Author: Franz Hagra
Date: 07:32:22 06/19/04
Go up one level in this thread
Wellknown experts and authors think about testsuits: Chrilly Donninger (Hydra) ...die rating forumla is not important to me... Stefan Meyer-Kahlen (Shredder) ... its simple to optimize a engine for testsuits and nobody will knew ... Rob Hyatt (Crafty) ...Forget about test sets predicting ratings... ...This shows that such tests are basically flawed. ... Personally, the idea of using a test set to estimate ratings is just nonsense. It's never worked for humans, it will never work for humans. If you want a wild guess of "is it 2200 or 2600?" then you might get something that comes close to answering that. But is A 2550 or 2600? Forget it... Ed Schröder (Rebel) I don't think it is possible to create a testsuite that measures a true elo rating for a chess engine. If it were possible it would already exist, if someone manages to creates one he/she should sell it, I am ready to pay good money for it. My best,Ed Sune Fischer (Frenzee) ...The problem is that the implication "higher testscores" => "stronger engine" is often false... Anthony Cozzie (Zappa) 1. Ultimately what matters is that the program plays the move on the board. If it takes 1 second or 10 seconds or 1 minute, its not that important, from a game score sense. 2. I could increase Zappa's rating on such a test by 500 points by 1 line of code change: eval += 5*kingsafety(p); in german look for: http://f51.parsimony.net/forum203977/ Best Hagra
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.