Author: Will Singleton
Date: 17:22:42 10/19/02
Go up one level in this thread
On October 19, 2002 at 15:37:00, Louis Fagliano wrote: >On October 19, 2002 at 13:17:27, Will Singleton wrote: > >>On October 19, 2002 at 12:28:20, Fernando Villegas wrote: >> >>>What would be a serious match in your opinion? >>>I ask you this because too many times I have seen that good results for >>>computers are at once relativized with some kind of "this is not serious" kind >>>of argument. On the contrary, if the GM gets the upper hand, then at once it is >>>considered a proof of "how much superior they are". >>>fernando >> >>Hi Fernando, >> >>Hard to say how to do a serious match. The trick is to get the GM motivated to >>win. I don't believe a champion GM would agree to a winner-take-all computer >>match, but that might do it. Or if the outcome of the match was to decide the >>official world champion, that might be good enough. >> >>Perhaps in the future when computers get a bit stronger, the humans will have no >>choice but to play serious matches. Not now. >> >>I don't know if you were listening to the chess.fm commentary, but the IM >>commented that Kramnik controlled the entire match, whether by crushing Fritz in >>the early going, or losing by blunders, or taking the gm draw. Fritz was never >>really a factor, and the clear implication was that a motivated Kramnik could do >>very much better than a drawn match. >> >>The silver lining for me is that the game of chess appears to be particularly >>and stubbornly difficult to master for computer science. The game goes on. >> >>Will > >I don't think you say that you can "control" a match by losing by blunders. >Otherwise I could play Kramnik and lose 8-0 by leaving pieces hanging in every >game and you can say I "controlled" the match. That's exactly what I meant. If you lose 8-0 by hanging pieces, you certainly did control the match. Any opponent would beat you. > >It's true that Kramnik played in a much higher creative plane than Fritz, but >then again the same can be said for the 1978 World Championship Match where >Korchnoi played more creatively than Karpov but lost the match because he >blundered more often. The bottom line is still wins and losses. > >And why do you say Kramnik wasn't motivated? Playing a move like 19. Nxf7!? in >game six sure looks like motivation to me. If Kramnik had a gun to his head, he would never, ever play that move. That's what I meant about motivation. It was pure speculation for the fun of it (he actually admitted as much), because he knew the outcome of that game wasn't so important. > >Human GM's and computers have different strengths and weaknesses and what's neat >about this match is that both the strength and weaknesses of human GMs and >computers were present in different games of the same match. Games 2 and 3 show >the human GM's mastery over the computer in long range planning in closed >positions with the queen's exchanged. Games 5 and 6 show the computer's mastery >in tactics. > >You can say that in games 7 and 8 when Kramnik didn't want to take serious >chances to play for a win that Fritz wasn't able to do anything about it, but >that same can be said for any other strong GM playing Kramnik that they can't >beat him either if he chooses to play quietly and safe. So it can be said that >Fritz is as least that strong. Nothing of the sort. Kramnik demonstrated clearly that he could win at will against the computer. Do you remember? Everyone was saying the computer was like a child against a master. Fritz was on the fritz. Should have been <insert program here> playing. Remember? Then Kramnik decided to go soft. The last game clinched it, an early gm draw. I don't mind that, particularly, but I wish they'd just admit it. Perhaps it's reflective of a European view of the world, I don't know. Just play it safe, give up, make nice, no fights. Put nothing on the line, nothing is worth risk, safety is paramount. Shake hands and pat each other on the back. Take the money, screw the fans. Will
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.