Author: Dana Turnmire
Date: 10:04:13 10/02/02
Go up one level in this thread
On October 02, 2002 at 11:42:35, John Merlino wrote: >On October 02, 2002 at 09:02:06, Dana Turnmire wrote: > >>On October 02, 2002 at 04:45:06, Jorge Pichard wrote: >> >>>He said that what matter is to develop the most unusual tastics possible. >>>This statement could have been acceptable if the actual 8x MP system and program >>>was not given to him before the match, but everybody know that what he will >>>actually do is to replay games that he already knows what the outcome will be, >>>since he already had plenty of time to play many games specially certain >>>openings and he already had played different moves out of the opening to see >>>what deep fritz would play etc... >>> >>>http://www.chessbase.com/newsdetail.asp?newsid=530 >> >>Larry Christiansen had almost a month to study Chessmaster 9000 but it didn't >>seem to help him that much. Since Kramnik doesn't know what openings Fritz will >>play how does he have that much more of an advantage than GM Christiansen? Is >>the learning feature overrated in chess programs? If not would it not prevent >>repeat wins by Kramnik? > >CM9000 was out for a month before the match, but Christiansen never got the >product. He told me a day or two before the match that he was very familiar with >CM6000, and he just assumed that CM9000 was "much stronger". > >Also, Christiansen didn't know which personalities he was going to play against >in games 1-3, so preparation against unknown opponents was pretty much >impossible. He DID draw the Chessmaster personality, so perhaps his preparation >was an advantage in the final game.... > >jm I suppose it comes down to what is considered to be fair for each side. Unfortunately no one can agree on that. In the example of Kasparov vs Deep Blue, Kasparov was not allowed to review any games his opponent had played yet the IBM team had a a strong GM (Nick DeFirmian) inputting the openings for Deep Blue. After studying hundreds of games by Kasparov it would be easy for a professional chessplayer of DeFirmian's strength to play moves that he knew Kasparov was unfamiliar with or uncomfortable playing. I never could understand why it is unfair for the human to study games played by his upcoming opponent but not unfair for the computer team using a strong human GM to guide the play into areas of the human's weaknesses. Will there be a human with a database of Kramnik's games playing the openings for Fritz? No one has ever explained to me why the learning feature in the modern programs don't compensate for the fear Kramnik will kill it using the same lines of certain openings. Is this feature strictly for computer vs computer games?
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.