sign in sign up
<<
Home , Blunder (chess), Bobby Fischer, Compensation (chess), Correspondence chess, Draw (chess), Eduardo Iturrizaga

The Complete Swindler

How to Save Points from Lost Positions

New In Chess 2020 Contents

Explanation of symbols ...... 6 Acknowledgements ...... 7 Introduction ...... 9

Part I What is a ? ...... 17 Chapter 1 When to enter ‘Swindle Mode’ ...... 30

Part II The Psychology of Swindles ...... 35 Chapter 2 Impatience ...... 38 Chapter 3 Hubris ...... 42 Chapter 4 Fear ...... 45 Chapter 5 Kontrollzwang ...... 50 Chapter 6 The Swindler’s Mind ...... 57 Chapter 7 Grit ...... 58 Chapter 8 Optimism ...... 62 Chapter 9 Training Your Mind ...... 68

Part III The Swindler’s Toolbox ...... 79 Chapter 10 Trojan Horse ...... 80 Chapter 11 Trap ...... 86 Chapter 12 Berserk Attack ...... 92 Chapter 13 Window-Ledging ...... 100 Chapter 14 Play the Player ...... 108

Part IV Core Skills ...... 121 Chapter 15 Endgames ...... 123 Chapter 16 Fortresses ...... 144 Chapter 17 ...... 160 Chapter 18 Perpetual ...... 168 Chapter 19 ...... 183 Chapter 20 Gamesmanship ...... 195

Part V Swindles in Practice ...... 207 Chapter 21 Master Swindles ...... 208 Chapter 22 Amateur Swindles ...... 254 Chapter 23 My Favourite Swindle ...... 273

5 The Complete Chess Swindler

Part VI Exercises ...... 277 Test 1 ...... 278 Test 2 ...... 289 Test 3 ...... 299 Solutions to exercises ...... 309

Epilogue ...... 349 Index of names ...... 355 Bibliography ...... 361

Explanation of symbols

The with its coordinates:

8 TsLdMlSt 7 jJjJjJjJ 6 ._._._._ 5 _._._._. 4 ._._._._ 䩲 White stands slightly better 3 _._._._. 䩱 Black stands slightly better 2 IiIiIiIi  White stands better 1 rNbQkBnR  Black stands better a b c d e f g h  White has a decisive advantage  Black has a decisive advantage q White to move  balanced position n Black to move ! good move ♔ !! excellent move ♕ ? bad move ♖ ?? ♗ !? interesting move ♘ ?! dubious move

6 Introduction

Chess is in the last resort a battle of wits, not an exercise in mathematics. Theory helps you; but you have to fight. Hence our contempt for the stupid word ‘swindle’ in chess. – C.J.S. Purdy

At the strong Tallinn tournament the board after the game’s end and of 1971, the game between the asked Smejkal, ‘What would you grandmasters Furman and Smejkal have done after 91.♘b3 ?’ ‘Why, was drawing to a close. White had to a2 of course!’ was the reply. already mentally given up and was ‘Ah,’ said the Magician from Riga, simply going through the motions with a twinkle in his eye. ‘Then to conclude proceedings: you would only have drawn.’ And with his trademark nonchalance, he Semyon Furman pushed a few pieces: Jan Smejkal 91.♘b3 a2?? 92.♘c1!! Tallinn 1971 ._._._._ _._._._. ._._._._ ._._._._ _._._._. _.l._._J ._._._._ ._._._._ _.l._._J _._._.m. ._._._._ J_._._._ j._._.m. _.n._._K ._.n._._ † _._._._K A fantastic drawing idea is revealed, no doubt to both players’ horror. Position after 90...a3 to a queen or rook is 91.♘e4+ ♔f3 92.♘xc5 a2 and stalemate, while getting a new Furman soon resigned. For most of bishop is no help with a rook’s pawn, the spectators, the finish was rather even without the white knight. As dull. Tal started to walk away, Smejkal For most spectators, that is, except waved him back, announcing ‘Aha! for one: . The former I can still win! Pawn to a1 equals World Champion was watching knight!’ Unphased, Tal reached back on. As the British writer Leonard and played Barden recounts, Tal approached 92...a1♘ 93.♘b3!!

