Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Gulko's Tactics

Author: Dann Corbit

Date: 15:17:56 03/22/02

Go up one level in this thread


On March 22, 2002 at 17:36:34, Otello Gnaramori wrote:
>On March 22, 2002 at 14:48:15, Dann Corbit wrote:
>
>>I cannot see his mind.  I suspect you are wrong, though.  I also think he will
>>win a game, now that he has the right idea.  In fact, it is too bad that he did
>>not start out with anticomputer style from the beginning.
>
>C'mon Dan ,
>Don't you think that Boris Gulko had many opportunities to train his marvelous
>(sic..) anticomp style against comps in the comfort of his sweet home ?

I don't have any idea.  But in all the preceeding games, he did not try to use
anticomputer tactics.  Clearly, he did not prepare for this match with
'classical' anticomputer tactics.  I think he prepared on a slower machine, and
got a bit stunned.  Being a smart person, he did a bit of research (or someone
told him about it) and found some better ways to beat the silicon brute.  Of
course, it is pure speculation on my part.  I have talked to other GM's who have
had high profile matches against machines and sometimes they do not make special
preparations at all and some of them are completely ignorant of the things that
cause problems for computers.

>>I think he has made the right choice.  Play anticomputer chess against a
>>computer.  Astonishingly, there are still a lot of GM's that don't know about
>>simple things like long pawn chains against computers.  I think that they will
>>watch these high profile matches and figure it out, though.
>
>Long pawn chains are good only for draws, if you want to win you have to try
>some other anticomps techs like stonewall or trojans (good luck).

Long pawn chains are good to lock up the formation.  Sometimes, you can engineer
it in such a way as to hold the key in your hand, so that you can open the gate
if you want to.  The purpose (in the case of trying for a win) is to delay any
actions that are decisive to more than the 7 full moves that a computer can
easily fathom.  If you can make it ten full moves away on a crowded board, the
computer simply won't see it.



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.