Author: Tony Thomas Karippa
Date: 21:17:11 02/25/06
Go up one level in this thread
On February 25, 2006 at 22:57:56, Uri Blass wrote: >On February 25, 2006 at 21:33:17, Tony Thomas Karippa wrote: > >>On February 25, 2006 at 21:20:15, Tony Thomas Karippa wrote: >> >>>On February 25, 2006 at 19:20:36, Uri Blass wrote: >>> >>>>On February 25, 2006 at 18:51:15, Tony Thomas Karippa wrote: >>>> >>>>>On February 25, 2006 at 18:22:19, George Tsavdaris wrote: >>>>> >>>>>>On February 25, 2006 at 18:14:20, Tony Thomas Karippa wrote: >>>>>> >>>>>>>On February 25, 2006 at 17:50:25, Uri Blass wrote: >>>>>>> >>>>>>>>On February 25, 2006 at 17:25:05, Tony Thomas Karippa wrote: >>>>>>>> >>>>>>>>>Why are you so surprises? Yes, I know it was a connection problem, but Zappa >>>>>>>>>isnt weaker than fruit, at least not when running on a quad. >>>>>>>> >>>>>>>>We do not know it because nobody can test zappa on a quad against latest fruit. >>>>>>>> >>>>>>>>Uri >>>>>>>Didnt Zappa ran on a quad in WCCC 2005 and in Leiden?? >>>>>> >>>>>>Do you call one game against Fruit testing......? >>>>>>Also that Fruit played there, was not the latest.... >>>>>Dont bring that argument, Zappa wasnt the latest either. >>>> >>>>It is another reason that we cannot compare the levels between zappa and fruit. >>>>It is possible that previous zappa is better than fruit but latest zappa is >>>>weaker than fruit. >>>> >>>> The last three times >>>>>they played against each other Zappa one twice and draw one game. Isnt that >>>>>clear evidence that they are on the same level to say the least. >>>> >>>>No >>>>player a1 beated player b1 >>>>player a2 beated player b2 >>>>player a3 drew with player b3 >>>> >>>>Is it an evidecne that player a4 is at least in the same level as player b4? >>>>No. >>>> >>>>We simply do not know. >>>>We did not claim that zappa is weaker than fruit and of course we have no >>>>evidence for that. >>>> >>>>You claimed that zappa ia not weaker than fruit and we criticized this claim >>>>because you also have no evidence for that. >>>> >>>>Uri >>>No no no, that isnt the case >>>1)player a1 defeated player b1 >>>2)player a1 defeated player b1 >>>3)the next game was a draw >>>I dont see why you are using different names for the same engines(Based on >>>versions?) >>>To me its enough evidence to say that they are within the same elo range. When >>>was the last time Fruit won a major tournament?? Have you seen the elo list >>>posted by Jouni Uski?? It indicated that Zappa is one of the top engines. Also >>>Zappa has the advantage of the hardware. >>In case you missed it here is the link >>http://www.talkchess.com/forums/1/message.html?488168 >>I heard that the new Movei is stronger than Fruit, when do we get to test it?? >>:-) :-) > >I can promise you that the new movei is stronger than fruit1.0 > >I thought to release it next week after seeing some positive results but at this >point of time I am disappointed by it's rating in the CEGT list so I decided not >to release it. > >I already sent a new version to the CEGT(352) but I am not sure if it is strong >enough to be released. > >I believe that there is a room for improvement of more than 50 elo by changing >parameters. > >I decided to document every parameter that is possible to change and explain >exactly it's meaning and hopefully it may help me to discover some bugs and fix >them and make the thing stronger. > >there are parameters that I programmed a long time ago and I do not remember >their exact meaning. > >Uri I dont think I really care about a big elo jump, I am satisfied as long as the new version scores about 5 points higher than the previous version in a 250 games test.:-)I only play 2/0 games. Some people say it is pointless, but heck I am not bothering anyone by posting results all the time. Also considering that I have a pentium 2800MHZ, I think its as good or equal to some of the 30 minute games played by testers on 500MHZ pentium. I would say you should go ahead and release it, its been a while since we have gotten a new version of your engine. Tony
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.