Computer Chess Club Archives

Search

Messages

Subject: Re: Number of games to test

Author: Frank Schneider

Date: 22:15:54 03/08/02

Go up one level in this thread

Hi,

On March 09, 2002 at 01:01:02, TEERAPONG TOVIRAT wrote:

>Hi,
>
>How many games I've to test between 2 versions. I think it varies as
>the score ratio.

that's right, I guess you need about 500 games to be 90% confident that
program A is better than program B if A scored 52%. But if A scored 60%
30 games are probably enough.

> Suppose program A beats program B 4-1.Can I say
>A is superior to B? Or the number is too small?
>

The number is too small.

Frank

>Thanks for any comment,
>Teerapong

• Re: Number of games to test TEERAPONG TOVIRAT 23:46:42 03/08/02
  • Re: Number of games to test Christophe Theron 08:03:38 03/09/02
    • Re: Number of games to test Kurt Utzinger 09:09:54 03/09/02
      • Re: Number of games to test Christophe Theron 13:37:29 03/09/02
  • Re: Number of games to test. Here is your table ! Leen Ammeraal 07:59:11 03/09/02
    • Re: Number of games to test. Here is your table ! Martin Schubert 09:13:36 03/10/02
  • Re: Number of games to test Frank Schneider 07:50:43 03/09/02
  • Re: Number of games to test Edward Seid 00:59:35 03/09/02
    • Re: Number of games to test Miguel A. Ballicora 06:45:56 03/09/02

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.