There are many ways in which chess literature suffers from the fact that the game is not a recognised academic study. Take, for example, the imprecision of our terminology; we all use various chess words but seem to mean different things when doing so. The confusion continues in perpetuity because there is no recognised authority to define these terms for us.
Perhaps the best example of an abused chess word is ‘swindle’. Some people say that they have ‘swindled’ their opponents whenever they have won (or drawn) games which they should have lost; but this cannot be the technical meaning of the word. A swindler is more than a passive recipient of an undeserved point – he has done something fiendishly clever, and has made his own luck. But this does not get us far in the search for a definition. What is the difference between a swindle and a cheapo (by which I mean a tactical trick); or between a swindle and a well-concealed threat; or between a swindle and a methodical fightback, or a general randomisation of the position which confuses the opponent? Is there a difference between A swindling B, and B effectively swindling himself? Swindling is a skill which is distinct from, and higher than, these other facilities; indeed, I think that a true swindle is something of an art. My definition would be:
When a player plays a move which is designed to (and does) provoke an error from his opponent, and this error makes possible a profitable combination, that player has swindled his opponent.
This formula should immediately clarify one point; the swindler must surreptitiously enlist his opponent’s help in making the combination possible. So if A simply makes a threat which B misses, then B has not been swindled. The combination was possible before B’s move; B may have allowed it to happen but he did not create the opportunity for the combination – it was there already. Take the example Gosling–Rudd (BCM, February 1996):
Black’s co-ordination is pretty poor, and his 1…Nf4! must have looked like desperation. White decided to meet the obvious threat of 2…Ne2+ by 2 Bd8?? (2 g3! Gosling) which threatens 3 Qe7+ and 4 Qe8 mate. Only then did he see Black’s real threat, viz 2…Qf1+! 0-1 (3 Kxf1 Rh1 and 4 Rxf1 Ne2 are both mate). As two-move cheapos go, this is about as good as you can get. But I would not call it a swindle, because the loser simply allowed his opponent to execute a threat made on his previous move.
Nonetheless, my formula prompts other questions.
How might an opponent’s error ‘make possible’ a combination?
The opponent must play a move which enables the swindler to play a combination which was not previously possible. Take this position:
Here Black played 1…c5?, a natural move which intends 2…b4, but which loses a piece to 2 Rxd3! (2…Rxd3 3 Qxa8+ or 2…Qxd3 3 Re8+!). The combination was only possible after Black’s error; had Black, say, doubled rooks on the d-file, Lasker would have had to try something else.
How does the swindler ‘design to provoke’ this error?
The difference between a swindle and a cheapo is the provocation of the opponent’s error. A ‘cheapo artist’ simply finds a tactical trick which the opponent has overlooked; the swindler sees the trick coming, and induces the opponent to ‘set himself up’ for it. Returning to Lasker–Scheve, had White only found the combination after Black played 1…c5? then he would have ‘cheapoed’ Black. But if White had already seen that 1…c5? loses, and played a move which was designed to provoke it (perhaps his rook was on a1, and he played 1 Rad1, unprotecting the a2 pawn and thus prompting Black to adopt the plan of 1…c5 and 2…b4), then he would have swindled Black.
So a swindle has two ingredients: the ‘provocative move’ and the combination itself. It is evident from the Lasker example that the provocative move includes an element of psychology. The opponent is never forced to make the combination possible; he should always have a decent alternative available. There must therefore be something about the provocative move which gives the opponent the idea that he should adopt a certain plan, or that he can play a combination himself. My own best swindle is perhaps the following:
You will notice that I am a piece down, and it was only on account of Miroslav’s acute time shortage (about a minute for the next ten moves) that I was playing on. Since White had even used quite some time over his last move (1 Qxa7) I guessed that something had to be wrong with the obvious 1…Qxc3, and saw that 2 Qxb7 Rxd3 3 Rxd3 Rxd3 4. Rxd3 Qxd3 5 Qb8+ Kg7 6 Qg3+ wins for White. But I then saw that 1…f4! (the ‘provocative move’) was possible, and immediately played it. Houska, presumably thinking that I had blundered whilst trying to blitz him, flicked out 2 Bxh7+? only to find that after 2…Kxh7 3 Rxd5 Rxd5 4 Rxd5 Qg6+! 5 Kf1 Qb1+ 6 Ke2 Qe4+ I had regained my rook. White still stands better, but it would have taken a superhuman not to have been affected by this turnaround, and after 7 Kd2 Qd5+ 8 Qd4 Qxa2+ 9 Kd3 Qb1+ 10 Ke2 Qc1 11 Kf3? Qh1+ we agreed to a draw.
Paradoxically, this swindle only worked because Houska is a good blitz player and was thus able to see 2 Bxh7+? instantly – without having the time to check and double check his calculations. A less strong player might not have seen 2 Bxh7+ so quickly, would have played a safer move such as 2 Kh1, and would probably have won! This example shows that it is the provocative move which constitutes the artistic element of the swindle – the swindler has to assess the probability of his move provoking the desired error, which in turn requires a psychological assessment of the opponent. In this case, I felt instinctively that Houska would be tempted; up to this point, his tactical play had been so strong that he would be bound to see 2 Bxh7+, and my own tactical play so poor that he would not be immediately suspicious at my supposed blunder (though he might have been had I delayed over the move).