9 The Complete Chess Swindler

._._._._ myself among them, the result _._._._. seemed clear on both players’ faces. ._._._._ The white player had his head in _.l._._J his hands, looking completely ._._._._ dejected as his final seconds ticked _N_._.m. down. From the sidelines, I spotted ._._._._ a cute swindling motif. Yi is a s._._._K pretty good tactician, and I briefly thought his expressions were all leaving the grandmasters aghast. part of some sort of theatrical A picturesque swindle that, alas, bluff he was pulling on his older never made it onto the scoresheets. opponent. Alas, the gestures were legitimate angst. As I was trying I’ll never be able to play like Tal. to evaluate a complicated queen But let me share with you the endgame, Yi let his clock run down closest I’ve come to imitating him to zero and resigned in the same as a spectator. Fast-forward almost motion. The spectators began to 50 years to 2017 and the game disperse. that inspired this book, at a less Allowing a respectful pause, as I esteemed tournament with less assumed Tal had done, I approached esteemed players and certainly a and asked Aleks what he would less esteemed kibitzer: have played if Yi had continued with Yi Liu 38.h7!! Aleksandar Wohl Aleks looked at me as if I had Gold Coast 2017 suggested the moon was made of cheese, and reached out his hand ._._.m.k for the b-pawn. But the Australian _._._J_. veteran had known me long enough ._._._.i to appreciate my swindling ways, _.t._.i. and he paused to take a second look .j._._._ at the position. _.j._._. ._._.m.k ._R_._._ _._._J_I † _._._._. ._._._._ _.t._.i. Wohl, a legendary Australian IM, .j._._._ had been steadily outplaying his _.j._._. young opponent in this endgame and had assumed complete control. ._R_._._ As a crowd of spectators watched, _._._._.

10 Introduction

‘Ahh!’, he exclaimed. It was then that both players realised the game ._T_.r._ was far from over, and that White _._._.kI has some serious self-stalemate ._._._J_ chances: 38...b3?? 39.♖xc3! ♖xc3 _._M_._. 40.g6! is immediately a , for ._._._._ example. Another pretty line is _Jj._._. 38...♖b5? 39.♖xc3 (anyway!) 39... ._._._._ bxc3 40.g6! c2 41.gxf7 and there _._._._. is still no way to avoid the draw. After a few minutes, Wohl cleverly With an endgame that we agreed suggested gave White excellent practical 38...♖c8!, chances. It is notoriously more which is the only try for Black to difficult to calculate how to escape keep winning chances. However, the checks than to give them in White’s not out of tricks yet: these sorts of queen endgames, 39.♖f2!1 especially in . In fact, with the help of tablebases, it turns ._T_.m.k out that White can hold the draw _._._J_I even against best play: ._._._._ 44...♖c7+ 45.♖f7 ♖xf7+ 46.♔xf7 c2 _._._.i. 47.h8♕ c1♕ 48.♕d8+! .j._._._ with, apparently, a theoretical draw. _.j._._. Perhaps Yi would have held the ._._.r._ draw from here, or perhaps not _._._._. – perhaps it would be lost ninety percent of the time. But this isn’t Now g5-g6 is an unstoppable threat. the point. The lesson here is that 39...b3 with a little more grit and a dash 39...c2 40.♖xc2 ♖xc2 41.g6! leads to of optimism, White could have a draw, while 39...♔e7+ 40.♔g7 b3 deployed a fiendishly cunning 41.♖xf7+ ♔e6 42.♖f6+! ♔d5 43.♖f8 defensive resource that, at worst, ♖c7+ 44.♖f7 ♖xf7+ 45.♔xf7 c2 would have made Black sweat 46.h8♕ c1♕ also splits the point. hard for the full point, and at best 40.g6 ♔e7+ 41.♔g7 fxg6 42.♖f7+! would have secured an immediate ♔e6 43.♖f6+! ♔d5 44.♖f8 draw. Ten percent is better than

1 At the time I suggested the inverted move-order 39.g6 ♔e7+ 40.♔g7 fxg6 41.♖e2+ ♔d6 42.♖f2, but the irrepressibly thorough Belgium FM Helmut Froeyman, who often comments on my blog, later pointed out that the ending after 42...g5!! 43.♖f8 ♖xf8 44.♔xf8 c2 45.h8♕ c1♕ leads to mate in 57 moves!

