Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: GM Smirin vs 4 comps - Match Predictions

Author: Roy Eassa

Date: 14:42:23 04/16/02

Go up one level in this thread


On April 16, 2002 at 17:19:13, martin fierz wrote:

>
>>I am still amazed that people claim the programs are low rated (although the
>>number of people claiming this gets smaller), the combined TPR for the 16 games
>>will be very strong.  GM Smirin would need to win this match by two points just
>>to bring the comps down to 2650 and some of the programs will be above that
>>number and a couple below.  This hw is not the fastes either, reinforcing that
>>rating of 2700+ by top programs on fastest hw.
>
>you seem to misread the claims intentionally. i for one would never say they are
>"low rated". oh, and this is rapid chess, don't forget...
>computers have HUGE weaknesses, and if properly exploited, they play like
>beginners. if your own rating is like 1000, you cannot see that. if it is only
>about 1700 like hyatt's, you can. a program which plays a game like shredder vs.
>smirin is just not GM strength. it is 3000+ in tactics and 2000- in positional
>play. it may perform at GM level in tournaments, and in matches. but if a
>program evaluates dead drawn positions like the ones i posted under "strange
>fritz evaluations" as winning, how can you call it "GM strength"?
>GM strength to me would imply that it understands that such a position is a
>draw. even you can understand that it's a draw! so how come the mighty computers
>can't?
>
>aloha
>  martin


Martin, I think it's much easier for them to debate a misrepresentation of
others' statements than the actual statements.  :-)



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.