Author: Robert Hyatt
Date: 09:46:50 04/11/02
Go up one level in this thread
On April 11, 2002 at 10:34:13, Uri Blass wrote: >On April 11, 2002 at 09:41:10, Robert Hyatt wrote: > >>On April 10, 2002 at 18:04:54, Uri Blass wrote: >> >>>On April 10, 2002 at 17:57:25, Robert Hyatt wrote: >>> >>>>On April 10, 2002 at 17:28:34, Amir Ban wrote: >>>> >>>>>On April 10, 2002 at 01:24:55, Joshua Lee wrote: >>>>> >>>>>>http://www.chessbase.com/newsdetail.asp?newsid=255 >>>>>> >>>>>>Kramnik may not know anything about programming but when he says Fritz was >>>>>>suggesting objectively better moves than Deep Blue Played.....For someone of his >>>>>>level of play to say such a thing. DB might have had extra knowledge than fritz >>>>>>or any other program doesn't who really knows.....there is not enough >>>>>>disclosure. Still i would like to see if anyone has found a position from either >>>>>>match inwhich DB played a move that is out of Commercials grasp. >>>>>> >>>>>>I am pretty sure this is like the thousandth message that asks this question. >>>>> >>>>>He was stating the obvious. Fritz6 and Junior6 were already stronger than DB, >>>>>and I said so at the time. >>>>> >>>>>Amir >>>> >>>> >>>>Obvious to _whom_??? >>> >>>It is easy to find people that it is obvious to them (for example Vincent >>>Diepveen and Kramnik) >>>I also believe that Fritz7 on good machine is better than Deeper blue. >>> >>>> >>>>not anyone in "reality" I don't think... >>>> >>>>Or did I miss where Fritz6 and Junior6 beat kasparov in a 6 game match at >>>>40/2??? >>> >>>It proves nothing. >>>Junior was also leading 5-0 against Fritz. >>>Does it mean that Junior was better than Fritz? >>> >>>Uri >> >> >>So what. Junior and Fritz are _programs_. Were junior leading Kasparov 5-0 >>it would be _so_ much more important... > > >The point is that if 5-0 does no tell you which side is better >than 6 games with minimal result also does not tell you much >and it is even possible that the loser is 200 elo better >and it is impossible to know that Deeper blue is better >Than Deep Fritz based only on the result. > >The performance of other programs against humans also suffer >relative to deeper blue or deep thought >from the fact that humans have the possibility to buy the program. > >Not all of them use it but it is enough that >part of them do it in order to reduce the performance of programs. > >Uri 5-0 between two nearly equal _programs_ doesn't say much. 5-0 between a program and a human says more... The closer two players are to being equal in every way, the more games you need to separate them...
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.