11 The Complete Chess Swindler zero, which is all you can get from chess. Make no mistake: computers resigning. have drastically improved chess training, and today’s players *** advance faster and are much more That evening, I wondered whether likely to reach their full potential today’s energetic young talents, thanks to technological tools. But even with – or perhaps because of – an unexpected side effect of these their use of computer engines and advancements is that we have vast online materials, are somehow forgotten the practical nature of less motivated to look for swindles the battle. We are obsessed with than players of former generations. engine evaluations, treating them After the wide research I conducted as gospel (which, incidentally, has for this book, I am surprised at how led to the rise of armchair critics – little attention has been paid to but that’s another story). And that swindles in training. influences our own games, in which This is especially strange given that we strive to always play the ‘best’ this is a part of our game where move as evaluated on the screen, improvement yields immediate to the exclusion of almost all other dividends. You may never reap factors. the benefits of, say, learning 20 I completely agree that a player moves of an obscure opening should try to play the best sideline or memorising the Philidor moves – but I disagree that ‘best’ manoeuvre in the notorious rook- always equals the engine’s first and-bishop versus rook endgame.2 choice. The computer evaluates On the other hand, we all get a position assuming that our into lost positions often, and thus opponent will play perfectly at stealing an extra half or even full every turn3; it doesn’t (and cannot) point every now and then will consider the myriad of important make a dramatic difference to both human factors in a contest, such your rating and your tournament as fatigue, time pressure, risk- performances. aversion, complacency, frustration, There are two reasons why the impulsiveness... the list goes on. majority of players ignore the The second reason for the neglect study of swindling. The first is, of swindle training is simpler: it’s surprisingly, the rise of computer hard! For a coach or a self-taught

2 Though, see the Endgames chapter. 3 This is true of engines that use alpha-beta pruning algorithms, which includes most engines such as Stockfish, etc. But this is not exactly true of the new generation of engines that use Monte Carlo tree search, such as AlphaZero, Leela Chess and its commercial clone Fat Fritz. In fact, these engines spit out evaluations in terms of percentages that are much more similar to how humans think about evaluating a position than the cold ‘0.00’ we are all used to seeing.

12 Introduction player, the amount of material to collect examples, I started to available for other parts of the appreciate why past authors have game, such as openings, endgames been wary to touch the subject. or tactics, is overwhelming. There are a few collections But where does one start training showcasing chess curiosities that how to swindle? You can’t search a have been meticulously assembled database for games with swindles, by noted enthusiasts such as and even if you could, what would Horowitz, Tim Krabbé, Ger van you learn? Perlo, Yochanan Afek and Amatzia I sympathise with coaches who Avni. I now appreciate what these want to help their students study authors must have gone through swindling and chess psychology in their research, and you will in general but simply can’t find find details of these superbly any relevant materials. Most of the entertaining resources in the advice I’ve heard or read has been bibliography. vague, ranging from optimistic Luckily, thanks to the fact that sentiments such as ‘hang in there, I am addicted to watching chess because blunders do happen’ to tournaments online4, and that the slightly more useful ‘train I meticulously file interesting your tactics so you are ready when examples on my computer5, my opportunities arise’. However, database of modern swindles was there’s a big difference between already quite large. But for this winning a game because your book, I wanted more, and so I opponent blundered all on his took a modern approach: crowd- own, and setting up a swindle that sourcing. On social media, blogs actively encourages the decisive and the like, I put out a call to blunder. But a student seeking to the chess community to send me learn the valuable art of swindling their best swindles. The response is liable to wind up disappointed; was overwhelming. Soon, my engines are clueless, databases inbox was flooded with stories of are useless, and there have been games featuring horrific blunders, virtually no good books devoted to outrageous gamesmanship, and the topic since the 1950s. even blatant cheating. This book has grown considerably The filtering task was much more since my initial concept of an laborious than I expected. Part of anthology of beautiful swindles. the problem, as Afek noted to me, As I embarked upon my quest is that there is not a consistent

4 Not a good personality trait for a researcher. 5 A significantly more useful personality trait for a researcher.

13 The Complete Chess Swindler definition of ‘swindle’ in chess Swindler needs to exploit these terminology.6 I painstakingly biases. Finally, we bring these filtered through the games until I parts together to focus on how to was left with over 800 examples of train yourself to become a better the purest gems, resulting in what Swindler. I am sure is the largest collection To that end, I’ve added some of chess swindles in the world. swindle-specific puzzles at the end Along the way, I discovered some of the book that are designed to miraculous escapes and mouth- help train your swindling skills. watering trickery. But more They are different from most significantly, I noticed familiar chess exercises in the sense that patterns among the games. I began computers won’t be any help to to realise that this book could not you; in fact, quite often the engine’s only redress the glaring lack of best move will not be the correct material in the chess literature answer, as it often doesn’t reflect a on pure swindles, but could also player’s best chance of setting up a be a useful instructional tool for swindle. (On the other hand, there would-be Swindlers. are some puzzles in which you will I don’t mean to trumpet this book try to avoid a swindle and win a as the godsend to fill this vacuum; won game, where, reassuringly, the I would be very happy if more computer typically agrees with the (and better!) books were written solutions.) on swindles in the future, as this And along these lines, you’ll find is a topic that fascinates me. But that my annotations in the main until then, I’ve done my best to games often don’t match the put together as informative and as computer’s evaluations. I might entertaining a guide as possible to award a brilliancy (‘!!’) for a move this wonderful world. that doesn’t even make the engine’s top choices but sets up a devious The general flow of the book is as swindle, or call a move a blunder follows. First, we look at the most (‘??’) when it technically forces common psychological biases but drastically narrows that affect chess players over the the margin for error.7 Proponents of board. Next, we discuss the most may shudder important attributes and skills the at such blasphemy in a chess book –

