+ 8 Hab Hb1 9 Ld3 Hel 10 Af1 Bc6 11 Bad cS 12 Bad Sd4 13 LbS Bb1 14 Ap3 ke3 15 Atl+ cf4 and it does not appear possible for Black to make progress. C2) 2...€c6 is an extra option, but after 3 Dbs+ Sc5 4 DaG+ eb 5 LASH (we do not even need to see this since we are happy with a draw) 5...€2xd5 6 b7 and White even wins. (Thanks to John Shaw for his help with analysing this study). OF course it is difficult to calculate all this, and T imagine that an average grandmaster between 2500 and 2600 would take 10-15 minutes to make the right choice, and occa- sionally fail. But is it impossible to learn how to calculate it all? No. First of all, let us look at what really needs to be seen: The difference between 3.2 and 3..xa7+ is important. In two out of three lines everything becomes clear very quickly, so they can be disregarded altogether as inva- lid. Just as easily we would not be able to refute 2 &b3! which, in a game, would alone be enough to inspire us to play it, leaving the accurate calculation to the opponent. As you will learn in this book, one of the main points of calculation is that we should only calculate when we have a choice — although strong players know that we often have more possibilities than we see at first glance. In this book I have tried to give a good overview, to lay out some useful tools, and to explain how we can train to improve our cal- culation. For those frightened by analysis, this book will be a true horror story. But it is im- possible to study tactics without once in a while venturing into analysis, to understand what we are really dealing with. For my other books I have generally refrained from long variations and tried to explain everything with words. But here it was the variations them- selves that needed to be explained. Also, for those interested in working with this book seriously, it will be an advantage to be able to follow the author’s reasoning more closely. As always I would like to use a little space to thank those who have helped me in writing this book: Danny Kristiansen for his assis- tance and proof reading; Mark Dvoretsky for allowing me to use from his lectures four ex- amples which I felt suited my agenda well; John Shaw and Esben Lund for occasional advice; Byron Jacobs for postponing the dead- line; and Anne Faith James for moral support. Finally, T would like to apologise to the reader for using my own games to such a large extent, I have no illusions as tegards their quality. It is merely easier for me to target specific themes when I know what the player was actually thinking during the game. I hope that the lack of chess beauty presented in them will not ruin the overall impression of the book. Jacob Aagaard, Glasgow, July 2004. 9gCHAPTER ONE Before you can think, you need to learn how to see The most famous book ever written on cal- culation is Alexander Kotov’s legendary and much discussed Think Like a Grandmaster, originally published in Russia in 1970 and subsequently translated into several lan- guages. ‘This book covers a lot of important topics such aslhafidliig 6 fimelEoUblglhow to assign thinking time in calculation, and attention to general concerns: bishops pro: tecting or restricted by their own pawns, and so on, But more than anything the book is famous for its first third dedicated to the calculation of complicated variations. Other later important books on this sub- ject are Mark Dvoretsky’s School of Chess E: cellence 2 — Tactical Play, Jonathan Tisdall’s Improve Your Chess Now, and Dvoretsky and Yusupov’s collection of essays from a train- ing seminar for very talented youngsters in Russia 1992: Aitack and Defence. All the au- thors are in an inescapable dialogue with Kotov, never fully agreeing with the old mas- ter, but always deeply respectful. In the cur- rent book I will use Kotov’s terminology too, as well a5 wying to show the differences be- tween the v: ~ for the practical player — deeply important issue. But here I will also give attention to dif- ferent phases of analytical technique that the ious viewpoints on this previous books have not sufficiently covered. The problem for most chess players wanting to improve their results is that books are aimed at a level above which they currently perform; there are certain abilities they need to acquire before they can benefit fully from complex calculation tools such as the method. Work with eager pupils in all age-groups shows that some simple abilities are not natural and need to be developed, but which can be trained with success all the way up to the very best players in the world. At a seminar in Denmark in 2002 Artur Yusupov told how he once, while walking around the tournament hall, came to stop at of compatison or the tree of analys the board where Alexander Morozevich was playing. Morozevich was deeply concentrat- ing on the position, thinking and thinking, Yusupov did not understand why Moro- zevich was thinking, as the position had one completely obvious move and no real alter- natives. Assuming that Morozevich, rated in the world’s top ten, knew what he was doing, Yusupov started to look for alternatives, and found that Morozevich actually did have a very interesting alternative. Morozevich kept on thinking and Yusupov returned to his board and