Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Result based standards Vs. Subjective standard for computer GM status.

Author: Fernando Villegas

Date: 10:45:18 06/16/01

Go up one level in this thread


Hi:
You are right in the main of the argument, results matters, but then we have a
problem: which results we have? The pool where GM's gets his scores is huge in
number of members, games and time. Hundreds of thousands of games giving meaning
to the entire structure. Chess programs has games againts GM - at 40 in 120-
only in the order of dozens. So I would say there is no enough statistic
evidence against or in favor of GM tesis. All We have are some games won, some
lost, that's all.
Also you are right that GM's are just human beings prone to mistakes, but if you
are coherent with your own argument, "only results matters",  then to mention
the mistakes or failures or ugly moves they sometimes commit is out question for
your cause. Fact is that in terms of results they are far superior to other
guys.
My guess is: there are three programs with weak GM level in 40-120. But I have
no facts to support that. Only a guess. Programs are Gambit Tiger, Deep Fritz
and Shredder. Probably also Junior.
Fernando



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.