6 Jonathan Rogers, writing in Kingpin magazine, calls it ‘Perhaps the best example of an abused chess word’. 7 Speaking of grammar: I have capitalised ‘Black’ and ‘White’ when these words refer to the generic players, while the corresponding adjectives are not capitalised (as in ‘the black queen’). I generally stick to the present tense when annotating a game, except when it seems to make more sense to tell an anecdote or a story in the past tense, like in the above examples.

14 Introduction but here, too, I will demonstrate the anything instructional; there are benefits of learning to swindle by many new and unusual themes showing some remarkable rebounds discussed in these pages, and I find within their hallowed world. that it is much easier to absorb a A quick note on the structure. I key idea if it is self-contained in its have tried to follow some sort of own chapter. natural flow, as well as keeping Above all, I’ve tried not to forget congruence between the sections what got me interested in this – for example, each psychological project in the first place: swindles bias maps directly with a specific are fun! I hope the examples in this type of trap that the Swindler can book will inspire and entertain you employ to exploit it. The book as much as they’ve delighted me. is made up of several parts, with May you always win your winning each part containing several ‘mini- positions, and may Caissa smile on chapters’. Some of the chapters are you for the rest. quite short, containing a few or even just one example. This reflects GM David Smerdon, my own preference when reading Brisbane, January 2020

15 PART II The Psychology of Swindles

For the chess struggle nowadays, one needs a subtle knowledge of human nature, an understanding of the opponent’s psychology. –

Psychology is one of most debated yet least understood parts of our game. famously scoffed, ‘I don’t believe in psychology. I believe in good moves.’ Yet, despite this being one of his most quoted sayings, Fischer spoke often about psychology in chess, stating that ‘the object is to crush the opponent’s mind’. Judit Polgar has argued that chess is 30 to 40 per cent psychology; but then again, Richard Teichmann quipped that chess is 99 per cent tactics – so somebody’s wrong. Top players often talk about psychological factors in their writings and interviews; and are prominent examples. The classical masters were even more willing to share their thoughts on the subject. In the following game between two of the game’s greats, we can learn some fascinating insights into how even a World Champion can be swindled.

Tigran Vartanovich Petrosian its powerful and threatens Viktor Kortchnoi to mop up Black’s weak pawns on tt 1963 (2) h6 and e5, while he has his own monster-passer on d5. Black’s position seems completely devoid ._T_._._ of counterplay. j._._M_. In Petrosian’s own words: ‘For .j._._Rj a long time I had regarded my _.jIj._I position as a winning one. Thus ._I_Ij._ the whole opening phase of the _I_._K_. struggle, when Kortchnoi was I_._._._ unable to get out of trouble, had „ _._._._. psychologically attuned me to the idea that the ending would This game was played shortly be favourable to me.’ And indeed after Petrosian won the World it is – perhaps too favourable, for Championship. White is the world’s best player no doubt dominating the endgame. His rook expected a quick capitulation from controls the only from his opponent. Note (in Petrosian’s