made his draw. He went to watch Morozevich again and found he had made a third and equally interesting move! 10Before you can think, you need to fearn how to see This little story — and others like it — tells us that even top grandmasters do not auto- matically see all valid possibilities on the first move! However, €xperience|shows that this ability can be developed and continuously trained — and indeed should be, as it is probably the most important technique the tournament player can acquire and master. Kotov (and most of his followers and crit- ics) begin with lining up the candidate maes. This is an important aspect of analysis, but it cannot be done vety well without some kind of training in how to find the candidates! There is more about this in the next chapter, bur first take a look at the following position and, as a quick test of your ability to see, try to select three possible continuations for White. Nohr-M.Nielsen ‘Taastrup 2002 » a nome Baten ate a White to play. In this position White was soon lost after 19 2g? exd6 20 De6?, since Black, had he been trained in looking for candidates, could have used the unprotected rook on the first rank to eliminate the unpleasant knight by the odd-looking 20...We8!, pinning the knight and preparing ...2xe6, thus securing the king and his material advantage. ‘The more natural 20...We6l, forcing an endgame with two minor pieces against a rook, should also have won quite easily. Instead of 19 Ag5?, White has two equally interesting opportunities, both leading to suf- ficient compensation for the piece. But before we turn to the concrete options, let us linger a second to consider the abstract objectives in the position. White has sacrificed a piece in order to attack. He needs either to regain his material, or equivalent material, or deliver mate in order to justify his actions. There is no doubt that he has the initiative, but in order to take advantage he needs to bring all his pieces into play quickly, to gain a majority ar the scene of action, in otdet to fulfil his objectives. Less abstractly, this means that he needs to get his finger out and get the queenside into play! This is what we would expect a grandmaster would most likely try to achieve. The first of the two relevant candidates is the following: 19 dxe7!? Wxe7 20 &g5! This is the most logical and simple way of mobilising the queenside: £6 is under attack and White develops the bishop with tempo. Soon the rook will enter the action and White will start a direct assault on the black Kingside, before Black is able to develop his own pieces. Quick analysis shows that Black should probably respond with: 20...Hib4! After 20... Wd7? 21 &.£6!! Black cannot de- fend himself. a) 21.88 22 Wc3 Ah6 23 We5 Eps 24 17Excelling at Chess Calculation Bad] &xf6 25 Axf6 and White wins. b) 21.65 22 We3 &h6 23 Ado Wh5 24 Asxé7 Exf7 25 Hest 28 26 Hxf8+ SxfB 27 Wb4+ and Black is mated. 20..We7? 21 Bact Wa7 22 266 is similar. It is not easy to imagine a line where the rook is worse on 1 than it was on al. Now, if he has nothing better, White has 21 2d2 We7 22 &g5 with repetition, though 22 Dp5!? looks far more tempting now than it did earlier. Sometimes it is enough to know you have a definite draw in order to go for a line, especially if that line includes a substantial sacrifice. All this is not so surprising. That White went for 19 Ag5 in the game should be ac- credited to a lack of attacking technique (Finn Nohr is mainly a positional player) rather than an inability to calculate variations. However, some unimpressed youngsters and the equally unimpressed Fritz 7 found 19 Snel! SSS, pris “a aang a ne eran AS ~ Black to play. 19 &AGH might be a surprise, but what now? ‘The main idea behind 19 &h6! is not so hard to spot. If Black plays 19...&@xh6 then 20 dxe7 and White wins a lot of material due to 20..Wxe7 21 DFG+ and Black has to give up his queen for insufficient compensation. But Black of course needs to respond. He cannot play 19...cxd6 because of 20 DrGr, nor can he play 19.845 (an attempt to free the queenside and relieve the pressure on the king by exchanging queens) since after 20 dxe7 Wxf3 21 exf8W+ 2xf8 22 exf3 &xh6 23 Ad6 the pin costs Black a piece and he ends up the exchange down. But what is White really threatening? Nothing in particular; his next move would be 20 Bad1, after which he does start to have a lot of threats, but mainly 19 &h6! is about domination. Black cannot get out of this with 19...c6 since after 20 Sxe7 Sxg7 21 Weer be8 (21..2h6 22 He3 or 22 ApS illustrates the need to sce little other than candidates) 22 Wd4 White has a very strong, ptobably decisive attack. Black's only move is to force White to ex- change on g7 without relinquishing his weak, but still existing control of £6. 