Author: Tim Mirabile
Date: 15:00:00 11/30/99
Go up one level in this thread
On November 30, 1999 at 03:08:14, blass uri wrote: >On November 29, 1999 at 22:48:38, Len Eisner wrote: ><snipped> >>I think this is largely a myth. The current crop of programs are much better >>positionally and strategically that CCC people give them credit for. Otherwise, >>they could not beat strong masters at any time control. Combinations are only >>possible if you have a positional advantage. So these programs must be getting >>better positions against masters to make use of their tactical abilities. Keep >>in mind that I am not talking about GM's and IMs. I'm saying that today's >>programs hold their own in all aspects of the game against everyone below IM >>strength > >Masters and even candidate masters are superior in positional understanding and >computers can get good positions against humans because the search push them to >see good positions that they understand that they are good. > >They avoid positional blunders not because of good positional understanding but >because the search push them to see that the positional blunders are wrong >because the positional blunders lead to a bad position that even they understand >that it is bad(they have small positional understanding) or to material loss. Not only that, but a human can wind up with a bad position (with out losing material) because of a tactical oversight. Perhaps he was depending on a certain positional move being tactically unplayable for the program, and the program proving otherwise, or vice versa. The human might see the tactic early enough to avoid material loss, but not early enough to avoid having to accept a positionally bad game. No, it's too hard to separate positional play from tactics. A better tactician will have more resources available to achieve positonal goals, and this will be a big help even if it's positional knowledge is rudimentary. I think it is thins kind of effect which surprized Kasparov in his second match with Deep Blue, especially in game 2.
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.