35 Part II – The Psychology of Swindles assessment) the recognition of the real swindle after spotting the psychological factors at play. thin veneer of an alternative trick. 32...♖f8! Thinking that the trap has been Rather than passively defending averted, Petrosian decides to put the h-pawn by 32...♖h8, Kortchnoi the game to bed: ‘threatens’ to create some 35.♖xh6?? f3!! counterplay after 33.♖xh6 ♖g8, though in reality the activity ._._.t._ amounts to nothing. Rather than j._._M_. calculate, Petrosian prefers the easy .j.i._.r path, keeping control of the g-file. _.j.j._I 33.d6 ._I_I_K_ Now 34.♖xh6 is certainly _I_._J_. threatened, as 34...♖g8 is met by I_._._._ 35.♖g6!. _._._._. 33...♖h8 34.♔g4 ♖f8! The true trap is revealed, executing ._._.t._ the swindle in fine style. Petrosian’s j._._M_. comments are telling: ‘I did not .j.i._Rj even see the threat ...f4-f3, possibly _.j.j._I because it was in contrast to Black’s ._I_IjK_ hopeless position. Personally, I am _I_._._. of the view that if a strong master I_._._._ does not see such a threat at once _._._._. he will not notice it, even if he analyses the position for twenty or A cunning switchback, and the real thirty minutes.’ set-up of the swindle. Kortchnoi 36.♔g5 ♔e8 appears to have given up hope. But White resigned. perhaps this is a desperate attempt to prevent 35.♔f5 – thanks to the Psychologists have identified sneaky 35...f3! 36.♖f6+ ♔g7! – at literally hundreds of biases that the necessary cost of the h6-pawn the vast majority of us have. They because, as mentioned above, are part of what make us human, 35.♖xh6 ♖g8+ 36.♖g6! wins for and it’s important to remember White. that they are not always bad – in ‘That must be it!’, thinks Petrosian. fact, some of them are useful rules This, as we will see, is a fine of thumb that can help us make example of a decoy trap, whereby an decisions quickly and that work unwitting victim can be encouraged most of the time. But biases can to ignore his suspicions about the also be exploited, as anyone in

36 marketing, politics or real estate circumstances, the position, or will attest. even just their mood on the day. It’s your job to stay attuned to For our purpose, there are four your opponent’s psychological main biases that can be targeted weaknesses and be aware of the by the would-be Swindler. I’ve possibilities of exploiting them. named them the ‘Four Flaws of the Later in this book, we’ll talk about Chess Psyche’, but only because I different techniques to do this. couldn’t think of a catchier name. But first, let’s meet the four flaws: Every player is susceptible to Impatience, Hubris, Fear and one or more of them, depending Kontrollzwang. They will soon be on their personality, the match your new best friends.

37 Part II – The Psychology of Swindles

CHAPTER 2 Impatience

So I made what some called my biggest blunder of the tournament. I awakened Fischer. – , on noticing Bobby Fischer had fallen asleep during their game in the 1963 US Championships. Fischer went on to win both the game and the tournament.

A delayed flight, a slow barista, a queue for the toilets – it can be enough to drive a person crazy. Impatience is one of the most common human foibles, and chess is no exception. How often have you sat behind a winning position while your oblivious, soporific opponent just sits there instead of resigning? ‘My God,’ you think to yourself, ‘He’s not just wasting my time; he’s wasting his own, too. What an idiot!’ You stare at him until, finally, painstakingly slowly, he reaches out and makes a move. Instantly you bash back your reply and glare back in his face, silently demanding his resignation. Alas, instead, he goes back into his favourite thinking pose, and the dance continues... Such situations can seem exasperating, but warning lights should be going off in your head. These are first-grade, prime-rib conditions for a swindle. The desire to finish off a game as quickly as possible is perhaps the most common cause of failures to win a won game. One of the most famous examples features the great in the 1942 US Championships. It’s so well known that you’ve probably seen it before in a puzzle book somewhere:

Carl Pilnick Queen endings are perhaps the Samuel Reshevsky most protracted of all endgames, New York ch-USA 1942 (12) and are frequently a catalyst for impatience (and swindles!). It might ._._._._ seem that the position is devoid of mJ_._._. resources, but White finds a way J_._._._ to set up one final trap. Key to its i._Q_.j. success is the fact that while queen ._._._.j endings are tedious, pawn endings, _._.d._. of course, are much simpler. ._._._._ Pilnick’s move offers Reshevsky a † _._._._K chance to be clever and finish the game immediately.

38 Chapter 2 – Impatience

92.♕f5!! 2451 And into the swindle he falls! With Bonelli 2652 92...g4?? Benasque 2018 (4) he intends to meet 93.♕xg4 with 93...♕e1+ 94.♔g2 ♕g3+, after which his king saunters around to capture ._L_.t._ the a5-pawn. However, his short- _._._.t. cut to victory has one huge, gaping J_._Jm._ problem: _J_._J_Q 93.♕f2!! .i.d._._ _._._.r. ._._._._ I_I_._Ii mJ_._._. † _._._R_K J_._._._ i._._._. Hopelessly lost, White gives his ._._._Jj higher-rated opponent a chance to _._.d._. be clever instead of patient: ._._.q._ 33.♖d1! Now almost any sensible move _._._._K wins, but the strong thinks he sees a cute timesaver. After the forced 93...♕xf2 it’s 33...♖h8?? stalemate: ½-½ ._L_._.t We’ll see plenty more _._._.t. cropping up throughout this book. J_._Jm._ But it is the lead-up play – the _J_._J_Q swindle set-up – that should really .i.d._._ interest us about this example. _._._.r. Offer your opponent a fraudulent I_I_._Ii shortcut to victory, and you will be _._R_._K surprised how often they veer off the winning path. The rook cannot be captured on Reshevsky’s impatience caused him account of back-rank mate, while to look for a simple solution. But the retreats 34.♕f3? ♖xg3! and ironically, impatience can also lead 34.♕e2? ♕h4! lose on the spot. a player to look for complications And all other moves simplify the instead of playing simple chess, so position to a trivial win. All other long as it seems to hasten the end. moves, that is, except for one. Here’s a recent example. 34.♖g6+!!

