Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF and question for Tony

Author: Harald Faber

Date: 02:17:48 07/24/01

Go up one level in this thread


On July 24, 2001 at 04:34:07, Uri Blass wrote:

>On July 24, 2001 at 04:23:15, Harald Faber wrote:
>
>>On July 24, 2001 at 02:44:38, Tony Hedlund wrote:
>>
>>>On July 24, 2001 at 01:18:47, Harald Faber wrote:
>>>
>>>>So you just need to take care that DeepFritz wins the last two games and we get
>>>>the standard-Fritz-result. ;-))
>>>
>>>I'm writing this from my job. Tonights game was won by Deep Fritz, and it had a
>>>small advantage in the opening in the second game.
>>>
>>>Tony
>>
>>Hehe, so DeepFritz wins again and remains unbeaten in long term matches. :-))
>>Isn't it strange? It suggests superiority which in fact is not present...
>
>It is present in long mathces.

Yes, maybe, but not in chess. :-)

>It is possible that other program may have superiority in the opening book but
>after Fritz learns to avoid bad lines it can win matches.
>
>Uri

So SSDF honours the best learner and not the best chess program. ;-)
For extreme cases it may lead to DeepFritz playing 5-10 opening lines for black
and for white and win matches. Isn't this scenario fantastic?

Actually, what is/was the original intention of the SSDF?

For Tony: Do you reset the weights in the opening book before you start new
matches or do you take over the learned move preferences from former matches?




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.