Author: Drexel,Michael
Date: 19:37:12 10/09/03
Go up one level in this thread
On October 08, 2003 at 11:49:53, James T. Walker wrote: >On October 08, 2003 at 06:31:38, George Tsavdaris wrote: > >>On October 08, 2003 at 05:40:40, Torstein Hall wrote: >> >>>http://www.chessbase.com/newsdetail.asp?newsid=1229 >>> >>>Is realy programs getting stronger compared to humnan GM's? >>> >>>(In my view they are) >>> >>>Torstein >> >>"Although computers obviously must be improving in recent years, the strongest >>humans seem to also be improving at about the same rate." >> >>a) The ELO performance in the graph, of the strongest human player(Kasparov), >>in the last 6 years is almost a straight line that stays the same. >>b) The SSDF maximum ELO performance in the graph is inceasing in the last 6 >>years, although the way SSDF handles it's rating list makes this less important >>for belying his above statement. >>c)Every year, we have an increase at the plies a chess-computer searches at a >>given time(due to improved search techniques and hardware speed) and also >>programmers add more knowledge at chess-computers. Humans can't improve so >>much every year to compensate this one(or two) plies and in fact they improve >>only 5 % - 10 % in relation with comps. >> >> Due to the above 3 reasons and especially the c) , his statement seems to >>me wrong. >> It's ridiculous to say that computers are not getting stronger compared to >>GM's (Mr Jeff Sonas didn't(yet?) said that). And it's ridiculous to say >>that: "I don't believe that computers will inevitably surpass the top humans". >>This thing is inevitable. >> Of course this has nothing to do with todays strength of computers, as indeed >>may be lower, than that of top GM's. > >I think that the basic "idea" of his article is correct in that computers have >not made much progress vs top humans lately. But he seems to ignore the real >reason. The top GM's are taking computers more seriously now. They are >learning how to play vs computers which they didn't take seriously several years >ago. Computers are getting stronger but GM'a are adjusting. The problem is >they cannot adjust forever. The computers will surpass them all (humans)very >soon. Not in the near future. >I think if humans were to play computers now without knowing the opponent was a >computer then they are already behind the curve. That is if they played them >straight up like a normal GM game. This is NOT GM chess. In GM chess you can almost always study the games of your opponent before the game. Of course GMs always try to benefit from weaknesses of the human opponent. Their anti-computer tactics are holding on >for now. As programmers clean up the holes in anti-computer tactics then humans Clean up the holes? Are there just some bugs to fix? :) >will fall by the wayside. >Jim Kasparov and Kramnik did not play classical Anti-Computer chess. This is just wrong. Michael
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.