19...8h8! is therefore the only defensive option that gives Black any hope of surviving. White should now proceed with 20 &xg7+ sxg7 21 Had1! bringing the rook into play as de- scribed in the abstract objectives above. As yet we have not managed to find a defence here for Black, and most likely there isn’t one. One attempt starts 21...Wd8 but then 22 We3+ eS (Black has no choice here as after 22...h6 23 dxe7 Wxe7 24 Ado! he is under heavy artillery fire, unable to find cover anywhere) White has time and active pieces enough for launch a deadly attack on the dark squares, weakened by the exchange of the bishops. Best is 23 f4! ‘yy Mig wea ne t Wi Py Yay x, 12Before you can think, you need to learn how to see and Black seems to be quite lost. The vatiations go something like this: Exdit. cashing in) and White has a win- ning attack without having invested any seri- ous material in the process. So the conclusion to this exercise is that White wins after 19 &h6! — truly a stunning move that neither of the players considered during the game, nor any of the spectators. “knowledge of attacking technique to find, but i i It is = ably no surptise to the reader that it was Frit, 7 which originally discovered this possi- bility. Strong chess-playing programs have transformed the way we think by showing us corridors between our logic and our eyes that previously were dark and hazy, if not com- pletely closed to us. Though a computer makes a lousy chess teacher it can be a useful tool in the hands of the serious chess player and chess trainer, by pointing out different alleyways and possibilities. But more about that later. Let’s return to the issue of secing through the eyes of a human. When Kotov writes that we should select a limited number of candi- dates to calculate, he actually skips the most important part of the process, Finding moves, is by no means easy. What young players normally try to do when they follow Kotov’s advice is look quickly at a few possibilities and then try to calculate them one by one, as directed by Kotov. Again and again they then flip to the solutions (or these days compare their thoughts with those of a computer) and find that they are way out in the haystack, and that if the needle is anywhere to be found, it is sitting in their own behind. Often they simply do not find the best first moves, and instead of accepting this as a lack of technique, they get frustrated and lose belief in themselves. Unsurptisingly they quickly lose interest in a type of exercise which is very difficult and gives them nothing but failure. This can be compared to people starting out in jogging, unaware of the time it takes for the body to build muscles or simply to get warmed up. Many are the cases of people stopping again, assuming that the pain they get in their calves is due to a problem with jogging or with their legs, and not simply jogging technique. Because, how difficult can it be to run a mile or two? Well, it takes about 8-12 minutes for the body to warm up and if you start too quickly, then you will inevitable suffer and have a disappointing experience. The same goes for training calculation. Searching for candidate moves is not an automatic process, The mind usually does not work logically, but intuitively. Searching the board for possible moves is neither automatic, not something that will happen automatically once you know you need to do it. You need to train at it. Some people claim that they end up play- ing their first idea 75% of the time or even more. Does this mean that we actually see our best move immediately in 75% of the cases? No. This phenomenon can be ex- plained by sor ce i ed factors involved in the decision process. First OF all) we Have a natu- ee quent thought processes ee often become an attempt to prove the correctness of our intuitive choice, ot else we simply play the ~ Sas roy ty Ba 13Excelling at Chess Calculation in a position, we are not likely to find alterna- tives to out first thought; while if we do not select candidates at all and simply skip the search function, we are not likely to have many alternatives. Furthermore, the 75% does not tell how many times this first thought actually is the best move in the posi- tion, nor about how many of these positions might be deemed ctitical. Obvious moves do exist (even though we might try to keep the above Yusupov example in mind) and often the obvious move is the best move, whether we tty to find superior alternatives or not. How much of the 75% remains if we take all these restraints into consideration? Probably not a number significant enough to support the claim that we should value the first move we see that highly. Finally, it is very possible that the majority of mistakes committed by players who have problems selecting candi- date moves are to be found in the remaining 25% anyway. It is obvious that some players are intui- tive, while others are very concrete and logi- cal in their approach to chess. And though this indicates completely different thought processes and different bases for decision making, both types of player will benefit greatly from learning the selection of candi- date moves. For logical playets there is great need for a sound basis of candidate move selection in all kinds of positions (based on logic or gen- eral concerns), because the moves that just come to them can hardly be trusted in great numbers. The very reason for defining them as analytical players is due to their relative weakness in intuitive decision making and their reliance on a concrete analytical ap- proach. They need to really look for the op- portunities in order to know what to analyse, and they need to analyse in order to know what to play. Perhaps more surprisingly, there is great