Author: Mike S.
Date: 06:50:57 09/29/02
Go up one level in this thread
On September 29, 2002 at 07:52:52, José de Jesús García Ruvalcaba wrote: >He talks about his upcoming match against Fritz, among other things. >http://www.spiegel.de/spiegel/0,1518,216103,00.html SPIEGEL: "Won't you be able just to repeat games which you have won during the preparation?" (Können Sie jetzt nicht einfach jene Partien wiederholen, die Sie in der Vorbereitung bereits gewonnen haben, um erneut über den Rechner zu triumphieren?) Kramnik: "The probability that a game may be completely identical, is nearly zero. The preparation against a human is much more specific, because you know the favoured openings of your opponent." (Die Wahrscheinlichkeit, dass eine Partie absolut identisch verläuft, tendiert gegen null. Die Vorbereitung gegen einen Menschen ist wesentlich konkreter, weil man die Lieblingseröffnungen seines Gegners kennt.) --- I wonder how many people will still stick to their opinion that it's "so unfair... (blah...)" even now when Kramnik himself explains that issue. Normally you should expect that even anyone who knows (a) that there are about 6897997394 fantastillions of opening variants, and (b) the match will consist only of 4 games where K. has White and 4 with Black, should be able to reject that idea of repeating preparation games. You don't even have to know that a program may play differently for various other reasons after the opening, i.e. slightly different response time of the opponent, other engine settings... But I've read the "repetition" idea *very* often. I wonder if anybody who thought so, has ever tried to repeat games against a program. Even that simple practical test should immediatly show what was wrong with that idea. :o)) Regards, M.Scheidl
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.