Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF CM8000 - Deep Fritz: 0,5 - 3,5 Now: 3,5- 15,5

Author: Jonas Cohonas

Date: 04:40:13 04/04/01

Go up one level in this thread


On April 04, 2001 at 06:36:47, Daniel Chancey wrote:

>The CM program is not weak   but it's not a good anti-fritz program.   Also the
>personality used is slightly inferior.      CM has a small book, no book
>learning, no tablebases.   Much of Chessmaster was focused to help human players
>inprove their game.      When the best Chessmaster personality is found,  a
>match between Deep Fritz under same conditions is necessary.
>
>Castle2000

Why all these excuses as to why CM8 plays poorly, you are right it was intended
to be a learning program and it does a heck of a job at just that, but then
again it takes away from the playing strenght, which becomes more and more
obvious. The idea of testing a program in the first place is to test the
strenght of a prog, and you dont see Deep Fritz needing anti CM8, CT14, GT2,
Nimzo8 etc in order to be a dominating factor in the world of cpu chess!
When i tested Gandalf against CM8 gandalf absolutly crushed CM8, so did
Crafty1714, yace and wbnimzo2000b, CM8 simply failed to win a tournament against
any of the above mentioned, this was at long tounament time controls with a
litlle fischer time added to compensate for CM8's poor handling of time.
On my T-bird 1.3ghz 512 ram.
If you manage to find a setting that will match the play of Deep Fritz (which i
doubt) or any other top prog, then it would suggest that CM8 play very
unbalanced if it needs a different setting for each top program.
If CM9 will support book.lrn, tb's etc. have a better book and is bugfree to the
point where it pays to buy the prog, then i think it might be up there with the
other progs!

Regards
Jonas



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.