Author: Uri Blass
Date: 06:45:09 12/30/05
Go up one level in this thread
On December 30, 2005 at 09:15:18, Albert Silver wrote: >On December 30, 2005 at 08:41:56, Uri Blass wrote: > >>On December 30, 2005 at 08:37:24, Gian-Carlo Pascutto wrote: >> >>>On December 30, 2005 at 08:19:44, Uri Blass wrote: >>> >>>>On December 30, 2005 at 08:11:49, Austin A. Biberink wrote: >>>> >>>>>On December 30, 2005 at 07:57:16, gerold daniels wrote: >>>>> >>>>>>Hello Tony.Thanks for your testing. Nice to see you are testing Fruit. >>>>> >>>>>Yes, but it would be even more nice to see that SSDF would be testing: >>>>> >>>>>1. the strongest freeware engine >>>>>2. the currently strongest engine >>>>> >>>>>(1.) = Toga II 1.1a >>>> >>>>It is not clear if the strongest freeware engine is TogaII1.1a or GambitFruit >>>> >>>>The CEGT rating list has >>>> >>>>9 Toga II 1.1 2767 15 15 1375 60.5 % 2693 36.9 % >>>>14 Gambit Fruit 1.0 Beta 4bx 2740 26 26 468 52.4 % 2724 33.3 % >>>> >>>>The possible error is too small to know if TogaII 1.1 is better than Gambit >>>>Fruit1.0 Beta4bx in CEGT conditions and based on Ryan, Toga is probably weaker >>>>at long time control because of some change that is productive only at blitz and >>>>the SSDF use longer time control than the CEGT. >>> >>>Uhm, I'm not so sure. Error margins don't add linearly. From eyeballing, the >>>CEGT list gives >90% probability that Toga is stronger than Gambit Fruit. >>> >>>Where is the proof that "some change in Toga is only productive in blitz"? >>> >>>-- >>>GCP >> >>I do not know the proof because I did not learn the differences between Toga and >>Fruit2.1 but I trust Ryan about it. > >I don't. CEGT shows Toga 1.1 to be 50 points ahead of Fruit 2.1 in the 40/40 >list. > > Albert Only note that Ryan did not claim that all the changes from Fruit2.1 to Toga1.1a were counter productive at long time control. He only claimed that some change in toga is counter productive at long time control and this change is not in GambitFruit. Note that there are also other differences between GambitFruit and Toga and my only claim is that it is not clear if Toga1.1a is the best free engine. Uri
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.