This is a lament that every chess player utters at some point… some of us more often than others. My turn to utter it was yesterday.
FEN: 7k/7p/P2R2p1/5p2/2p5/1K6/6rP/8 w – – 0 48
Position after 47. … c4+. White to move.
Round one of the Western States Open in Reno. I’m playing White against Samir Alazawi, a class A player. Here is where I had my moment of chess blindness.
I probably spent 10 minutes thinking about this position. The first move I wanted to play was 48. Ka3, to keep Black’s rook off the a-file. Obviously if 48. … c3 49. a7, and White’s pawn wins the race. But then I started looking at 48. … Rxh2, and I couldn’t see a win! After 49. a7 I suddenly thought, “Wait a minute! He can play 49. … Rh1! If I queen my pawn, he’ll play 50. … Ra1+ and skewer my king and queen!”
So then I started looking at the alternative, 50. Rd8+ Kg7 51. a8Q Ra1+ 52. Kb4 Rxa8 53. Ra8 (diagram).
FEN: R7/6kp/6p1/5p2/1Kp5/8/8/8 b – – 0 53
Position after 53. Ra8 (analysis). Black to move.
At this point, I have to say, my mind just boggled. I wasn’t sure whether I was winning, drawing, or losing — and a difference of a single tempo can easily change the result. I couldn’t really force myself to analyze this position, because there are a hundred million different variations and the crucial moments won’t come for another six or seven moves. So finally, I just gave up and played the safer move back in diagram 1, namely 48. Kxc4. Of course he then brought his rook to the a-file with 48. … Ra2. Maybe White can still win somehow, but I wasn’t able (with limited time) to figure out how, and the game was soon drawn after 49. Kb5 f4 50. Rf6 g5 51. Rf5 h6 52. Kb6 Kg7 53. a7 Rxa7 54. Kxa7 Kg6 55. Rf8 Kg7 56. Rf5 ½-½.
Those are the facts. Now, detective, what did I miss?
Answer: In the variation highlighted in red, after 48. Ka3! Rxh2 49. a7 Rh1 50. a8Q+ is check. He doesn’t get a chance to play 50. … Ra1+ and skewer my rook. Instead, to add insult to injury, I fork his king and rook!
It’s so head-smackingly simple. And the irony is that in the other variation, 48. … c3 49. a7 c2 50. a8Q+, I did realize that I was queening with check. But somehow I forgot it when I got to the other variation.
The only thing that makes me feel a little bit better is that I was completely lost earlier in the game, so my opponent could say exactly the same thing as me: “How could I be so blind?”
P.S. If you’re wondering about the rook-versus-three pawns endgame in diagram two, it is in fact won for White. I checked in the Nalimov tablebases. In order to draw, Black needs to get his pawns to h5-g4-f3 or to f5-g4-h3, and he’s a tempo short. For example, 53. … h5 54. Kxc4 h4. If it were Black to move, then … g5 would draw. But with White to move, 55. Kd5 wins (and if 55. … h3 56. Ke5!)
The moral here is that if you know a key position, you don’t have to analyze a hundred million variations. You can just count tempi. Unfortunately, I have never studied rook versus 3 pawn endgames, so I didn’t know the key positions.