need for intuitive players to learn to select candidates as well. To be an intuitive player does not mean that you necessarily see more moves. Rather it means that you have a good feeling as to which direction the position going, that you know where the pieces be- long and, most importantly, that you have automatically make good valuations of the different positions reached in your calcula- tion, The problem with this way of thinking, despite its obvious advantages, is that if the solution is not in the direction your intuition tells you it is, you will probably not be able to solve the position at all. The important thing to be aware of about secing is that) we continuously change the way we look at a chess position. If you try to return to the position above, you will find it impossible to look at the position with the same eyes as you did before solving it. Before you read the solution, 19 &h6 may not have been within your horizon, while after seeing the solution, it takes a great effort to look at the position and not see 19 &h6. This is because the way we see the world is struc- tured not only by light and colours, but also by our brain which organises the information imported through the visual cortex. Chess positions will have no meaning to someone who does not know the rules, while they might seem overpowering with all their possibilities for the beginner. But, in time, the number of possibilities are reduced to a limited number, until one day you may feel like Yusupov, that there is an obvious move in the position and nothing else. ‘This is the natural progression for a chess player, but it is not the end of the line at all. A learning process has usually four phases: ‘The first phase is called unconscious incom- petence, the second conscious incompetence, the third onscious competence, and the fourth phase is called unconscious compe tence. Related to chess vision this can be translated into the following four phases. © Chess pieces are pieces of wood. No dinovledgeoftheralesatally 14Before you can think, you need to learn how to see Knowledge of the rules, but with a _great amount of possibilities that can ‘not truly be evaluated: This i This is the situation of the beginner. All legal moves are possible candidates. A greater knowledge of chess and of the evaluation of moves. The amount of possibilities becomes greatly re- duced. ‘This is the level of the club Chess positions immediately make sense and obvious moves appear in the subject’s head. This is the situation of the very strong chess player with a fully developed intuition, If we were unaware of the Yusupov story above, we might be led to believe that the fourth phase was an attainable goal. To look at a chess position and immediately read it in a way that makes absolutely sense is of course an advantage, but it will also be a pre- text for doing nothing, something that makes you feel there are no other possibilities in the position, This actually comes pretty early on in a chess player’s progress. In his entertain- ing and rather philosophical book, Creative Chess, Amatzia Avni divides the development of the chess player’s imagination into three stages: © Everything may be right; everything is possible, There are no principles, no guiding laws. Everything that falls inside the frames. is possible; the rest is wrong. Everything that falls inside the frames. is possible; the boundaries between the frames and their surroundings are not tigid. So what is outside the frames can also be possible. Or to put it another way: Any move is possible. The number of possibilities is highly limited and no other possibilities are considered. Few moves are considered, but all op- tions again become open. In chess, realising that we cannot and do not see everything immediately is highly im- portant for the development of our thinking abilities. Before we can say ‘select candidates’, we need to talk about the ability to find moves. This is a kind of chess vision and is a sacticelagy that will prove very helpful in _ _ vision. A much more common type is solv- ing simple combinations. Some people claim that you should not solve too easy combina tions, which is true — as long as we add the word ‘exclusively’. Exercises that can be solved simply by finding the right combina- tional motif are good for improving your chess vision. You can never do enough. An. other way is by solving difficult studies. The following study is a good example of the possibilities in chess, and how we need to open our eyes to them. M.Liburkin 64 magazine 1939 Zz all ae “a ame Y gnats oe me ie ae yy White to play and win, 15Excelling at Chess Calculation ‘The prime component of this study is the ‘threat’ of being left with two knights vs. nothing, and thus being unable to force your will upon the opponent. That the black pawn can be stopped is easy to see, but both knights are able to catch up with the pawn, and only one of them wins. Before we give the solution, let us have a look at another great study, which is dedi- cated solely to finding the tight move, and with not so much to calculate. J.O’Donovan British Chess Magazine 1939 "s_ =“ a a White to play and win, The solution to the first study should in- clude seeing Black’s resources as well. 1 ATS Using the other knight, eg. by 1 De6 g2 2 Dyg5 fails to 2...a4+1 3 a3 (Gf 3 Wc3 a3 ete.) 3..g1W 4 D+ Sd 5 Axgl Lxc2 with a book draw. 1...92 2 Bh4 giW 2.14 is of course also possible, but it will not make a draw in the long run, so we stop there after a few general observations, such as that the knight cannot immediately catch the c-pawn and force the undesired endgame. 