Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: electric GM already here super gm 5 years

Author: Robert Hyatt

Date: 14:17:36 12/13/00

Go up one level in this thread


On December 13, 2000 at 13:35:29, Pete R. wrote:

>On December 12, 2000 at 16:19:27, Robert Hyatt wrote:
>
>>I would totally disagree.  You won't _ever_ find a GM that is willing to
>>sit down and play a game against a totally unknown opponent.  Because they
>>like to prepare for the opponent directly.  Preparing for a computer is
>>simply another type of preparation.  And saying that doing so is not legitimate
>>means that _no_ GM game is legitimate, since they do this for each other all
>>the time.
>>
>>The real problem is that GMs that are not in the top 100 give programs fits on
>>ICC all the time.  I won't mention names, but it is common.  Because they tend
>>to play the opponent, which is perfectly normal.  I don't think a GM would care
>>_which_ computer he has to play, but he would certainly want to know that he
>>is playing a computer (I think computers are more similar than most would give
>>them credit for being).
>
>I agree that it's normal to play the opponent, however this is a much larger
>advantage in the case of playing against a program because there is such a stark
>difference in their strengths and weaknesses. They have a choice to face a
>machine gun or a pea-shooter if they can steer the play to certain positions.
>GMs play simuls all the time,

I would agree.  But these games are nothing more than "skittles" games... IE
they are not played in a tournament where winning wins money...


> so it is not the case that they always know their
>opponent. In that case they simply rely on their "pure chess strength", if we
>can define such a term, meaning just the sum of their positional knowledge,
>tactical ability, and experience, including familiarity with opening lines.
>While knowledge of the opponent is important, if you remove that knowledge,
>there is still a spectrum of pure chess strength, and the top GMs will still be
>the top GMs if they didn't see their opponents in a match. That would be easy to
>prove with such a match, and I doubt anyone would disagree that there would be a
>strong correlation of the results with current ratings. I believe that if a
>computer participated in such a "blind" tournament where players simply move
>pieces on a screen, its strength would be very obvious. I don't dispute that
>there would be a larger difference if the players knew they were up against a
>machine, but that's the difference between actual intelligence and a piece of
>silicon. I'm simply defining chess ability as something separate from the extra
>advantages of preparing for a specific opponent.

If your hypothesis is (a) a GM would likely do better when playing a computer
if he _knows_ he is playing a computer; then I would agree.  I would add that
(b) a GM would likely do better against _any_ opponent if he knows who he is
playing beforehand.  I also believe the inverse is true.  If the GM didn't
know he is playing a computer, he would almost certainly do worse.

By the same token, if a frog had pockets, he could carry a gun and not have
to worry about snakes.  :)




>
>It would also just be funny to have Kasparov give a simul somewhere, where a
>number of people get moves fed from a computer, and see how he does. :)

This has been done somewhere.  I don't know if it was Kasparov or someone
else.  But whomever it was, he got pissed about 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.