Author: David Rasmussen
Date: 05:02:18 12/24/00
Actually, it has nothing to do with christmas, but here goes: [D]8/8/5n1p/p1k1pBpP/1p2P1P1/1P2K3/P7/8 b - - 0 1 This position is from Damjanovic-Fischer, 1970. I've gotten it from Silman's masterpiece, How To Reassess Your Chess. He uses it as a demonstration of how important it is to penetrate the enemy position with the king, in such an endgame, and that bishops aren't superior to knights in endgames with pawns only on one side of the board. The line played was 1...Kb5 2. Kd3 a4 3. bxa4+ Kxa4 4. Kc4 Ka3 5. Kc5 Kxa2 6. Kxb4 Kb2 Here, Silman says that black is winning, because the black king gets into the enemy position, and because the knight is now equal to the bishop. My problem is that Chezzz doesn't find this line. Now Chezzz is no genius, especially not in endgame situations, but when I run a search from the resulting position, it gives black a +0.60 advantage, so it should be able to see this. Never the less, from the starting position, after searching 15 plies deep, it still evaluates the position as 0.01 with a boring line of just moving the pieces around. How do other programs do in this position? It doesn't seem very difficult. What is the most important evaluation factor, that Chezzz might be missing, that will decide this position? Chezzz doesn't know that a bishop is better when there are pawns on both sides. Maybe that is the problem.
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.