Author: Mark Young
Date: 10:49:12 10/07/98
Go up one level in this thread
On October 07, 1998 at 12:49:12, blass uri wrote: > >On October 07, 1998 at 12:10:23, Mark Young wrote: > >>On October 07, 1998 at 11:58:09, blass uri wrote: >> >>> >>>On October 07, 1998 at 09:49:33, Robert Hyatt wrote: >>> >>>>On October 07, 1998 at 05:04:15, Jouni Uski wrote: >>>> >>>>>On October 07, 1998 at 04:31:48, Bruce Moreland wrote: >>>>> >>>>>> >>>>>>On October 07, 1998 at 04:23:59, odell hall wrote: >>>>>> >>>>>>>I am interested in knowing whatever Rebel 10 will top the swedish rating list. >>>>>>>And if Die Hard Sceptics such as Robert Hyatt will concede that Rebel 10 Is >>>>>>>indeed Grandmaster Strength? >>>>>> >>>>>>The question is not whether Rebel or Hiarcs or Fritz or any of the others is of >>>>>>GM strength, but of whether they all are. >>>>>> >>>>>>bruce >>>>> >>>>>And the answer is: YES. >>>> >>>> >>>>If the answer is "YES", please produce some 40/2hr games where the >>>>programs have beaten (or broken even) with GM players. The last four >>>>games I personally watched ended up with the GM's ahead 3-1 in points, >>>>and the last time I checked my Elo performance rating for that, I get >>>>GM-200 for the rating... which is still in the 2400 range, where I've >>>>said they are all along... >>> >>>The GM's were better than average GM's >>> >>>I know that Junior drew with 3 grandmasters before the match against yudasin >>> >>>I know that Zungzwang did a good result in a tournament (performance more than >>>2500 and I think it played against GM's in the tournament) >>> >>>I think that we should look at all the games that were played by computers >>>against GM's last year in tournament time control. >>> >>>I am not sure if computers are of GM strength but they are very close >>>and maybe part of them are GM's >>> >>>I am not impressed by the ability of computers >>>They do too many mistakes that I will not do so the reason that computers are >>>close to be GM's is not that computers are strong but GM's are weak. >>> >>Many things come to mind when I think about GM's....but Weak is not one of them. >> >>By the way what is your record playing any of the top 3 or 4 programs at 40/2hr, >>from the way it sounds it must be pretty good with all the mistakes they do, >>that you do not do. > >I admit that I am weaker than top chess programs > >I did not try to play against top programs in 40 moves per 2 hours >I am almost sure that I will lose because I will not go to the positions that >they are weak but when I look in games between programs I see sometimes that >they do mistakes that I will not do. > >When I use computer programs to help me to analyze my correspondence games I am >angry about mistakes that they suggest me to do. > This is the point a want to make. People should not fixate just on the computer’s weaknesses. Yes computers do have weaknesses, but so do humans and human GM’s. What you should look at is how the computer programs perform, as in do they win or lose and to whom are they winning and losing to. Because computer programs also have great strengths, that counter balances their weaknesses. Therefore, it is better to give performance greater weight then just fixating on weaknesses that can taint your judgement about a programs real strength. >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.