Author: Robert Hyatt
Date: 07:01:59 06/24/01
Go up one level in this thread
On June 24, 2001 at 06:45:40, Mogens Larsen wrote: >On June 23, 2001 at 22:48:28, Robert Hyatt wrote: > >>This _absolutely_ happens. In a WC match, after a sealed move/adjourned game, >>the GMs go to bed, their "seconds" stay up all night analyzing the position, >>and the next morning they show their GM the analysis on how to win, or draw >>the position. The GM commits it to memory and off he goes to the board to >>play those moves he has _never_ seen before. > >You wrote that the GM was shown the analysis. Thereby he or she is aware of the >general idea with the move. That suggests some kind of inevitable comprehension, >not just memory alone. There will always (IMO) be a certain degree of >understanding of the move, not just the isolated move. The degree is arguable of >course. > >Mogens. Let's go back to DB vs Kasparov match 2, game 6. Kasparov claims "he mixed up the move order". Does that suggest comprehending the opening or does it suggest simply memorizing a particular set of moves in a particular order? Suppose a computer took its opening book, played thru each one, did a real search at every position, then minimaxed the entire thing so that its evaluations were used to choose from the book lines? Early Crafty versions did this. Cray Blitz did it too. IE they didn't just blindly follow the book moves, they actively participated. Or in the current case of crafty, with book random 0, it does a search over the set of known book moves and plays the one _it_ thinks is best, not the one that is most popular or whatever.
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.