Author: Louis Fagliano
Date: 12:54:00 02/06/03
Go up one level in this thread
On February 06, 2003 at 13:54:08, Bob Durrett wrote: >DISCLAIMER: I am not a chess grandmaster and do not have any first hand >knowledge about what GMs do in the privacy of their own homes. > >Anybody who has played around with chess engines and chess databases, will have >experienced many cases where the chess engine wants to play a new move in the >opening. By "new move," is meant a move which does not appear in the games of >the database. These new moves may or may not be any good. > >Believe it or not, there was a time when chess engines, and electronic >databases, did not exist at all! [This may come as a shock to many.] > >It seems reasonable to assume that GMs have always devoted a lot of time to >developing their opening repertoires. What do they do, specifically, in the >process of developing their repertoires? > >One thing seems obvious. They come up with good new moves, never played before, >[i.e. TNs] and then prepare lines from that move which will give them an >advantage over their opponents. Ideally, the unsuspecting opponent would have >to refute the TN over-the-board with the clock running. A good TN would be one >which presents very difficult problems for the opponent, so that it would be >very likely that the opponent would go astray in the ensuing positional &/or >tactical complications. > >It also seems reasonable that GMs would be aware of the possibility that their >opponents might prepare TNs. One must surmise, then, that GMs not only prepare >TNs for their own use but also spend some time in identifying possible TNs which >might be used by their opponents. I would expect that each new opening line a >GM plans to use would be exhaustively analyzed by that GM, both using >conventional methods and using computers. For example, the GM's computer might >be tasked with the overnight analysis of a line as if it were a game. The chess >engine would be given plenty of time to analyze each move. > >New moves for the opponent identified by the computer, including new "second >best" moves, would then be studied by the GM in preparation for coming events so >that the GM would be prepared for potential TNs played by the opponent. > >As an example, it was noted that Kasparov sprung a TN on DJ in the fourth game, >and DJ then surprisingly sprung a "counter-TN" against Kasparov, effectively >turning the tables on Kasparov. Kasparov then spent a long time trying to >figure out what to do and finally had to settle for a draw with White. It was >Kasparov who had to try to find a refutation to his opponent's TN over-the-board >with the clock running. Actually quite funny! > >TNs really do make a huge difference in the serious games of the top GMs. > >When preparing his TN, Kasparov had to look at all plausible responses by his >opponents. It is remarkable that DJ seems to have come up with a refutation to >Kasparov's TN. But was it really the chess engine that did that? Could it be >that the people who made DJ's opening book had already identified Kasparov's TN >and prepared for it, providing a refutation in the opening book? > >Bob D. If that were so, then with the move in the computer's book, it would have replied instantly. I don't think Deep Junior replied instantly with 10... Bxh2+.
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.