Author: Eelco de Groot
Date: 15:46:49 03/09/01
Go up one level in this thread
On March 09, 2001 at 16:59:41, Thorsten Czub wrote: >one major things chessbase invented that is that good that others >HAVE to copy it. > >thanks to chessbase. thanks to ed and christophe. Hi Thorsten, Yes, I agree. I think there is also much potential in position learning and the way Ed has programed that wit CAT in Rebel Century 3, actually. If you can combine that with all the analysis that Dann has been collecting for the Chess Analysis Project and if people could send in findings of their own, with 3-best lines (I don't think that is possible with CAT at the moment) you would have three good lines already for a position and the first three moves in very good sorted order too which does not happen in alpha-beta, it should be possible to speed up the search for the next iteration, make it possible for a program to vary it's play by choosing a good second best move every now and then, give comments about other lines in a position, lines to avoid etc... And it could act as a repository of chess knowledge, copyright questions of published grandmaster analysis lines would have to be considered of course. Robert Hyatt said position learning was first mentioned by David Slate and Tony Scherzer in ICCA Journal, I mention it because somebody asked about women in computer chess and Tony and Linda Scherzer were another couple, with their program "Bebe". I believe Tony Scherzer doesn't live anymore. Anyway it doen't seem terribly difficult to program such a thing, position learning, wouldn't it be something for Chris Whittington to put into Tal? If Chris ever decides he wants to try something new for his program again...
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.