Author: Thorsten Czub
Date: 03:38:34 08/22/05
Go up one level in this thread
On August 22, 2005 at 06:24:44, Majd Al-Ansari wrote: >To me it sounded like a even headed gracious report ... no midlife crisis here. aha. for me it sounded different. >I still think that Shredder will kick any programs butt in a tournament with >many games. we will find out about this later. They are close. > I am looking forward to trying out the new Zappa in a long >tournament with normal even books. I am still not convinced it is the strongest >engine, even though I thought it played great. But it really was never in any >trouble out of the openings and was extremely well prepared. I don't want to >belittle Zappa's achievement at all, I think they were great. But a WCCC title >doesn't automatically mean you have the strongest engine. of course. but i am happy that zappa and fruit and toga and glaurung and spike and and and appear. they bring new wind into this crusted business. >I believe one of the main reasons Junior and Shredder's performance seemed >sub-par was because if you go to the engine room on Playchess, almost every >engine is playing one of these two. So it is easy to play thousands of >automatic games and see which lines these 2 programs play poorly and cook books >against them. Now Zappa is the new champ and people will probe for its >weaknesses once the program is available for public download. right. >By far the most impressive program in my opinion was Fruit. It is using a >revolutionary type of evaluation that in many ways seems kamakazee. Can you give an example ?! > Yet it >still manages to prove that its evaluation was correct. I think it will be the >most difficult engine for other engines to deal with because most engines are >not even considering moves that it chooses. why is the program different concerning evaluation. give an example. > This is why it can get away with >dramatically inferior hardware and no EGTB and books off the shelf that are not >tuned to its weird evaluations and still almost win the WCCC. A Fruit with EGTB >and multiprocessor support with a Playchess engine room tested book will be a >formidable opponent. i had fruit in a longer tournament with s9 and j9 etc. http://f50.parsimony.net/forum200336/messages/4971.htm
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.