Author: Christian Koch
Date: 07:41:37 02/18/05
Hi all :-), The CEGT websites are up and running. - http://www.koch-in-holm.de/CEGT/ - http://www.husvankempen.de/nunn/ CEGT (Chess Engine Grand Tournament) is currently in progress with a lot of double round robins by Heinz van Kempen and me with a time control correspondent to AEGT conditions (that means 40/40 repeated adapted to 2 Ghz CPU via Crafty benchmark), so that we can combine the rating lists and have a better comparison commercials vs. amateurs. Tournaments are running on five faster computers for weeks already, but we wanted to wait until we could add the new Shredder 9 and also replaced ProDeo 1.0 after more than 100 games by the new and stronger version ProDeo 1.1. There are also other aims we have. Statistically relevant results with many games, that is why we do not use tournament time controls. These games here anyway with repeated time control sometimes last hours. Additionally we want to have a comparison of books. So Heinz is taking Nunn 2 Suite and maybe later Noomen positions and I take shorter books and Nunn 1 Suite. The reason not to take own books is that we do not want to have only games that are already in late middle game or even early endgame stage after the book moves and avoid also own books as they are in many cases adapted to certain opponents. Furthermore we want to give statistics over the time in Excel sheets, showing the performance of each engine with open, half-open and closed games and some other things, for example: - which engine performs better with Nunn/Noomen or short books - difference in ELO between best professional and best amateur engine, where at this point it is not clear if this will be ProDeo, List, Aristarch, Fruit or the new SOS hopefully coming after Paderborn - which commercial engines does best against the amateurs and which amateur engine does best against the commercials - how many amateurs are doing better than the worst commercial in CEGT (probably this will be Patriot) - which engine performs best in open, half-open and closed games - which engine won the most short games (tactical strength) - which had the most draws We also want to see how many games are needed to have more or less the same range of engines in our respective overall crosstables, if this will happen over many hundred games per engine. Best regards Christian
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.