39 Part II – The Psychology of Swindles

Instead of a dazzling victory, Black The World Champion is two will find himself a whole rook pawns down with a horrible down after 34...♔e7 35.♖xd4 ♖xh5 position to boot. His bishops are 36.♖xg7+ ♔f6 37.♖c7. And the more hopelessly hemmed in by White’s straightforward point is 34...♖xg6?? commanding pawns on d5 and 35.♕xh8 is CHECK. The stuff of f6, and his kingside is in tatters. nightmares. 1-0 Knowing that he has minimal survival chances following a normal In this example, the strong GM course of events, Carlsen makes a playing Black displayed traits of calculated gamble: both impatience and hubris, and 34...♗xd5!! indeed, these two are often found ._.t.l.m together when a player is winning _._._D_. against a weaker opponent, as J_._.i.j we will see in the next chapter. _._Lj._J But impatience also occurs when .t._._.i playing against a stronger opponent. _.n.bR_. The possibility of a memorable victory can lead a player to try to Ii._.qI_ force the issue, perhaps under the _._Rk._. impression that ‘the longer the Snatching White’s trump card, game goes on, the more chances Black also puts himself in a there are for me to go wrong’. This seemingly paralysing d-file . The may even sometimes be true, but pin is so debilitating that White your wily opponent might be wise wins slowly with any ‘normal’ to your mindset and take advantage move, such as 35.♖f5! (eyeing up of your agitation. the pawns on e5 and h5) or 35.a3!, evicting Black’s most active piece. 2772 But Aronian thinks he spots an Magnus Carlsen 2870 immediate execution, and, after London 2019 (2.2) only 90 seconds of thought, he plunges head-first into the trap: L_.t.l.m 35.♗c5?? _._._D_. Seemingly crushing. If 35...♗xc5 J_._.i.j 36. ♕xc5, Black loses his bishop _._Ij._J on d5. But moving the rook on .t._._.i b4 anywhere allows 36.♗xf8 and _.n.bR_. Black would in any case need to Ii._.qI_ remove a key defender from the _._Rk._. d5-bishop, losing material. Did I say „ ‘anywhere’?

40 Chapter 2 – Impatience

35...♖d4!! A followed Aronian’s choice as well: ._.t.l.m 36.♗xd4 ♗xf3! 37.gxf3 exd4 _._._D_. And, although White retains an J_._.i.j edge, the tables have certainly _.bLj._J turned. The game was eventually ._.t._.i drawn, Aronian’s impatience having _.n._R_. cost him a famous victory. Ii._.qI_ _._Rk._. *** There are many causes of The brutal point of Black’s play. impatience at the board: trying to With one beautiful swoop, Carlsen win a won game, catching the last simultaneously cuts off the lines train home, a hot date17... but for of communication on the d-file some people, impatience is just an and the g1-a7 diagonal. No matter unshakeable personality trait. As how White captures next, a cruel we saw above, it often occurs along intermediate move awaits him. If with hubris, which is no less a 36.♗xf8 ♖xd1+! and Black saves weakness for the eager Swindler to the piece, while 36.♖xd4 ♗xc5! exploit. Turn the page. turns the pinner into the pinned.

17 See the chapter on Gamesmanship.

41 Part III – The Swindler’s Toolbox

CHAPTER 13 Window-Ledging

You must take your opponent into a deep dark forest where two plus two equals five, and the path leading out is only wide enough for one. – Mikhail Tal

We’ve seen how a Trojan Horse exploits impatience, the Decoy Trap exploits hubris and a Berserk Attack exploits fear. Only a creative name will do to describe the tool to exploit Kontrollzwang. Unfortunately, I couldn’t think of one. But a friend of mine, a fiction author, came to my rescue. In many a classic action film with sword- fighting, the weaker swordsman, facing imminent death, climbs out onto a precarious window ledge. The implication for the dominant swordsman is clear: follow into a terrain where risk, uncertainty and deadly perils are rife for both, or give up the pursuit – and with it the advantage. That’s the idea of Window-Ledging that I want to convey in this chapter. Our opponent is cruising towards victory, completely in control. We need to change the environment into one where both sides can easily slip up. We need complexity; we need complications. We need chaos.

