Author: Robert Hyatt
Date: 06:52:59 01/21/00
Go up one level in this thread
On January 21, 2000 at 01:36:53, Amir Ban wrote: >On January 20, 2000 at 22:23:15, Michael Neish wrote: > >>On January 20, 2000 at 02:18:51, Ed Schröder wrote: >> >>>You misunderstand. It's the DB main-line for 36.axb5 what made Kasparov >>>suspicious. In this main-line DB sacrificed 3 pawns for no direct win >>>but for a dangerous looking king attack, all very human-like. Kasparov >>>could not believe his eyes (he still can't) and started the accusation >>>human intervertion took place as he could not believe a computer was >>>able to produce such a (super) main-line. >>> >>>So this whole issue is NOT about the move 36.axb5 but about the asthonising >>>main-line DB produced. >> >>Ed, out of interest, do you happen to know what was the line that DB came up >>with? Was it the same as what was played in the actual game? >> >>I just did what a lot of you must have done: I set my pet program thinking on >>the position where DB chose 36. axb5. In this case I'm using HIARCS 7. This is >>probably very naive, but at 8-ply the score for 32. Qb6 and 32. axb5 was almost >>the same. At 9-ply though it favours Qb6 quite clearly (about +1.7, compared to >>+1.0 for axb5). Maybe the key points in the search lie at deeper plies? Can >>anyone tell me what a program should be able to spot in order to determine that >>axb5 is the best move? And has it been proven beyond doubt by grandmaster >>analysis that axb5 is best? >> > >Qb6 wins a pawn and axb5 doesn't. There's a lot of analysis done on this >position, to great depths. Much of it was done here, on various threads, and you >can search for it in the CCC archives. Kasparov was aware of much of it during >the game, and certainly a day later with his helpers. What would he have done if we had substituted Crafty on a 16-way alpha for DB? Because Crafty would definitely sacrifice a pawn for an attack in the right positions, without any hope of getting the pawn back inside the search horizon. Or it would sacrifice the exchange in the right positions without being able to see that it can recover the material later. Is sacrificing material suddenly a sign of cheating? IF so every GM on ICC is going to go public with the fact that Crafty cheats. Because it has done such positional sacrifices for years. Counting pawns isn't the only way to evaluate a position, of course. At least for humans. Why not for computers??? > >We don't know if axb5 is the best move, and it doesn't appear that it is. It >seems to be a distinctly human move, giving up on a clear tactical chance in >order to keep up the pressure. The same goes for the next move 37.Be4 but there >the problem is not so great because the alternative does not win material. > >Amir > > >>Not trying to prove a point either way. Just interested. >> >>Cheers, >> >>Mike.
This page took 0.01 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.