Author: Danniel Corbit
Date: 07:33:34 07/15/98
Go up one level in this thread
On July 14, 1998 at 22:19:22, Robert Hyatt wrote: [snip] >However, I hope we don't consider "old GM" players in the Fritz argument, >because I *know* Fritz can beat Botvinnik. The game goes "1. e4 <time forfeit> >since he won't be able to make a move. :) > >I'm talking about "active GM" players that are still "GM" strength. Playing >an 80 year old GM that is now rated 2300 won't do much toward a GM norm... Even this is irrelevant. If beating a GM made a person a GM, there would be ten times as many GM's. GM's don't win *all* of their games, even against non-GM opponents. Beating a GM under non-tournament conditions means nothing. Beating a GM under tournament conditions does _not_ [reapeat _not_] make a person a GM. It simply adds points to the player's rankings according to a prescribed formula. The original statement is akin to: Fred: "Joe is a policeman." Barney: "How do you know?" Fred: "Joe has a pair of handcuffs. Policemen have handcuffs. Joe is a policeman." In other words, playing well against a GM is a necessary condition to become a GM. But it is not sufficient. Computer programs will be GM's when they are GM's. Right now, none of them are. That includes Deep Blue. Deep Blue would probably be a GM if it played in enough tournaments. On the other hand, the GM's might learn a weakness and pound the stuffings out of it.
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.