When I was a graduate student, I once heard this story about one of the professors in the department, a world-famous mathematician who was originally from Japan. He was giving a lecture in an undergraduate course, probably calculus, and a student raised his hand and said, “Excuse me, I don’t understand that last step.” The world-famous professor stopped, thought for a moment, and then said, “Understanding is psychological thing.” And then proceeded with his lecture.
I’m not exactly sure what the point of the anecdote is — actually, I think there are several points — but anyway, I was reminded of it on Tuesday, when I went to a lecture by David Pruess at the Mechanics Institute.
David is one of the rising stars of American chess, who holds one of the two Samford Chess Fellowships, which pays him a salary so that he can study chess full-time. David says that he studies about 30 hours a week, plus playing in lots of tournaments.
Last year David earned two GM norms at tournaments in France, which prompted John Donaldson to say in his introduction that David has a 2600 rating in French-speaking countries! (He also had a spectacular 8.5/9 result at a tournament in Canada, which John was including in that statistic.) John said that he is working on making French the state language of California, so that David will play equally well here! 😎
Anyway, the main theme of David’s talk this week was the psychology of chess, and in particular how psychology can cause you to lose a perfectly decent position. He illustrated this with a game he recently played against a Belgian master named Yves Duhayon, who self-destructed in spectacular fashion.
The game (in which Duhayon was White) began 1. Nf3 c5 2. c4 g6 3. Nc3 Bg7 4. e3 Nf6 5. d4 cd 6. ed d5!, reaching the following position:
The exclamation mark is for the psychology of 6. … d5, which was sort of an announcement to White that his opening has not gone according to plan. Instead of the English Opening he was looking forward to, we have now transposed to the Panov-Botvinnik Variation of the Caro-Kann (1. e4 c6 2. d4 d5 3. ed cd 4. c4 Nf6 5. Nc3 g6), except that it’s an inferior version for White. Usually White wants to play 6. Qb3 and capture the d-pawn, followed by moves like g3, Bg2, Ne2, and Nf4, clamping down on d5. With Nf3, White has developed his knight to the wrong square and blocked the long diagonal for his white-square bishop, so this main line loses its punch.
So, as David said, the opening has been an “unpleasant surprise” for White. And it got worse: after White played 7. Bg5 David hit him with 7. … Ne4!, another unexpected move that seizes the initiative.
What should you do when your opponent plays an unpleasant surprise? David said that a common response is to lash out and try to violently change the course of the game — but this is usually the wrong thing to do. In such positions, he said, you should just calm down and play simple, defensive chess. White has given away maybe 0.1 pawn of his advantage by his poor move order, and maybe another 0.1 pawn by playing Bg5, but that’s no reason to panic. Here David suggested that White should play 8. Be3, basically admitting that his move Bg5 was too optimistic. But such a move is psychologically very difficult to play.
Instead of trying to “calm down” the position, White continued to flail around, trying to assert his control in a position where he was no longer in control. The game continued 8. cd Nxg5 9. Nxg5 Qb6 (hitting White where he is weakest, the dark squares) 10. Bb5+ (again too optimistic, in David’s opinion) Bd7 11. Qb3. David felt that this move misplaces the queen. White is trying to play his normal Panov-Botvinnik stuff in a position that doesn’t call for it. David said White should have played 10. Qd2, followed by Nf3 and Be2. It’s okay to play defense!
The game continued 11. … Bxd4 12. O-O O-O, reaching the following position:
Now White completed his self-destruction with 13. Rad1?? Bxc3 14. Bxd7 Qxb3 and White resigned. If 15. ab Bf6, White has two pieces hanging and must lose one of them. In some ways this catastrophe was the logical consequence of White’s faulty psychology, his inability to admit that something had gone wrong.
However, there’s an irony here, which is that in spite of his multiple psychological sins, White is still perfectly all right in the diagrammed position! In fact, he’s more than all right after 13. a4! I wanted to ask David about this move during the lecture but didn’t. When I got home and looked at it on the computer, it turned out that 13. a4 was even stronger than I thought. The position is actually a minefield for Black. Here are the possibilities:
a) The most obvious way for Black to go wrong is 13. … a6? 14. a5! Qc7 15. Bd3, and now the development of Black’s queenside is really messed up, and the pawns on a5 and d5 give White a nice space advantage.
b) The computer likes 13. … a5?! but I don’t; it allows White complete control over b5, and now White can continue with 14. Rad1, the move he wanted to play earlier.
c) 13. … Bxb5? is also horrible for Black after 14. Nxb5 Bg7 15. a5 and the only place for Black’s queen is back to d8.
d) The really devilish trick comes in the line 13. … Bf5 14. a5 Qc7?!, which is exactly what I would have played if I were Black. Amazingly, this natural retreat gets into trouble after 15. Nxf7!? (a total computer move, no human would ever consider it) Rxf7 16. Be8! You might wonder, “So what’s the big deal? White gets a rook and a pawn for two pieces, which is a bad trade for him.” But if you look a little bit closer, you realize that after the exchange of bishop for rook on f7, Black’s king will be exposed to the discovered check d6+!, winning Blacks’ queen! That’s why the retreat to c7 was no good. Black now has to play 16. … Qc8 or 16. … Qd8, but neither one gets him out of trouble. For example, 16. … Qd8 17. Bxf7+ Kxf7 18. d6+ followed by 19. Qxb7 (possibly with check) and Black is feeling a bit like a punching bag. The computer gives White a one-pawn edge, although this could be debated.
Instead of 14. … Qc7 Black could play … Qd8, … Qd6 or … Qf6, but none of these are really natural moves, and it’s not surprising that the computer gives the edge to White in all of these lines.
(e) The only really solid move for Black, I think, is to liquidate some material with 13. … Bxc3 14. bc a6 15. Bc4 Qxb3 16. Bxb3 Rc8, with at most a nominal edge for White. (Actually I think it’s equal.) But speaking of chess psychology, would David really have been able to play this move? He was thinking at this point that he had outplayed his opponent, and the clearest sign of this was his control over the dark squares. So it would have been quite difficult for him to adjust his thinking and give away the dark-square bishop.
The interesting thing about analyzing on the computer is that it has absolutely no idea of the psychology involved. So it finds variations like the critical line above, line (d), which I think that no human except maybe Morozevich would ever have played. When you factor in the psychology, David is probably right — Black stood better in the diagrammed position. If you ignore the psychology, White has good chances to get an advantage! If David ever gets to this position again, he’d better watch out, because it will mean his opponent has prepared 13. a4. Then the psychology will be completely different.
So… Yes, indeed, chess is psychological thing! No chess computer will ever change that.
