Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Learning programs -Old ones that have learned compete with new programs?

Author: Bertil Eklund

Date: 02:47:13 07/07/99

Go up one level in this thread


On July 07, 1999 at 00:47:30, Ed Schröder wrote:

>On July 06, 1999 at 18:33:40, Bertil Eklund wrote:
>
>>On July 06, 1999 at 18:22:07, John Wentworth wrote:
>>
>>>It would be interesting to know if programs that have been "learning" over the
>>>past few years, if they could compete with programs that have not "learned"
>>>anything yet. For example, MChess 5.0 which came out several years ago was
>>>supposedly able to learn, if someone has been continually using it and it has
>>>been learning, could it compete with a program like Hiarcs 7.32 or Fritz 5.x ?
>>
>>Hello!
>>
>>No!
>>
>>But it don´t loose the same game twice. This is the case with programs with no
>>learning, especially Rebel that often play the same variation over and over
>>again. Rebel has some primitive book-learning but it don´t work in narrow lines,
>>where there isn't any allowed alternatives. Mchess have a good learner but it's
>>not easy to save a game from a rotten opening-line, which is the problem in many
>>cases.
>
>That is because it is proven you have tested Rebel WITHOUT the learner.
>There is nothing wrong with Rebel's learner. There was something wrong
>with your testing.
>
>Ed Schroder
>
>Hello!

What kind of learner is that? Rebel saves the games including the result i.e.
0-1. Other programs is aware of the result if the game is SAVED including the
game-score. In 99% of the cases the game is saved including the correct result.

Regards Bertil SSDF
>>Regards Bertil SSDF



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.