Author: Timmay
Date: 12:49:16 11/20/03
Go up one level in this thread
On November 20, 2003 at 14:03:37, Dana Turnmire wrote: >It seems obvious that computers get lost in closed positions and anti-computer >play. Even low ranked players have had some success against top programs using >unorthodox play. How in heavens name can one of the greatest players in history >not prepare openings that are known to give computers problems? It seems any >grandmaster could lock up the position where long term strategy is the key. Is >this not possible against Fritz? If not then Kasparov should not be criticized >because when it comes to tactics the computers rule. Well, computers are getting much better against the stonewall dutch for instance. Also, playing anti-computer chess is contrary to his active aggressive style (the reason he's number one). If something isn't natural to you, you won't play it right. Game 3 was a natural ebb of the game to Garry (and coincidently happened to be a discovery into what Fritz currently doesn't understand). c5 is the best way to attempt to refute the strong Meran-type ideas of a6, dxc4 Bxc4, b5, a6, Bb7, and c5 etc. That's a move he'd play against any human grandmaster. Beginners may occasionally be successful with special anti-computer strategies (though I question how often that is truly the case), because they do what they're told to do with little regard for the actual quality of the position. In other words the position doesn't bother them. However, Garry always is concerned with his quality of position (the reason why the deep blue match was so poor for him, he ignored the quality of his positions and "did what he was told to do" by computer experts). In this match against X3D Fitz he chose openings he thought the computer wouldn't do well in, but each of the openings were openings he'd play against strong humans! In other words, he chose a certain opening out of his current repertoire that Fritz would be least likely to do well in. For instance, the g4 move in the semi-slav is strong because its a battle for the f1 bishop tempo. Instead of giving in to black with Be2?! or Bd3?! losing a tempo against the bishop leading into the Meran just so white can castle kingside, he delays the f1 development and sacrifices the g4 pawn (which can't be taken without consequence) and castling queenside, never allowing the Meran deployment for black. My point is g4 is not something he prepared only for computers, it's a strong, novel idea in that opening altogether. Also, if you remember, Kramnik didn't play closed, locked up positions either. He played in his positional comfort zone. You never saw Kramnik or Kasparov in a bad position in all these recent computer matches (with the exception of Deep Blue 1997). Speaking of that I can't believe John Fedorowich thought the computer was much better in Game 2. He had a very bad assessment of the game that game (normally he's real good). Black had a clearly better position (which Anand also agreed with mind you), and as Garry mentioned the h4 Qc4 Nd4 Bxd4 exd4 idea was absolutely critical to using the benefits of his position. I agree Fritz didn't press as hard as a good human in that game, but I would still prefer black in any case. The point of me talking about this being that even in the game he lost he had a fine position. So the top Grandmasters are learning that they only need to make small adjustments, and make choices from their normal repertoire to battle computers.
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.