Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: What convinces YOU of the strength of a program???

Author: stuart taylor

Date: 17:45:10 10/17/04

Go up one level in this thread


I feel that if it shows a certain kind of consistency in results, even through a
relatively small sample, then it will show to be very strong, possibly even
moreso, after more extensive testing, and this is what I mean.....

If it gets a plus result, even a small one, against ALL competitors (even only
computer competitors).

If it hardly ever loses, even though draws are boring, but the very fact that it
doesn't lose should indicate that it doesn't have many weak points, and I would
normally think that if it hardly loses against a selection of top program after
150 games or more, then it might climb even much further throught the next 1000
games or more. I can't promise this, but.....

It gives an impression that if I want to consult that particular program in any
position, that what ever THAT computer shows in analysis might be as goodas any
other program would show, and I don't need to consult other programs because
this one might have "missed" something.

I hope all this is interesting to CCC readers, as what I have written grips me
very strongly, when I try to decide which program to buy.

I also tend to wait until a few test results are out, and if a small sample
would be tested according to the way I would like to see, this (the above) is
what I would be looking for.

All this, combined with what I wrote in a nearby post, (in this thread) below
regarding what someone Peter.... wrote about Tiger.

S.Taylor



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.