Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: old deep blue's tactical mistake against kasparov

Author: Ernst A. Heinz

Date: 10:40:34 12/04/98

Go up one level in this thread


On December 04, 1998 at 06:14:38, blass uri wrote:

>3r2k1/p4bp1/5q1p/8/3Npp2/1PQ5/P2R1PPP/6K1 w - - 0 1
>
>In this position old deep blue played 29.g3 and lost after Rd5
>
>Junior5 can find Ne2 in a few hours and I think that black cannot win after
>29.Ne2
>
>When I give Junior5 the position after 29.g3 the evaluation after 10 hours is
>more than a pawn advantage for black.
>
>I think that this position proves that deep blue as a program is sometimes worse
>at tactics than top programs
>If top programs were 500 times faster they could reject 29.g3 because of
>tactical reasons at tournament time control.

No 500x times speedup needed -- "DarkThought" happens to lock onto 29. Ne2
in iteration #12 with a score of -0.25 after roughly 2 minutes on a 600MHz
Alpha-21164a.

But who knos if it does so for the right reasons ...

=Ernst=



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.