Must the swindler have foreseen the combination when he plays the provocative move?
Quite often, one sees a player complicate (‘randomise’) a position, generally because he expects to lose in the ordinary course of play. He may blitz out a few aggressive moves, without calculating specific variations, but then slow down if he suspects that his opponent has made an error. If he then finds a winning combination, has he ‘swindled’ his opponent? In my view, no: the swindler must have designed to provoke a specific error and have prepared a specific combination in case that error is played. By contrast, the randomiser takes what comes his way, and may be every bit as surprised as his opponent at what turns up:
D. Rosen – J. Rogers
Not feeling very happy with my compensation for the pawn, I had staked everything on a kingside assault and now played 1…g4. White could now refute my play with 2 hxg4 hxg4 3 Nh4, and the threat of 4 Nxf5 is hard to meet: 3…g3 is well met by 4 Rxg3+. I had not seen this – in fact I had not dared to analyse my prospects too closely and was just making the only moves which could conceivably cause problems. But my opponent played 2 Nh4? without swapping h-pawns, and I now saw that I was winning: 2…g3! 3 Rxg3+ (3 Nxf5 gxf2+ 4 Kf1 Qh2) 3…Nxg3 4 Rxf4 Re1+ (my opponent had assumed that I would play 4…Ne2+, picking up two rooks for the queen but with a hopelessly exposed position) 5 Kh2 Nf1+ 6 Kg1 (had he swapped h-pawns, the king could go to h3) Ne3+ 0-1.
Of course, I could claim some credit for having complicated the position, which resulted in my opponent’s error; but when I played 1…g4 I had not seen why 2 Nh4? was bad. It was White’s idea, not mine, that he should initiate the line of tactics starting with 2 Nh4. For all I knew when I played 1…g4, 2 Nh4 might have won, so if anyone swindled my opponent, it was my opponent himself. Only if I had seen that 2 Nh4 lost and had played 1…g4 in order to provoke it would I have ‘swindled’ Rosen.
Indeed, I subconsciously perceived the difference between swindling and randomisation at the time. In the post mortem with Rosen I was rather subdued, feeling somewhat apologetic for having won in this way. But against Houska I felt that I had fully deserved the half point in spite of my previously atrocious play. The difference was all in the mind: I knew that I had deliberately ‘set up’ Houska but felt that I had not swindled Rosen.
Of course, one objection to this is that a third party can never tell when a swindle has been perpetrated without reading the swindler’s mind. However, one can often tell what the swindler has foreseen from his demeanour – I stopped to think after Rosen’s 2 Nh4, but replied instantly to Houska’s 2 Bxh7+. Also, the problem of ‘mindreading’ is not unique to swindling. Whenever one decides a best game competition, for example, one is having to guess that the winner had seen certain points, or had intended certain consequences.
What is a ‘profitable’ combination?
Ideally, the swindler emerges from his combination with a clearly winning (or clearly drawn) position; but it may be that the combination simply improves his position without securing the desired result. This was the case in Houska–Rogers: I was still a pawn down after my swindle. Thus, it should be enough that the swindler’s combination has substantially improved his chances. It is quite possible to speak of the eventual loser having swindled his opponent at some earlier point in the game.
Must the swindler have a lost position when he plays the provocative move?
Many would answer ‘yes’, but this entails some difficulties. For a start, as any adjudicator knows, it can be difficult to distinguish a lost position from a rather poor one. More importantly, those who insist that the swindler must have been doing badly tend to miss the artistic element of the swindle. They assume that a swindle is a combination played in a bad position. Once one focuses on the swindle as the art of provocating errors as well as the combination, then there is no reason to limit the definition of the swindle in this way – it is already distinct from the ordinary cheapo. I believe that one can ‘swindle’ from a level position too. But, as always, the opponent must have a decent alternative to his erroneous move; so someone who has such a good position that his opponent has nothing obviously better than to make possible the swindler’s combination cannot be said to have swindled him.
The good and the bad swindle
So much for the definition of a swindle; but if a swindle is an art, then we should have some criteria for distinguishing a good swindle (i.e. a pleasing one) from a bad swindle (an unimpressive one). Whilst it is not necessary for the swindler to have had a lost position, clearly the worse his position, the better the swindle – because there are fewer resources in bad positions, and the swindle may even be the best chance. We can also say that the more plausible the opponent’s error, the better the swindle. But, given that it is the ‘provocative move’ which contains the artistic element of the swindle, it is the subtlety of this move which reveals the good swindle. Ideally, the provocative move will make a threat of its own, albeit not a very dangerous one, or one which can safely be neutralised. In neutralising this threat, the opponent makes possible another combination – the real point behind the provocative move. An example of a good swindle is given by Glenn Flear (Chess, January 1989):
White, clearly losing, played 1 Rb3! This ‘obviously’ threatens 2 Rb7, which is not enough to save the game but is an irritant. Wicker decided to annex White’s a6 pawn as the cleanest way of wrapping things up, and played 1…Re6?? However, this made possible White’s real idea, viz 2 Rb8+! 1-0 (2…Kxb8 3 Rh8+).