Author: Dann Corbit
Date: 12:37:22 07/23/00
Go up one level in this thread
On July 23, 2000 at 10:41:06, Jorge Pichard wrote: >I just Installed YACE 0.21 but decided not to include the Nalimov Table Bases. >Can somebody give me a good estimate of how stronger will YACE play with the >Nalimov? Look at the tablebase hits for a series of games. Each tablebase hit (not just probe) returns perfect knowlege. And if you have (for instance) 5 pieces left on the board, then tablebase files will cause perfect play if you have all of hte 5 piece files. On the other hand, they will make no difference at all, in the early game. Actually, they might make it play a tiny bit worse, because memory that could have been used for hashing is now used for cacheing the tablebase files. Overall, it seems that they don't add a huge amount of strength to programs in general. But if you make it to the deep endgame phase (fairly rare) it will have a very large impact there. Opening books are far more important than endgame tablebase files. If you can create a huge opening book without defect it will add a tremendous strength boost to your program. I't's the 'huge' and 'without defect' that are a problem. But I did say "if..." ;-)
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.