Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: more examples for search-based stupidity

Author: Robert Hyatt

Date: 14:19:49 06/14/98

Go up one level in this thread


On June 14, 1998 at 14:45:32, Thorsten Czub wrote:

>On June 14, 1998 at 10:51:37, Don Dailey wrote:
>>We could simply use this definition:
>>
>>    Tactics:  Things we can directly calculate.
>>
>> Positional:  Things we must guess at.
>
>>- Don
>
>Strong chess players "calculate" positional stuff as good as tactical
>stuff.
>They can show with evidence and facts that the move was a positional
>blunder.
>You cannot call tactics SAVE and positonal things GUESSING.
>
>Positional stuff can be transformed into material.
>Tactical advantage is material.
>Positional stuff is material transformed in something else.
>
>Like Energy. Tactics is physic-laws on materia, positional is energy in
>those processes.
>
>But tactics are not more accurate than positional.
>Both are as accurate.
>Thats what MY opinion is over the years studying people like Seirawan,
>Bronstein, Kosashvili or Kohlweyer/Schaefer fighting against the machine
>I operated, or i watched them killing the machine.
>
>How can you say a positional advantage is less real, just YOU are unable
>to COUNT - better - bean-count - it accurate !
>
>YOU are the problem. You cannot measure energy as good as measuring
>mass.
>Thats YOUR problem. Buy a new scales.


It's trivial, and your argument against it is weak.  A backward pawn is
a common positional consideration.  But *if* I can search deeply enough
to see that (A) I can trade the pawn (B) it is useful to block a file or
diagonal for a critical instant or (C) something else, then your
positional
(mis)understanding whill haunt you and help me... because you are using
a
general principle that says "backward pawns are bad" while I ignore that
and simply search to see how it ends up.

Tactics are *absolute* to a program.  If it sees mate, it is *mate*.  On
the other hand, positional terms are simply scientific will-ass guesses,
that may or may not work out as we expect.  That's true of *any*
positional
term.  Even something nearly invincible as having two rooks *and* a
queen
on your opponent's 7th rank.  And getting mated yourself because none of
them can get back at the critical instant to defend what needs
defending...



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.