Author: Uri Blass
Date: 11:44:52 05/27/02
Go up one level in this thread
On May 27, 2002 at 14:31:16, Bo Sjögren wrote: >Many years ago I was outplayed for most of the game, but my opponent >allowed me to "close the shop": > >[D]8/p7/6kp/1p2p3/1P2P2P/2P5/1P3K2/8 w - - > >White must play 1.h5+!, which I think we concluded after the game wins >rather easily. But instead he played 1. b3?? and after 1...h5! he >thought for five minutes, and shook his head (and my hand)... > >Does the current crop of engines recognise the danger? It seems CT14 >prefers 1.h5+, but is almost as happy with 1.b3 -- just a one minute >check. Will any engine conclude that 1.b3 is indeed a draw, and how >will it do it; by some kind of intelligence, or brute force? (I of >course have the feeling that brute force is not enough for this type >of problem.) > >/Bo S I think that brute force+hash tables is enough for this type of position. It may not be enough to see the draw after 1.b3 but it is going to be enough to see that 1.h5+ is better than 1.b3. Here is some analysis of Crafty18.07 on a slow p200 Crafty can see significantly better score with h5 New position 8/p7/6kp/1p2p3/1P2P2P/2P5/1P3K2/8 w - - 0 1 Analysis by Crafty 18.07: 1.Ke3 +- (1.97) Depth: 1/3 00:00:01 1.Ke3 Kf6 +- (1.87) Depth: 2/3 00:00:01 1.Ke3 Kh5 2.c4 Kxh4 3.cxb5 µ (-0.92) Depth: 3/4 00:00:01 1.Kf3 ³ (-0.52) Depth: 3/4 00:00:01 1.Kf3 Kf6 2.Ke3 +- (1.87) Depth: 3/4 00:00:01 1.Kf3 Kf6 2.Ke3 Ke6 +- (1.77) Depth: 4/6 00:00:01 1.Kf3 Kf6 2.Ke3 Ke6 3.h5 +- (1.80) Depth: 5/6 00:00:01 1.b3 +- (2.20) Depth: 5/6 00:00:01 1.b3 Kf6 2.c4 Ke6 3.c5 +- (3.46) Depth: 5/6 00:00:01 1.b3 Kf6 2.c4 Ke6 3.c5 h5 +- (3.43) Depth: 6/8 00:00:01 2kN 1.b3 Kf6 2.c4 Ke6 3.c5 a5 4.bxa5 +- (3.56) Depth: 7/8 00:00:01 5kN 1.b3 Kf6 2.c4 Ke6 3.c5 Kd7 4.Ke3 Kc6 +- (3.46) Depth: 8/10 00:00:01 10kN 1.b3 Kf6 2.c4 Ke6 3.c5 Kd7 4.Ke3 Kc6 5.h5 +- (3.49) Depth: 9/11 00:00:01 18kN 1.b3 h5 2.c4 Kf6 +- (3.43) Depth: 10/12 00:00:01 39kN 1.b3 h5 2.c4 Kf6 3.c5 Ke6 4.Ke3 Kd7 5.Kd3 Kc6 6.Ke3 +- (3.43) Depth: 11/14 00:00:02 69kN 1.b3 h5 2.c4 Kf6 3.c5 Ke6 4.Ke3 Kd7 5.Kd3 Kc6 6.Ke3 a5 +- (3.43) Depth: 12/15 00:00:02 113kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc6 6.Ke3 a5 7.bxa5 +- (3.20) Depth: 13/17 00:00:03 192kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kd6 6.Kf3 Kc7 7.c5 Kc6 +- (3.33) Depth: 14/19 00:00:05 341kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kd6 6.Kf3 Kc7 7.c5 Kc6 8.Ke3 +- (3.43) Depth: 15/19 00:00:07 493kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc7 6.Kf2 a6 +- (3.23) Depth: 16/22 00:00:09 661kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc7 6.Kf2 Kb6 7.c5+ Kc6 8.Kf3 a6 9.Kf2 a5 +- (3.23) Depth: 17/25 00:00:15 1192kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc7 6.Kf2 Kd7 7.Kf3 Kc7 8.c5 Kd7 9.Kf2 Kc6 +- (3.23) Depth: 18/27 00:00:18 1537kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc7 6.Kf2 Kd7 7.Kf3 a6 8.Ke2 Kc7 9.Kd3 a5 10.bxa5 +- (3.30) Depth: 19/29 00:00:25 2136kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc7 6.Kf2 Kd7 7.Kf3 Kc6 8.Kg2 a6 9.Kg3 Kb7 10.c5 Kc6 +- (3.30) Depth: 20/31 00:00:35 2945kN 1.b3 h5 2.c4 Kf6 3.Ke3 Ke6 4.Kd3 Kd7 5.Ke2 Kc7 6.Kf2 Kd7 7.Kf3 Kc6 8.Kg2 a6 9.Kg3 Kd6 10.c5+ Kd7 11.Kf2 Kc6 +- (3.30) Depth: 21/37 00:01:10 6248kN 1.h5+ Kg5 2.Kg3 Kf6 3.b3 Ke6 4.c4 Kd6 5.c5+ Ke6 6.Kf3 Ke7 7.Ke3 Kd7 8.Ke2 Kc6 9.Kd2 Kd7 10.Kd3 Ke6 11.Kc3 a5 12.bxa5 +- (3.69) Depth: 21/37 00:02:19 12356kN 1.h5+ Kg5 2.Kg3 Kxh5 3.b3 a6 4.c4 Kg5 5.c5 Kf6 6.Kh4 Ke6 7.Kh5 Kd7 8.Kxh6 Kc6 9.Kg5 a5 10.bxa5 Kxc5 11.b4+ Kc6 12.Kf5 Kd6 +- (3.93) Depth: 22/43 00:03:50 19447kN Uri
This page took 0 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.