38+ edt! ‘This is again the idea. After 3..d2f1 4 Axgl @xg1 the c-pawn cruises down to the eighth rank. 4xgi ad+! is a s oe aH as Pe ates ae An important thing to remember when one calculates is that the opponent is playing according to his plans, and not according to yours. If you tried to solve the study, this was the main thing you should have spotted, and what can be done about it. 5 &c3! Retaining the pawn is non-negotiable. 5...a3 6 Dd5 a2 7 De3+ sect 7.681 8 Sb2 and wins. 8 Se2+ sb1 9 ded2! Setting up the cage where Black will be smotheted to death. 9...a1H 10 Dc3+ Sb2 11 Ded mate. A very aesthetically pleasing study, and one that requires that you find some nice moves in order to solve it. The second study is even mote imagina- tive. White cannot win after the tempting zugawang operation 1 Hxg8t+ Axg8 2 WE? as Black has the stalemate defence 2... WfB!!, So White needs to find another way to proceed. Thete is hardly any methodology that can be used for solving this study that you could take with you into a practical game. The only thing is open eyes to additional opportuni- ties, ie. the conscious incompetence as a chess player that gives you the right to look at all possibilities afresh. Here that might assist you to find 16Before you can think, you need to learn how to see 1 Wxg8+! @xg8 2 Bf7! aoa Y D Y ae fate “ae “ and White wins, as the stalemate trick is gone, eg. 2.266 3 exfo Wxc5 4 Hh7+ des 5 £7+ Sf8 6 Bh8t 4y7 7 Byst+ Bho 8 wwe and so on. One problem with study solving is that you cannot help but wonder why there are pawns on c5 and c6, since for a study to be aesthetically sublime, all pieces on the board need to have a function. However, although training is training and should not be a competition, and although simulations of real playing situations are de- sirable at times, the main function of training is to isolate particular abilities and improve them through specifically designed exercises. Nevertheless, for most players studies are difficult, and one of the important aspects of training is to build confidence and motiva- tion which is not achieved by trying to solve far too difficult exetcises. Instead, you should solve exercises of medium difficulty. In this way there is a balance between pleasure (for training should be pleasurable or you will stop doing it before you reap any rewards). and the development of particular abilities. By solving simple combinations your success rate should be high and the benefit in terms of developing chess vision should also be quite high. Let me add a final word in this chapter concerning methodology: “tical exercises whose main idea is simply to find moves rather than to calculate accu- rately, you will lean naturally to look for moves. This is an important ability to take with you to the board, where you will be able to find additional possibilities again and again. In this book we shall return to this ability repeatedly, at various stages in the calculation process. The important thing to remember is that moves ate not all visible by themselves, even after lengthy training in chess vision. Many moves will have to be found. Solving exercises with this theme, and a general awareness of the things we over- look, will help you to pay more attention to this phenomenon. 17CHAPTER TWO Candidate Moves We have already briefly touched upon this subject in the previous chapter and now we will go deeper into the aspects of the selec- tion of candidate moves, Kotow was the first to introduce the terminology of candidate moves, although his book is famous more for his ‘tree of analysis’; and while many au- thors have opposed the ‘tree of analysis’ method (for a frontal assault please read Tis- dall’s interesting book Improve Your Chess Now), they have generally acknowledged the importance of candidate moves (except per- haps for Tisdall, who at first agrees with Dvoretsky that candidates need to be se- lected, though not necessarily immediately, but then a few pages later expresses deep reservations about the whole concept). In this book we shall talk about candidates, which covers both candidate mores and can- didate ideas. Kotov wrote that he wanted to teach chess players to think like machines, but it can easily be stated that we do not teally think in moves like a machine, so much as we think in ideas. Moves arc just moves to us, but ideas ate pockets of meaning that makes sense. The following example illus- trates this. Karpov-Serper Dortmund 1993 ara? fH we fata a8 gM % 4 i at Eat tell a fa White to play. Many candidates can be selected for White in this position, but the move 23 @f5 would not make sense unless you see the idea be- hind the knight sactifice. Most experienced players would spot the idea of sacrificing the knight first and only afterwards find the rea- soning behind it. Abstract ideas like over- loading and the weakness of the back rank would not matter, but a quick scan over all possible meaningful moves would pick up on the knight sactifice immediately, and the consequences of it would soon become ap- parent — and once these consequences are clear, it is hard to imagine that White should play any other move, so probably Karpov 18| Candidate Moves played it rather quickly and focused his en- ergy on the resulting endgame. 