Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: IBM's chess chip approaches grand-master status

Author: Keith Ian Price

Date: 13:22:20 08/22/98

Go up one level in this thread


Hsu seems to admit that if Kasparov had been given a chance to study DBs games,
he would have trounced it, all the while denying it:

"Conventional thinking, Hsu said, has held that the evaluation speed is a
sufficient condition for a good chess-playing machine. If you can evaluate
enough combinations of moves to play farther ahead than your opponent, you
should win. But Hsu warned that this is a misleading assumption."

"'Such a strategy can work in tournament play, where no one has a chance to
study your games and identify your weaknesses,' Hsu said. 'But in match play,
your opponent will study your history. If you have mistakes, a grand master will
study them, find the underlying weakness and drive a truck through it. Speed is
important, but in match play it is more important to have few weaknesses and
weaknesses that are hard to exploit.'"

He goes on to say that that is why they made the evaluation of DB error-free.
But if that's what he did, why did he not let Kasparov see any of the games with
Joel Benjamin. The laughingstock reason he gave for this is pretty weak IMO.

kp



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.