Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Double games Fritz5-Rebel8 in the ssdf list

Author: blass uri

Date: 03:54:41 11/16/98

Go up one level in this thread



On November 16, 1998 at 06:12:48, blass uri wrote:

>I found that in  games 5225-5264 in the ssdf list(by tony hedlund) between
>Fritz5(p200MMX) and Rebel8(P90)
>
>Fritz5 won by learning in games 15,33(the same win as game 3),
>In games 16,20,26,32(the same as game 8),in games 35,39(the same win as game
>11),  in game 30(the same win as game 14), in game 21(the same win as game 17)
>,in game 34(the same win as game 28)
>
>There is something not deterministic with the times that Rebel used so Rebel was
>lucky to draw in game 23(the opening was the same as game 17,21) and in game
>31(the opening was the same as game 29 that Fritz5 won)
>
>There were also double draws in the match and I do not understand why fritz5 was
>happy to repeat draws against a P90 openent
>
>It is a bug in the learning of fritz5.
>
>I think that the reason that Rebel8 could not learn is the Autoplayer of Fritz5
>and if it is the case then Junior5 should be number 1 in the ssdf list because
>without 11 wins Fritz5 should be second and not first.

I found that Renel8 lost the same game twice against another opponent
I think that the reason is that Rebel8 has no learning code.

Uri

>
>Uri



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.