Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Now I think there must be something wrong here

Author: Tony Hedlund

Date: 23:36:06 07/02/02

Go up one level in this thread


On July 03, 2002 at 01:02:56, Harald Faber wrote:

>On July 02, 2002 at 15:19:24, ERIQ wrote:
>
>>On July 02, 2002 at 14:12:15, Uri Blass wrote:
>>
>>>On July 02, 2002 at 13:25:16, ERIQ wrote:
>>>
>>>>:(
>>>
>>>It is possible that Hiarcs8(A1200) learned from losing against Junior7(K6-450)
>>>and plays better against Junior7(A1200).
>>>
>>>Junior7(A1200) could learn nothing from the games of Junior7(K6-450) because
>>>these players are 2 different players.
>>>
>>>I did not analyze the games to find if it is the case.
>>>
>>>Uri
>>In that case how can this match be fare to junior ??
>
>This does not concern only Junior 7 in this case. Many new programs play vs.
>opponents who have learn values from some hundred games. IMO this falsifies
>several results. It may equalize after 500 games, but up to there the new
>program has disadvantage.

This is not true. If a programme have played 500 games, each copy have played
perhaps 80 games. And the new programme have an advantage of a upgraded
openingbook.

Tony



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.