23 Af5! Black cannot allow the knight to come to dG, so the following endgame cannot be avoided, 23...exf5 24 Wxe7 Exe7 25 &xb7 Hts 26 Lab! x UY WSs y ie \jo WS +8 zm a ‘The tactics ate now over and White has a clear advantage. With his last move White keeps control over the c-file and the promot- ing square for his c-pawn. So now, with a strong bishop vs. knight, with more active rooks and an extra pawn on the queenside, White had no problems converting his ad- vantage into a full point. 26...g6 27 c5 bxc5 28 bxcS Hc7 29 Ec He8 30 c6 &g7 31 Sf1 He5 32 fb7 Bad 33 Hed1 He5 34 Sd7 He7 35 Exc7 Xxc7 36 Hd8 Ke7 37 £3 g5 38 Ha8 94 39 fxg4 fxg4 40 Exa7 Ad5 41 Bat De3+ 42 gi of6 43 a+ 1-0 You could argue that 23 DFS is a candi- date move, but also, and just as important, that it is a part of a candidate idea. Of course you can also argue that the move makes no sense without the exchange of queens and the potential for a back rank mate, In that case can really talk about candidate ideas. The following is such an example. Breder-Aagaard Budapest 2002 What is White’s strongest continuation? Black has had an off-day and has ended up a pawn down with only a little counter- play. White should now try to find the best way to neutralise this play. To do so he will need to start selecting possible moves. 1 Wa8t forces 1...Wd8, after which White seems to have little choice but to exchange queens. This is certainly one candidate. (Note that 1 Wa8+ is connected to a few forced moves, and hence there is little point in talk- ing about a candidate move, but more a candi- date idea. For the sake of simplicity, I will use the expression candidates, since for the practi- cal player there is no difference.) The alternative to 1 Wa8+ is 1 BB+, an- other forcing move. After this Black has only ‘one legal move, 1...@7, but then 2 We6+ @#h6 does not seem to advance the white agenda. My opponent thought for some time and came up with the wrong solution. 1 Wa8+? ‘There are of course many ways to solve this position, but the simplest is to list candi- dates, In the game the critical position is that after (1..Wd8 2 Wxd8+ Hxd8) 3 23, when the opponent’s countet-chances should be taken into consideration. But first we should do our job of selecting candidates for calcula tion, as this often saves us time. 19Excelling at Chess Calculation ‘The second (and better) idea in the posi- tion is to play 1 Bet s2g7 2 Hbsl. This es- caped my opponent's attention completely. He only calculated at the tempting 2 Wet, but did not see any purpose to it. Had he taken the time to look for other possibilities 1 am sure he would have found 2 Hb8!. The deadly threat of 3 Hb7 is so terrible that Black will have to hurry into a rook endgame with 2..Wig4+ 3 Waxed Exgdt 4 B61. We now have two candidates which can be calculated, but actually we do not need to. It is obvious that White is worse off with a passive rook on £3 in the 1 Wa8t line, than with an active rook on b8 in the second line. ‘This kind of comparison can often save us time in evaluation and calculation. Whether White is winning or not, it is obvious that the second line is preferable, and for the practical player that is all the information required. But, in the game White played 1 Wa8+. 1...Wd8 2 Wxd8+ Bxd8 3 Hf3 b3! Tam not sure if White had overlooked this move. He probably did, which is just another reason why he should have focused on the starting position and found the solution to the problem. 4 axb3 cxb3 5 cxb3? 5 ¢3!? was worth considering, although af- ter 5..05 6 d4 a4 7 He3 Hc8 Black should have sufficient counterplay. 5...id4! 6 Ke3 df7 7 of1 Hb4 8 He2 Hxb3 9 Hd2 a5 10 be2 ad “» Yi, a mem mo wae eat ig a a oiere Lay Black has a very active rook vs. a passive rook. The extra pawn does not matter much, 11 te3 eG 12 ddd g5 13 hed h6 14 {3 Hb4+ 15 d4 Xb3 16 Hf2 cd6 17 Hc2 Hb8 18 h3 He8+ 19 wd3 BtB 20 de3 He8+ 21 dd3 HB 22 we3 He8+ 23 d12 HbB 24 we3 %-% The ability to select candidates is probably the most important for a chess player. Strat- egy, positional rules and concepts ate impor- tant too, but if you overlook simple tactics, you will not be able to use the others for much, To put it very simply: the things you overlook in your long calculation will seldom be as important as the move you overlooked at the start. S.Webb-Aagaard Sweden 2002 are e Ot gs 2 Fa Yay eas en a0 Play White against your computer from this position. This position arose between two disgust- ingly outdated IMs and authors in the Swed- ish League 2002 third round. The end was played in severe time trouble and the tactical possibilities never really surfaced before the game was put under the microscope of home analysis. Please note: I recommend that you try to play this position with White against a computer. The idea is not to calculate deeply, but to make good moves. White’s position is win- ning, but not without a lot