Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Let's do a Chessmaster-test! CM against the world!?

Author: Harald Faber

Date: 00:13:10 12/17/98

Go up one level in this thread


On December 16, 1998 at 11:40:38, blass uri wrote:
>>To have a closer idea of the match strength of Chessmaster what about testing
>>against the other top programs? With this series we can also see a difference
>>between CM6000 and CM6555 and will certainly get proof how strong CM6k
>>(probably) is.
>
>I think it is better to change every time only one parameter in CM6000 because
>otherwise maybe some of the changes are productive and some are counter
>productive and we do not know what was productive and what was counter
>productive

So far the CM6555 settings are known, they are known to be good so why test and
compare step by step? And another problem is which parameter to change?

>I suggest to compare between CM6000 and CM6000 with selective search=10

Everyone is free to do so. It was only a proposal.

>I am not going to do games but test them in some practical hard positions that
>computer programs need a long time to find the right move(positions from my
>correspondence games and positions from games between computers)
>Uri

Positions where elementary knowledge is requested are really interesting for
programs but it doesn't replace real games.



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.