Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: About head or tail (was Upon scientific truth - the nature of informati

Author: Daniel Clausen

Date: 06:50:34 07/16/00

Go up one level in this thread


Hi

On July 16, 2000 at 07:59:44, Alberto Rezza wrote:
>Yes. So the problem is: how much confidence do we need in the chess programs'
>strength? Are 300 games not enough?

Depending on the confidence you want, you need a different number
of games. [The exact formula is an exercise to the reader. ;)]


> If we were to apply such standards to human players, we would have to
>conclude that when a player gets a GM title from FIDE we really cannot say
>whether he is of GM strength

This is a different case. A human player gets a GM title from FIDE if
(s)he reaches 3 GM-norms. (where GM-norm is defined as <whatever>) This
clearly defines what a GM is.

Now, what you mean with "someone is of GM strength" is simply not defined.
Of course it's logical to assume that this is someone who has a similar
strength of "some GMs". But since a GM is not defined with his strength
but with "reaching 3 GM-norms", the definition is only vague.


>A program whose results are good for 3 GM norms has GM strength - and that
>should be all.

Sounds like a good definition to me. The fact though "that program X
beat the sh** out of GM Y in the famous tournament at Z just around the
corner" doesn't. :)

Regards,
 -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.