Author: Bruce Moreland
Date: 17:26:47 05/27/01
Go up one level in this thread
On May 27, 2001 at 16:43:27, Larry Clark wrote: >Has there been any analysis done to estimate the effect of endgame tablebases on >the programs' performance? Obviously, the programs play much better endgames >with the tablebases, but has this in any way been quantified into an approximate >ratings increase? What would you estimate the approximate ratings increase >would be with tablebases vs. no tablebases? Nobody has quantified it. Sometimes the increase is zero. For instance, in my CCT3 games over the past two days, none of them were affected by endgame databases. Other times, you'll get a clear 1/2 point shift because you have them. I've never tried to figure out how often this happens, but obviously it does. It may matter how strong your opponent is, because the chance of reaching an appropriate ending would seem to be greater if the two opponents are evenly matched. Sometimes you get a 1/2 point shift the other way, but this does not seem to be as common. I think it is interesting that so many people use the Nalimov stuff without any concern that at some point while the program is playing, it is not their program actually playing. That idea bothered me enough that I wrote my own stuff. bruce
This page took 0.01 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.