of difficulty. The 20Candidate Moves moves required to win this game will not be found in long calculations, but in a lot of short variations and through making good moves on each and every move. Before we turn to the game there is some valuable advice I would like to present you with. The first is simple. Look, don’t think Whenever I think 1 am doing something really clever itis usually just plain stupid. —Esben Lund. One the greatest dangers in chess is that cour brain gets in the way. We might some- time have convinced ourselves that we are intelligent creatures, but really we should not fall for it. We are not that smart. Often the only thing our brain does is to cloud our eyes and make it impossible for us to see what is really going on in the position in front of us. "The following example is taken from the new version of John Nunn’s first collection of his own games, Secrets of Grandmaster Chess, in which the experienced grandmaster and writer presents some of his early games. Nunn-Najer Teesside 1974 Don't think, look! Here it is tempting to start calculating the different lines after 28 Wl Del+ 29 Afg2 Dxc2 30 Axc2 Bxc2 31 Bd4 Abeb 32 f4 and the attack is repelled, though some play still exists. Actually this might be what you would end up playing no matter whether you picked all the candidates or not. But Nunn is and always has been a very aggressive player and here he came up with a nice combination. 28 Ba ‘This first move should be spotted as a candidate move, but as we shall see on the next move, candidate ideas are just as impor- tant a concept. 28...b8 29 Bxf7! Nunn gives this an in his book, un- justly, as we shall see below. Here we are really talking about candidate ideas. White could either have come up with this highly original rook sactifice, or he could have fore- seen the opportunity to exploit the back rank with the following sequence: 29 Edat! Exc4 30 Axc4 and White is winning. Black cannot defend his back rank so he is simply a piece down in the resulting endgame after he loses his own queen. One line could go 30...d5 31 Asbo Wxb6 32 H1a6 and wins. It is not really fair to talk about 29 Exf7 and 29 Eidal as candidate moves. They both include certain ideas and just seeing the moves themselves proves little. If you do not spot the idea behind 29 Hdal, which is the simplest win in the position, then you have no chance of actually including it in your list of candidates. You simply see that 29...2ixe4 wins your queen and you move on. Actually, you probably would not see 29 Hdal at all, unless you are well trained in finding surprising candidate moves. 29...De5 ‘This seems forced, even though it is ob- jectively no better than alternatives, eg. 29..Hxc4 30 Hxg7+ Bh8 (if 30.28 31 Axeb+ Seb 32 Axc4 and White’s position is completely overwhelming — the threats are so many that it seems they can hardly be counted!) 31 2b7+ Hc3 32 Exb8+ Exb8 33 ®xe6 and White is a few pawns up. 30 Exg7+! 27Excelling at Chess Calculation White of course had this in mind. 30...2xg7 30..@h8 31 Exes! Lixcd 32 Axc4 and Black’s position is collapsing. ae me mee y “ge Don't think, fell — Brace Lee. It is easy to see the knight fork and then start to calculate the complex lines after 31 Dxe6 g8l. But if Nunn had stopped for a moment and actually just looked at the posi- tion, he would undoubtedly have spotted a far stronger continuation. 31 Axe6+? 31 Wexe6! is more ot less winning on the spot. There is no way for Black to defend his king against all the intruding white pieces. In the position White is threatening Df5+, Wed, Bxe5+ and Exd6. Against so many threats there can be no defence, as analysis of the position proves: 31..Be8 (or 31..Wo7+ 32 Dig? BA 33 Exd6 Bxdo 34 Wado+ We7 35 &xe5 and wins; actually it is a more a slaugh- ter house than a chess game here) 32 @.xe5+ (the most human approach; 32 WfS and 32 Wed also wins easily) 32..dxe5 33 Hd7+ and White wins, as after 33..@h8 there is a sim- ple combination with 34 AgGH hxg6 35 Wh3+ and mate follows. Why did Nunn overlook this both in the game and in his later analysis? In the game there could have been all kinds of reasons: time pressure, assumptions, bad form, or whatever. In the preparation of the book it is another matter. No computer program would miss this move for the world, and yet Nunn still does not include it in his book (and Nunn is very famous for his use of computers in his writing). The reason is that he did not stop to think here. Instead, the text move is so obvious that it is easy to play it without thinking, Ivis a check, a fork even. By comparison the queen is not doing any- thing directly on eG, there is no immediate threat to the black king or queen. But if you do stop to think for a moment, then you will these great threats and start to calculate. And before we can calculate, we must realise whether we have a choice or not. 31...2f67 ‘This is just reckless. Black needs to get his king into safety and this can only be done with 31..&g8!, when the king and queen might be on the same line, but White cannot exploit it, After 32 We2 @xc3 33 fre3 We7+ 34 e4 Wa7 Black is only slightly worse. The reason for Black’s mistake is likely to be time trouble, though Nunn does not mention anything thereof in his book. 32 Qxe5+! It seems that Black somehow missed this, but even so 32 Ac7!? is also a winning move. The attack is simply too strong. 