Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Shredder evaluations

Author: Ingo Althofer

Date: 10:51:09 07/06/04

Go up one level in this thread


On July 06, 2004 at 13:38:31, Robert Hyatt wrote:

>On July 06, 2004 at 13:17:59, Ingo Althofer wrote:
>
>>Stefan Meyer-Kahlen writes a diary (in German language)
>>during his days in Ramat-Gan.
>>
>>On the very bottom of the page you find a link to
>>a download of Shredder games from rounds 1-3 (pgn),
>>with Shredder evaluations included.
>>
>>Very interesting for instance, how early Shredder
>>realized that the game against Crafty would give
>>a draw. And against Jonny search depth 30(!) was
>>not enough to see Jonny's drawing trap.
>
>Shredder could have _easily_ lost the game vs Crafty.  It had a definitely
>inferior position.  I've noticed that many programs are "overly optimistic" in
>"self evaluation" and that it is often wrong ...

Right. The evaluations are Shredder's subjective view of the "world",
sometimes correct, sometimes wrong - and often optimistic (see the +1.92
at moves 38 and 39 versus Jonny).

For me it was especially interesting to see, how far they differed from the
evaluations of me "old" Shredder 7.04 on a "slow" 900 MHz mono-AMD.


>But if you look at
>the game, you'll see what I mean.  That is the kind of game I prefer to see
>Crafty play as it was going to be hard to lose with an early edge, and it had
>serious chances to win, particularly if the opponent makes a single misstep.
>
>>
>>http://www.shredderchess.de/telaviv.html

Ingo Althofer.



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.