Author: Dann Corbit
Date: 16:30:20 10/14/04
Go up one level in this thread
On October 14, 2004 at 16:59:54, Uri Blass wrote: >Frank Quisinsky did a tournament between programs > >GreenLight was leading with 6.5/8 >Dragon had 2.5/8 > >see http://f27.parsimony.net/forum67213/messages/3474.htm > >He discovered that both programs played without opening book so he repeated >their game > >Results here: > >http://f27.parsimony.net/forum67213/messages/3496.htm > >GreenLight 4.5/8 >Dragon 2.5/8 > >You can see that the programs scored 2 points more without opening book. > >I can say about it that it is possible that opening book has negative value not >because of bad moves but because the programs simply does not know how to >continue after book when it could know to continue better without book. I do not think that there is enough data to make a choice from 8 games either way. I would run at least 300 games, and I would also look at what happened when the programs fell out of book (e.g. first few evaluations) It would also be interesting to see shallow wide books, narrow deep books, wide deep books, and things of that nature employed. It would also be worthwhile to test professional books, Arena books, and things of that nature.
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.