Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Computer Evaluations of positions

Author: Daniel Clausen

Date: 01:40:41 04/09/02

Go up one level in this thread


On April 08, 2002 at 18:49:22, Andrew Williams wrote:

>The point about negative being good for black isn't universal. For my program
>(and I think some others) a positive score is good for the side whose turn it
>is. So if it's Black's turn and the score is -10.77 in my program that would
>mean that White is winning by about a Queen.

Concerning ICC: Kiwipete (you know, the guy with that famous movegeneration test
position :) solved it nicely by saying "+whatever by white/black". It's clear
for everyone and it's trivial to implement... I don't know why no-one else uses
this scheme (afaik). Sinan (my chess engine) will definitely use this (hopefully
not copyrighted) scheme when whispering eval on ICC. Of course it will be a
ridiculous eval, but at least everyone will know for which side it is. :)

Or I could use + for side to move, when it's the engine's move, + for white when
in analyzing mode - except on Tuesday and Friday afternoon. :)

Sargon



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.