Author: Christophe Theron
Date: 10:56:17 12/11/00
Go up one level in this thread
On December 11, 2000 at 08:13:22, Mogens Larsen wrote:
>On December 11, 2000 at 07:08:56, Thorsten Czub wrote:
>
>>so the draw-score in this score stopped df from seeing
>>the point.
>>therefore it lost the game.
>
>No, you can't make a conclusion like that. What would an evalution of -2 have
>helped. Would it have used more time? No, not according to the game. Approx. the
>same amount of time used when behind. Would it choose other moves? No, it still
>selects the move it believes to be best. Change plans? No, programs doesn't make
>plans.
That's not true. Programs DO plan.
>The only real chance of changing direction IMO would have been a fail low,
>forcing it (presumedly) to search for an alternative. Transposing a consistant
>score up and down wouldn't change much, if anything at all AFAIK.
You would be surprised how little changes in the evaluation give very different
behaviours.
>>more examples needed ? analyse enriques games...
>
>If the purpose is to establish the obvious fact that programs misevaluates
>certain positions then it's a profound waste of time unless you're writing one.
I'm not in agreement with Thorsten and his point, but at least he has a point
and tries to demonstrate it.
I suggest that your post is at least the same kind of waste of time unless you
are writting a chess program yourself.
Christophe
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.