Author: Christophe Theron
Date: 17:04:52 04/30/01
Go up one level in this thread
On April 30, 2001 at 19:24:09, John Merlino wrote:
>On April 30, 2001 at 19:20:37, Robert Raese wrote:
>
>>On April 30, 2001 at 19:14:48, John Merlino wrote:
>>
>>>On April 30, 2001 at 16:27:00, Peter Berger wrote:
>>>
>>>>c.) The games are indepedent events ( again this is a simplification but looking
>>>>at the games suggests this isn't too far off either ) .
>>>
>>>Probably not far off at all, considering that DF offered a draw when it needed a
>>>win and DJ turned it down, meaning there doesn't seem to be any consideration
>>>for the match score in the programming.
>>>
>>>Has ANYBODY added anything like this to their engines?
>>
>>interesting... the machine understands the concept of a game, but not a match?
>
>This was only pure speculation, based on the one (very interesting) event, but I
>would not be surprised if no modifications were being made to the engines based
>on the match score.
>
>jm
The programs do not only have games in their scope. Learning (book learning and
position learning) takes partly care of matches.
Also, when the programmer is present he can adjust the "contempt score" (what I
prefer to call "score for a draw") depending of what's happening in the match.
But as far as I know no program knows the concept of "match". For example no
program asks for the length of the match (which is a critical information).
Christophe
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.