Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Why?

Author: Gian-Carlo Pascutto

Date: 04:27:20 04/15/02

Go up one level in this thread


On April 15, 2002 at 03:35:07, Slater Wold wrote:

>[D]2r4k/1p3p1p/1pn2q2/1N1p1p2/8/PP2Q2P/5PP1/2R3K1 w - - 0 26
>
>Why in this position is it not perfectly clear that taking the pawn (Qxb6) is
>the obvious move?  Instead I've noticed several programs wanting Qf4, a4, and
>other moves that to me, just seem to waste time.
>
>I've spoken with one GM, (2550ish) and he agreed that Qxb6 is the only move that
>he would emphasize on.  (It was a quick discussion, no detail.)
>
>This is move 26 of game one in the first match between Deep Blue and Kasparov.
>Deep Blue played Qxb6 and later won the game.
>
>
>
>**I've tested only a few "top commercial" programs.  Most want Qf4, a4 and
>others at first, and around 1:30 want Qxb6.  My question was, why isn't this
>obvious from the start?**

A program with a speculative evaluation might think, gee, after the pawn
capture, black's attack might just work! So better not grab it.

A program with a nonspeculative evaluation might think: Donuts! Yummy!
Eateateat.

So, all those 'great attackers/new padadigm-whatever programs' will now fail to
simply win a pawn.

This is the other side of all those speculative wild attackers. It works both
ways.

--
GCP



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.