32...dxe5 33 Ad5+ Gxe6 34 Wo4+ St7 35 Dxb6 Hd8 36 Wt5+ ve8 37 Web+ &e7 38 Hxd8+ Wxd8 39 Ad5 1-0 John Nunn is a great writer and has won the British Chess Federation’s ‘Book of the Year’ award many times, including both books of his best games (1987 and 1994). He is especially senowned for the high level of analysis he presents in his books. Neverthe- less I managed to spot another ~ and very similar — mistake in his analysis in his latest work Lear Chess Tactics (a good book for beginners on basic tactical ideas in chess). It should be said that I am probably the only one who will ever notice this mistake, and that for good reason... 22Candidate Moves H.Danielsen-Aagaard Taastrup 1999 In this position I played an obvious move, which is easy to find for an IM who has solved many combinations in his time. 29...d2! It always feels good to sacrifice the queen, and even better to under-promote a pawn as well. But the combination is so simple that it is really no combination at all. 30 d6+ @h8 Here White used a good deal of time be- fore realising that he had nothing better. 31 Wxc3 31 We7 Wa4t+ 32 dg3 West 33 2 We4+ 34 e2 Le4+ and mates is one line that Heini was forced to analyse. 31 Who diD+ 32 SF De3+ 33 der Wa2+ 34 es Wexp2H 35 dxe3 Wel+ was another possible line. All in all nothing looks alright for White. 31...d10+32 Ge2 Dxc3+ 33 Sd3 b4 34 bd4 g5 35 eS Ded+ 36 wb6 £5 37 2d5 Dxd6 38 c7 DeB+ 39 xd7 M6+ 40 we6 Axd5 41 Sxd5 Lg7 42 sec4 a5 43 a3 bxa3 44 wb3 14 45 c4 94 46 c5 sef7 0-1 Nunn describes this game in his own words but he appears to be of the belief (as I was for years) that White fell victim to a clas- sical combination. It is a little difficult to understand how he could make this mistake, but the explanation is probably that he looked at it without a computer program. I had thought about using the example myself in an article but when I rechecked the variations I realised that White had a major improvement. Go back and see if you can find the only defence instead of letting me do all the work! (Solution on page 26) Let us return to Webb-Aagaard (page 20). White should start with 28 Dxf7!! This is the only serious candidate. The method of elimination should make this clear, Webb played this with only five min- utes left on his clock, seeing lines like: a) 28...lWixe2 29 We5! and White will enter the endgame a full pawn up with no doubt about the result, e.g, 29..Wxe5 30 Axes Hs 31 Ad7 and Black is out of ideas. b) 28...x£7 29 Wxe6 which just wins; the tricky 29..WEAl? is best refuted by 30 Exf7 Wsf7 31 Wed5 with three extra pawns. a Gy is noe “ee a wreeer 23Excelling at Chess Calculation What is the only defence to this kind of aggression? 28...ae8! The chess equivalent to turning the other cheek. It is clearly the only move, as the al- ternatives all lose directly. For a practical player this is a nice move to find. Under pressure, losing another pawn and clearly close to complete failure, he finds a move that at least keeps the engine running for a while. Now White simply panicked and was not able to use his limited time to find a play- able continuation. But still, giving up the rook gives no chance at all of a playable posi- tion, and the practical chances are also close to nil. So it would have been better to choose one of the alternatives and then see if any- thing turned up. 29 Bxd5?? Wxd5+ From here on the game is over. 30 Wxd5 exd5 31 Ags Hxe2 32 Af3 Exa2 33 Hd7 Hb2 34 Exd5 Hxb3 35 4\g5 EaB 36 DeG Hb6 37 Ac7 HcB 38 @b5 a4 39 Ad4 BaB 40 Bd7 a3 41 AS a2 42 Uxg7+ &h8 0-1 '" i oat ge ms . a a ay How should White have continued? ‘This is a truly complicated situation and it takes some time to calculate all the possibili- ties, I have personally listed three candidates beyond the move Simon played in the game: 29 We5, 29 Hcc7 and 29 Wd7. The right thing, I believe, is to calculate them one at a time and come to conclusions as quickly as possible. Remember, we calculate not to indulge ourselves in variations, but to find the best move as quickly and as reliably as possible. Pseudo- thinking with concepts and ideas will bring you little satisfaction here. You need to be concrete and to find the relevant variations. Not too relevant is the piece sacrifice 29 Wed? Ext? 30 Hxf7 Sxf7 31 4 when it is hard to believe in the white attack, and for this reason alone it is possible to disregard it as a serious candidate. The combination also has an casy tactical refutation in 31...We2! 32 Wa4 f6 and Black is close to winning. More persistent is 29 Hec7!? Wxe2 30 Ah6+ (but not 30 Wes Wxed 31 Axed Axc7 32 Bxc7 Zic8 and Black is winning) 30..22h8 8377, ae yy "el ata na mes ee White to play! ‘This position could easily have been fore- seen as well. A closer look would have te- vealed that White is actually in trouble! If he plays 31 f7+? then 31...—xf7 wins on the spot. However, there is a nice resource in 31 Wxf8H! BxfB 32 AF7+ and White has per- petual check because of the vatious rook mates. Tt should not take too long to find the right first move. 29 Wa7! Now Black has to be careful, as after 29..e7 White will surely answer 30 Exd5l! Eixd7 31 Bdxd7 Wxe2 32 Ah6t hs. 24