Author: James T. Walker
Date: 07:59:43 04/10/00
Go up one level in this thread
On April 10, 2000 at 00:37:24, Vincent Vega wrote: >Does anyone have any data on the strength increase/decrease due to use of >n-piece tablebases by programs? Note that decrease is theoretically possible >because programs that blindly use tablebases could lose an opportunity to make a >drawn position into a win against an inferior opponent. I think such >experimental data on a value of tablebases would be very interesting. Hello Vincent, I have recently tested Hiarcs 7.32 vs Chess Tiger at game/5 min. I played 200 games with Hiarcs using tablebases (Approx. 4 gig). I then played 200 games with Hiarcs having no tablebases. The difference in performance vs Chess Tiger was 40 points advantage using the tablebases. Hiarcs 7.32 ran on a PII-333 and Chess Tiger ran on a K6-3-450. I was going to repeat the test using Fritz 6 vs Chess Tiger when I discovered that Chess Tiger was losing many identical games to Fritz 6. I was hopeing that Christophe would give us a solution to that problem before continuing the test. Nothing more heard on Chess Tiger's problem. It would also be interesting to test two programs both having tablebases and then test again with only one having tablebases. I may do that this week with Fritz/Hiarcs. Jim Walker Jim Walker
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.