Author: Uri Blass
Date: 10:01:31 12/20/01
Go up one level in this thread
On December 20, 2001 at 11:06:48, Sune Larsson wrote: >Congratulations to your win Uri! >This is a triumph against...?? >Well, It's a triumph for the computerchess vs ?? >Anyway - Happy Christmas and Good New Year! > >Sune The opponent is Dov Rozenberg and I believe that he simply used computers for less time. It also seems to me that he did not use his head at least in this game. He got into the 3/4 final of the world championship so it gives me the impression that a computer that is used only for about 30 minutes in every move is enough to get into the 3/4 final of the world championship but it is also possible that he treats these games more seriously. Here are example for typical computer error of my opponent when in part of the cases computers change their mind after a long search. Note that I am not sure that these are errors in all the cases and it is only my opinion [D]2r2rk1/p3q1pp/bp6/2bp1p2/4n3/1P2P1P1/PB1N1PBP/1QRR2K1 w - - 0 23 Rozenberg played Nf1 when I believe that Qa1 or Nxe4 fxe4 b4 (not Rxd5 Qf7) are better After 23.Nf1 Be2 white is losing a pawn or gets a positional demage. Rozenberg chose to lose a pawn by 24.Rxd5 and I do not criticize him for that(the only way to keep equality in material was 24.Re1 Bxf1 25.Bxe4). The game continued 24.Rxd5 Qf7 25.Re5 Rcd8 26.Rcxc5 bxc5 27.Qe1 Bxf1 28.Bxf1 Kh8 and now I believe that rozenberg again did a typical computer mistake and played 29.f3 when I believe that Qc1 to keep the pair of bishops was better [D]3r1r1k/p4qpp/8/2p1Rp2/4n3/1P2P1P1/PB3P1P/4QBK1 w - - 0 29 Some moves later white had a losing position but I found the best move that programs do not find [D]7k/p3q1pp/8/2p2P2/1P6/2Q1P1P1/r6P/5BK1 b - - 0 34 I played Qd6 when programs prefer Qf8 that probably is also winning but less convincingly(some programs change their mind to Ra4 after a long time and this move seems even worse than Qf8 to me). After 34...Qd6 my opponent replied with a typical computer error 35.bxc5? when Bd3 at least prevents me to trade the queens and seems to me better from human perspective. The pawn endgame is a result of the sacrifice Rxd5+ [D]7k/6pp/8/2rB1P2/p3P3/3K2P1/8/8 b - - 0 46 I doubt if it is the best move and a3 is more normal for humans when white will have to sacrifice the bishop for a pawn but my programs like it so I decided to play it. It is also possible that a later move of me 48...Kf7 that does not prevent g5) was not optimal but yace after many hours of analysis said mate in at most 31 when it was unable to see the mate after h6 probably because h6 does not lead to trading pawns and to tablebases. [D]6k1/6pp/8/3P1P2/p5P1/3K4/8/8 b - - 0 48 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.