Malcolm Armstrong 2126 importantly, quite easy to play for Thomas Rendle 2416 White, as Black has virtually no Liverpool 2007 (4) counterplay. Instead, Tom steers the game into the sort of positions in L_._.t.m which a lower-rated player feels very _.jN_.jJ much not at ease: being under attack! ._._.s._ 25...♗xe4!! 26.fxe4 ♘g4 27.h3 ♘e3 _._.j._D ._._.t.m ._B_Ij._ _.jN_.jJ _.i._I_. ._._._._ .iQ_._Ii _._.j._D „ _._R_.k. ._B_Ij._ _.i.s._I Tom later told me that he knew that 25...♘xd7 was objectively forced, .iQ_._I_ but he didn’t trust himself to hold _._R_.k. the draw against his lower-rated A colossal . The beauty of this opponent. Indeed, the resulting swindle is that White appears to position is pretty grim and, have such an obviously winning

100 Chapter 13 – Window-Ledging move in 28.♕e2 that he doesn’t 31...♕g1+!! bother to search for alternatives. The killer tactic, without which Without such an obvious candidate, Black can resign. perhaps he would have found the 32.♔g3 f1♘+ 0-1 crushing 28.♘xf8!! ♘xc2 29.♖d8, Black’s sacrificed knight rises capping off a splendid upset. like a phoenix to deliver a sweet Instead, the weaker player falls prey finish. The prosaic 32...f1♕ was also to his instincts and jumps at the sufficient, but who could resist the opportunity to swap the queens. underpromotion? An educational 28.♕e2 f3!! victory by one of England’s trickiest ._._.t.m masters. _.jN_.jJ ._._._._ Window-Ledging can be an _._.j._D especially effective technique when ._B_I_._ your opponent is short of time. _.i.sJ_I After all, every move requires a lot more care when you’re standing on .i._Q_I_ a ledge. _._R_.k. White’s kingside, so unassailably David Smerdon 2521 secure a few moves earlier, is being Jakob Aabling Thomsen 2370 ripped apart. In fairness, White may Helsingor 2013 (4) have seen this resource in advance, but missed Black’s later blow. After all, Black is sacrificing almost his ._T_M_.t entire army, while the promotion j._.j.lJ square is still defended by the bishop. .jL_.sJ_ How bad can things really get? _.jN_.b. 29.♕xe3?? ._._._._ 29.♖f1!! is a tough move to find, but _._R_N_I it was the only way to win. IiI_._I_ 29...f2+ 30.♔h2 ♕xd1 31.♘xf8 † _.kR_._. ._._.n.m _.j._.jJ White’s pawn has fizzled ._._._._ out and Black is only a few moves _._.j._. away from consolidation. Behind on ._B_I_._ material and behind on the clock, I _.i.q._I made a practical decision to muddy the waters. .i._.jIk 21.♘xe7!? ♔xe7 22.♖e3+ ♔f8 _._D_._. 23.♘e5

101 Part III – The Swindler’s Toolbox

._T_.m.t ._T_._Mt j._._.lJ j._._._J .jL_.sJ_ .j._.bJ_ _.j.n.b. _.j.n._. ._._._._ ._._._._ _._.r._I _._.r._I IiI_._I_ IiI_._L_ _.kR_._. _.k._._. 26...♖f8 Setting Black problems. My Hoping to get a favourable endgame talented opponent rose to the after 27.♗xh8, and some much-needed challenge and played the next few fresh air for his kingside pieces. moves accurately, but spent valuable 27.♘g4! minutes in the process. ._._.tMt 23...♔g8 24.♖f1! j._._._J .j._.bJ_ ._T_._Mt _.j._._. j._._.lJ ._._._N_ .jL_.sJ_ _._.r._I _.j.n.b. IiI_._L_ ._._._._ _.k._._. _._.r._I IiI_._I_ Blitzed out without a pause. This was important, as it gave my opponent the _.k._R_. opportunity to conclude that I had blundered. With only a momentary Hoping for 24...♘d5?? 25.♘xc6!, hesitation, he fell into temptation. when Black can’t take either piece 27...h5?? without allowing checkmate. 27...♖xf6! would have maintained a 24...♗xg2 decisive advantage. A natural clarifying move in light 28.♖e7! ♖h7 of White’s window-ledging. More ._._.tM_ precise was 24...♘d7!, which looks j._.r._T risky in light of 25.♘xc6 ♖xc6 .j._.bJ_ 26.♖e8+. But although he is tied _.j._._J up temporarily, Black can hold ._._._N_ fast with 26...♘f8! (not 26...♗f8?? _._._._I 27.♗h6!) and White’s will eventually evaporate. IiI_._L_ 25.♖xf6! ♗xf6 26.♗xf6 _.k._._.

