Author: Thorsten Czub
Date: 00:17:14 10/31/05
Go up one level in this thread
On October 30, 2005 at 18:15:51, Maurizio De Leo wrote: >Regarding different parameters....it is possible that they improve the default. >However from CEGT testing of CM 10000, Shredder and Fruit, it seems that >different settings are usually worse than default or at best in the same league. tuning is part of the development of an engine. there are engines coming on the market that are well tuned, and others that are not well tuned. >After many many settings of CM, not one has proven to be stronger than the >default in a clear way. that might be true. >I think that different styles can improve a program at most 20-25 elo points, so >toga II 1.0 would still be weaker than fruit 2.2. i don't understand why it should only be 25 in general ?! It depends on the switches you can do. if you have an engine that has many interesting switches the programmer never tried, it can be very much, if you only get 3 or 4 unimportant, even the programmer has tried but was not convinced, it can be only 25 possibly. IMO there is no GENERAL estimation you can do. i have the feeling that toga has much potential.
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.