Author: Dann Corbit
Date: 14:55:20 07/16/02
Go up one level in this thread
On July 16, 2002 at 17:22:03, William H Rogers wrote: >I agree that a 20 move deep opening book would be a lot greater than one that >was on 10 moves deep. In fact, last year one game in a tourneyment was won by a >program that did not ever leave its opening book. I personally think that that >is totally rediculus as the book included opening, midgame, and end game all in >one. I have maintained for over 10 years now that opening books should be >limited only to the first 10 moves as they are considered the openings, after >that you are in mid game, etc. How do you rate a program that might be only >rated at 1200 elo when its opening book complete destroys the top 5 or so >programs rated at over 2600 elo? The use of opening books should be limited in >tourneyment games so that the true strength of the programs can be seen, not the >extensiveness of the books. What is sold to the public on the other hand can >have anything goes so long as the option to controll is there. >I know that most of the programmers here believe that anything goes, but in my >opinion we are trying to build and test the best "chess engine" not a >combination of other junk that has been included. >Of course if you already have an enourmous book with your program, then you will >not agree with me but want to keep the winning combination. >I am sorry if I get a little riled about this subject, but to me chess >programming is just that "chess programming" not playing a game totally from >book without ever using your engine. I disagree. If they can solve chess with the book alone, then that is what they should do. It would be a pretty big book, I think. Also, I think that the quality lines should go as deeply as possible. If this extends to the tablebase files, then that is superior to stopping early.
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.