The Unseeing Mind
Why do we miss some obvious things in chess? Is it just because we are not good enough? Or are there some other reasons – perhaps psychological reasons – why we might miss something that is right in front of our eyes?
It seems to me that it is a combination of the two. As humans, we have psychological weaknesses that prevent us from seeing everything. The stronger we are in chess – and the more typical ideas we know – the more likely we could override some of these psychological weaknesses. Let me give you an example, which is what gave me the idea to write this article:
White has just played 51.Rh8. He is attacking my h3 pawn, so clearly there are a very limited number of possible moves: 51…Rg3, 51…h2 and…??
That’s right, did you see the move 51…Rh2? It’s pretty obviously one of only three moves that don’t lose the h-pawn. But I failed to consider it at all during the game. And I wouldn’t be surprised if there are other good players who would also miss this move. The reason is that it breaks one of the cardinal rules of rook-and-pawn-endgames: don’t put your rook in front of the passed pawn.
As a matter of fact, though, this move wins trivially. The reason is that by playing …Rh2, followed by …Rh1 and …h2, Black prevents the approach of the white king. Imagine Black plays 51…Rh2. White tries to approach with the king by playing 52.Kb3. Black continues with 52…Rh1 53.Kc3 h2. Now if the king moves closer by 54.Kd2, there is the standard trick 54…Ra1 55.Rxh2 Ra2+, winning (play through this on the next diagram if you have trouble visualizing it).
White’s rook will also then be tied to the h-file (if it leaves the h-file, Black will have a chance to move the rook and queen the pawn). While Black’s maneuver looks strange, immobilizing his own rook, the most important thing is that it prevents the approach of the White king.
I believe if I had seen the move 51…Rh2 at all, it would not have been hard to see that it wins easily. After you find that White’s pieces are immobilized, the only thing left is to check that there is no fortress for White. But since the Black king can easily invade via the g5 and f4 squares, he will win the f3 pawn and White will also not be able to defend e4. It could go like this:
Instead, I considered first the most natural move 51…Rg3 (attacking the f3 pawn). Experienced players know that it can be dangerous to take the next-to-last step with the passed pawn in rook and pawn endgames – you may need the h2 square to hide the king. But after 51…Rg3, white plays 52.Kb3 and brings his king over. Black wins the f3 pawn, but in the meantime the White king goes to d5. I saw variations where White draws with a passed e-pawn against a rook. Therefore I began calculating 51…h2. I made a mistake in my calculations, and that move failed to win either, and the game was a draw.
What I didn’t do though, was to first list the candidate moves. If I did (and there were only three) I would surely see 51…Rh2, and realize quickly that it won. Missing 51…Rh2 was partially caused by dogmatism (“don’t put the rook in front of the passed pawn”). Another part of it was some frustration: I thought the position was easily winning a couple moves ago, and didn’t imagine that such weird moves should be necessary.
Of course, it is not as if I had not seen such an idea – to push the pawn to the last step and put the rook in front of it, to prevent the enemy king from approaching. But in the current position it didn’t occur to me. Now that this theme has been more strongly reinforced, I would be unlikely to miss a similar idea in the future, despite the fact that it looks weird. Broadening your reserve of chess ideas is one way to fight against such oversights.
This is one common reason why we might fail to consider moves that are right in front of our eyes – the moves look unnatural or break some kind of general principle. Players often overlook retreating moves – thus we have the famous Christiansen-Karpov encounter:
Another kind of “blind spot” which is probably very common among class players is to overlook a “quieter” move in a position which seems to promise a lot of action. This reminds me of one old game of mine:
Black’s position at move 16 looks quite unpleasant, with his king in the center and facing the protected passed pawn on d6. It seems to promise a quick breakthrough with e5-e6. I can remember calculating for a while to try to make this breakthrough work – probably I looked at moves like 16.e6, 16.Nd4, or 16.Rae1.
I was getting bogged down in the variations and not finding anything convincing, when finally my mind cleared up. I saw that I did not have to break through immediately, but that I could simply prevent the intended …Nf4-e6 maneuver, as well as …b5-b4, with the prophylactic move 16.Qd4! This was totally decisive. I also realized that breaking through with e5-e6 was not necessary, and that I could play on the queenside as well, by a2-a4, which would either win a pawn on b5, or decisively allow my knight to come to a4 and then b6 or c5. Thus another reason why we might make some simple oversights is because of tunnel vision. I was able to overcome it in this game, but probably other times I did not.
How can we get rid of these blind spots? One important way to improve this is to get in the habit of listing candidate moves. This method was famously publicized by Alexander Kotov in his books Think Like a Grandmaster and Play Like a Grandmaster. While it is true he has been criticized, and it is true that real chess players do not think in such an ordered, robotic way as he suggests, the candidate move method is nevertheless valid. Many players simply jump into calculations of one or another line as soon as their opponent makes a move; but this might lead to tunnel vision. If you find that you are not getting anywhere with your calculations, go back to the starting position and try to “widen your horizon”. If I had done this in the first position (the rook and pawn endgame) I would have surely found 51…Rh2.