102 Chapter 13 – Window-Ledging

29.♘h6+! position is objectively lost in any An abrupt perpetual. case, does this really matter? More 29...♖xh6 30.♖g7+ ♔h8 31.♖f7+ importantly, he narrows the path to ½-½ victory that Magnus must follow. 49...e4! 50.♖xh5 e3! It’s uncomfortable out on the window ledge, especially if you’re ._._._._ the sort of player who loves being _.t.m._. in control. That pretty much ._S_T_._ describes all chess players, even up _._R_._R to the World Champion. ._._._.i _._.j.i. Magnus Carlsen 2840 ._._._Bk Anish Giri 2773 _._._._. Wijk aan Zee 2017 (7) Believe it or not, after this move White can force mate in no more ._._._._ than 25 moves (technology, hey?). _.t.m._. But there’s little room for error out ._S_T_._ on the ledge; Magnus needs to be _._RjR_J accurate, or else... ._._._.i 51.♖h7+ ♔e8 52.♖xc7 e2 53.♗f3! _._._.i. e1♕ 54.♗h5+ ♔f8 55.♖f5+ ♔g8 ._._._Bk _._._._. ._._._M_ „ _.r._._. Black has been thoroughly ._S_T_._ outplayed and now faces a lost _._._R_B endgame against the world’s best ._._._.i player. The most stubborn defence _._._.i. is 49...♖h6, but after 50.♖b5 this ._._._.k will almost certainly lead to a _._.d._. rook endgame where White is two pawns up, and Magnus would be 56.♗f7+? expected to win this 95 times out It is likely that Magnus calculated of 100 against any player on earth. this variation when choosing his Instead, Giri uses the concept of 50th move, because he played the Window-Ledging to create the bishop check instantly, overlooking maximum chances of a swindle. an elementary mate: 56.♖c8+ ♔g7 His next two moves are frowned 57.♖f7+ ♔h6 58.♖h8#. I’m not sure upon by the computer; but if his if Giri had also missed this, or

103 Part III – The Swindler’s Toolbox instead had seen it and decided it 66.♗g2+ and the queen is lost. Of was still worth the gamble. If the course, all of this was too hard to latter, then his swindle deserves fathom in advance, but the point of even more kudos. Giri’s play is clear: by sharpening 56...♔h8 57.♖h5+ ♔g7 58.♗xe6+ the position, the chances of a ♔f6 blunder (by either player) increase dramatically. ._._._._ 59...♔e5 60.♗h3 ♕d2+ 61.♗g2 _.r._._. ♕xh6 62.♖xc6 ♕h7 ._S_Bm._ _._._._R ._._._._ ._._._.i _._._._D _._._.i. ._R_._._ ._._._.k _._.m._. _._.d._. ._._._.i _._._.i. Magnus had calculated to this point ._._._Bk and evaluated it – correctly – as _._._._. winning. But one final accurate move is required, and here the The smoke has cleared, and the champion falters: swindle has been successful. While 59.♖h6+?? Magnus pressed for another 60 59.♗c4! is the only winning move, moves, Giri was able to hold the covering e2 and renewing the endgame with little trouble. A threat on the black knight. There is fantastic example of using the no defence: 59...♘e5 60.♖h6+ ♔f5 window-ledging technique to 61.♗d3+!! ♔g4 62.♖g7+ ♔f3 63.♖f6+ maximise one’s chances, and ♔e3 against no less than a World Champion. ._._._._ _._._.r. There are many examples of ._._.r._ Window-Ledging among the games _._.s._. of the great tacticians, such as Tal, ._._._.i Kasparov etc. But our final example _._Bm.i. is from one of my favourite amateur ._._._.k players, the German FM Olaf _._.d._. Steffens, whose unbreakably upbeat outlook on life is reflected in his analysis diagram optimistic chess. A diehard fan 64.♗f1!! with the lovely point that of 1.b4 with the white pieces and 64...♘f3+? loses to 65.♖xf3+ ♔xf3 similar offbeat openings as Black,

104