Author: Uri Blass
Date: 23:24:14 11/14/02
Go up one level in this thread
On November 15, 2002 at 01:29:06, Sandro Necchi wrote: >On November 14, 2002 at 12:48:11, Sandro Necchi wrote: > >>On November 14, 2002 at 07:49:14, Vincent Lejeune wrote: >> >>>On November 14, 2002 at 07:40:08, Sandro Necchi wrote: >>> >>>>On November 14, 2002 at 02:55:46, stuart taylor wrote: >>>> >>>>>On November 14, 2002 at 00:59:54, Sandro Necchi wrote: >>>>> >>>>>>On November 13, 2002 at 10:32:11, Christophe Theron wrote: >>>>>> >>>>>>>On November 13, 2002 at 01:26:19, stuart taylor wrote: >>>>>>> >>>>>>>>On November 13, 2002 at 00:41:41, Christophe Theron wrote: >>>>>>>> >>>>>>>>>On November 12, 2002 at 13:55:34, José Carlos wrote: >>>>>>>>> >>>>>>>>>>On November 12, 2002 at 13:30:49, Christophe Theron wrote: >>>>>>>>>> >>>>>>>>>>>On November 12, 2002 at 02:48:21, José Carlos wrote: >>>>>>>>>>> >>>>>>>>>>>>On November 11, 2002 at 22:08:13, Christophe Theron wrote: >>>>>>>>>>>> >>>>>>>>>>>>>On November 11, 2002 at 19:13:19, John Rice wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>>I realize these engines are not even available yet commercially, but I'm >>>>>>>>>>>>>>guessing they have been beta testing for some time. For those of you involved in >>>>>>>>>>>>>>the beta testing, or have access to beta results, which of these three engines >>>>>>>>>>>>>>do you expect to top the SSDF? With Deep Fritz 7 and Chess Tiger 15 added to the >>>>>>>>>>>>>>mix, how would you rank all five from top to bottom? Here is my guess: >>>>>>>>>>>>>> >>>>>>>>>>>>>>1. Shredder 7 >>>>>>>>>>>>>>2. Fritz 8 >>>>>>>>>>>>>>3. Chess Tiger 15 >>>>>>>>>>>>>>4. Junior 8 >>>>>>>>>>>>>>5. Deep Fritz 7 >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>>On what is based your guess? Crystal ball? Marijuana? >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> Christophe >>>>>>>>>>>> >>>>>>>>>>>> Do you use crystal ball and marijuana? Or do you think anyone who posts >>>>>>>>>>>>something you don't like is in drugs? >>>>>>>>>>>> >>>>>>>>>>>> José C. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>>It's not that I don't like it. >>>>>>>>>>> >>>>>>>>>>>It's just that rolling dices would be a more scientific method. >>>>>>>>>>> >>>>>>>>>>>But who cares about science here? It's all about impressions and fantasy. >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> Christophe >>>>>>>>>> >>>>>>>>>> Science is part of the fun. Impressions and fantasy are other part of the fun. >>>>>>>>>>None of both parts deserve the accusation of using drugs. Open your mind; >>>>>>>>>>there're more things in the world apart from science. >>>>>>>>>> >>>>>>>>>> José C. >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>>Open your mind too. Never heard about "humour" (or "humor", depending on where >>>>>>>>>you live)? >>>>>>>>> >>>>>>>>>It was an humoristic way for me to say that this prediction was bullshit. >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> Christophe >>>>>>>> >>>>>>>>Is my prediction (a few posts below) just as BS? >>>>>>>>S.Taylor >>>>>>> >>>>>>> >>>>>>> >>>>>>>Any prediction that includes non-released programs is bullshit. >>>>>>> >>>>>>It depends who makes it and how it is made. >>>>>>A general statement like this is bullshit. >>>>>>> >>>>>>> >>>>>>> Christophe >>>>>> >>>>>>Sandro >>>>> >>>>>And just wait and see if I was right! >>>>>Junior 8 >>>>>Tiger 15 >>>>>Fritz 8 >>>>>Shredder 7 >>>>>DF. 7 >>>>>And CM9K Utzinger very close by. >>>>>S.Taylor >>>> >>>>OK, here are my expectations: >>>> >>>>Junior 8 = + 60 points over Junior 7 >>>>Tiger 15 = + 45 points over CT14 >>>>Fritz 8 will be postponed to make it stronger... >>>>Shredder 7 = + 50 points over UCI Paderborn >>>>DF 7 = we know it already. My expectation after seeing some games was 35-40 >>>>points stronger than Fritz 7 >>> >>>1 Deep Fritz 7.0 256MB Athlon 1200 MHz 2763 >>>2 Fritz 7.0 256MB Athlon 1200 MHz 2742 >>> >>>only 21 points difference ... so you admit that your error margin is 180% ?? ;o) >> >>Yes, but I do not know this program as I do with Shredder and I think it will >>gain some points in the near feature as Fritz 7 will loose some. >> >>Sandro >>> >>> >>> >>>> >>>>Now you can guess the resulting SSDF list >>>> >>>>Sandro > >I mean that the final difference will be around 30 points. > >Anyway I AM NOT INTERESTED TO FIND OUT EXACTLY HOW MANY POINTS DIFFERENCE THERE >IS BETWEEN A PROGRAM AND ANOTHER. For me 10-15 points more or less means very >little. > >Today is difficult to see if a version is stronger than another and you are >pointing out that the difference is about 14 points? > >SSDF makes very many games and still leave an error percentage, so how can I be >precise 100%? Simply by giving more freedom for error in your estimates. Instead of saying that a program is 35-40 elo better you can say it is 5-60 elo better. I do not believe that you or somebody else can get a good estimate for the exact difference in an interval of 5 elo after seeing some games. 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.