Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Again about some details of Kasparov vs. DB in 1997

Author: Hans Gerber

Date: 12:02:08 06/22/00

Go up one level in this thread


On June 22, 2000 at 12:50:15, Pete R. wrote:


>Within certain limits of course. In general I think both "real" chessplayers and
>computer chess people are in agreement regarding the state of computer chess.
>Humans are still superior overall, but it's inevitable that this must change in
>the future as computers become increasingly powerful.

I believe that the best human chessplayers will always be better than a machine
as long as chess is not 'solved' and there is a possibility to prepare with the
machine.


>  So I don't believe that
>computer chess people look down on human GMs at all, and neither do I think that
>GMs scoff entirely at computers.


My fear on the base of several observation is that computerchess people do look
down on chessplayers. Of course for the wrong reasons. But it has something to
do with the tendency to confuse the performance of the machine with one's own
capabilities...





> I think there is natural respect on all sides.


For many reasons I see no reason for sort of symmetry in the relationships.
Chessplayers are the center of interest whereas programmers and the people in
computerchess do have their own field. For all a programmer is not dealing with
chess on the same level as the chessplayers.

To make a joke, the mothers of all the GMs do neither unit and take themselves
as equally important as the GMs!


The machines are interesting players in chess, not to replace for training, but
the programmers should stay in the background and not step into the world of
chess itself. In special they should not operate their machines. It looks
ridiculous when the smart programmer is just moving the pieces in order of the
machine's "will"... Finally it becomes absurd if the programmer does believe
that in principal he himself is now playing the GM simply because he had
programmed and decided wich move the machine should play in a concrete position.
The truth is that the programmer instructed the machine to simulate the play of
a game of chess, but in almost all concrete positions today's programs play on a
higher level than their programmers would be able to.




> But if there is a particular issue such as Kasparov vs. DB then I don't see the
>harm in discussing it, even though some may hold a negative opinion of Kasparov.


The irritating truth is that on the contrary it is often argued that
unbelievable and unproven attacks against IBM and the DB team should stop...


> You seem to be saying that your fear is that such an attitude will cause more
>GMs to shun dealing with computer people.


Yes, always in combination with the incredible loss of reality some programmers
are demonstrating when it comes to talking about their creations.



>  I don't know to what extent such
>fears are valid, or where the line should be drawn in negative opinions.  You
>seem to be advocating zero-tolerance in this regard, which I think is too
>extreme.
>


Let me give some details. I find it negative if most famous programmers use
incredibly low-levelled languages to characterize e.g. Kasparov for his behavior
in 1997. This way younger enthusiasts will lose their respect for our best
chessplayers. In the case of the Dutch event with Fritz you could read such
postings here. The invention of two equally important worlds (here computerchess
and there chess, "if _they_ don't like ... then let us forget them") was the
underlying emotional attitude.



>>To speak it out I am deeply concerned about the way some important persons here
>>have commented on Kasparov in the past. I am very sure that beyond all possible
>>deviances in his personality Kasparov was treated in a very unfair manner for
>>his behavior during and after the match against DB.
>
>You are sure, but others are not. So should they not be allowed to voice their
>opinions because they disagree with you?


Of course they should!! But reality is different. Here in CCC it is possible for
certain people to stop whole discussions just by ad hominem hints... In my case
the underlying attack that I am not even the one who I am! That it would be
making no sense to debate with me at all...

This is all meant to create sort of closed community with a lot of taboos and
forbidden activities. In science however openess and universality are keywords.



>
>>Because Kasparov asked the
>>questions Hsu and his team should also and primarily have asked! This is by no
>>means a case closed. Think of the damage for certain figures if it should be
>>proven that K's thoughts were justified... Think of the damage to the American
>>computerchess community if it becomes known that their best figures violated >the rules of science. For me it is already proven.
>
>I don't want to start this discussion all over again. You seem to say that
>Kasparov was "scientifically" justified to demand printouts or whatever proof
>might be deemed valid in order to show the machine played on its own.  In my
>view, and in the view of many I think, such proof is by definition impossible.
>Printouts are meaningless.


We have complete agreement about that. At least I will believe what R. Hyatt has
explained.

Now what does it mean for the event in 1997?? Will you agree with me that if
that is the truth that the secretness and opposition in view of Kasparov's
questions simply made no sense? Isn't it telling that such a trivial fact became
the reason for the whole agitation and dissence between the DB team and
Kasparov. What are your conclusions? Do you believe that the attitude of the DB
team was justified in relation to the fact that prints had no real meaning in
the end?






> Kasparov accused the DB team of cheating without any
>proof other than his own surprise that the machine played so well at times.
>This does not constitute evidence, and the burden of proof is on the person who
>makes such an accusation.  To make such an accusation based solely on personal
>conjecture, without the realistic possibility of evidence, is simply in poor
>taste.  It's poor sportsmanship.  If Kasparov feels that the test was
>unscientific because of lack of "objective" oversight of the machine, he has
>only himself to blame because he set the conditions and signed all the contracts
>to approve the match under those conditions.  To complain after the fact just
>makes him look bad, and destroys chances to have another match in the future
>under whatever "scientific" conditions he feels would be appropriate.


I am very happy to have you and your carefully put arguments to be able to show
what I wanted to explain for some time right now.

1. Kasparov did never accuse them of cheating. Please give your evidence for the
contrary.

2. Kasparov is not a scientist. He took part in a series of experimental tests
to see what the new machine could do. However Kasparov had all the right to
believe in the seriousness of the American scientists, the DB team. He does not
bear the responsability to prove anything.

3. The innocence and readiness to cooperate is apparently speaking for Kasparov.
If he just had wanted to chicken out and to find some weak excuses for a
possible loss of the match he would have had many other medication. However he
openly told them when he felt that something strange was happening. Strange in
his mind as one of the best players in the world. My opinion: they should have
taken more serious Kasparov's questions.

4. The often repeated argument that Kasparov simply missed his chance to make
appropriate rules in the contracts and that later he should have kept his mouth
shut is itself a weak argument. For me the unpreparedness, his naive belief in
the scientists, is speaking for Kasparov not against him. We do not talk about
kustice and the law. Nobody said that Kasparov should appeal to the High Court
in New York. This is all about science and standards in science the DB team
should have respected.

5. What has science to do with Kasparov vs. DB? Simple answer: if you want to
test what your machine could do against a chessplayer you should not irritate
the chessplayer during the event otherwise you can no longer test the strength
of your machine but you are suddenly testing what this concrete human being
could be irritated with during such a match. Sounds easy but seems to be almost
not to understand for many computerchess experts. Irritation is when you at
first reply "ok you will have the prints" and then suddenly out of the blue sky
you tell him "No, sorry, we are not allowed to give you the logfiles". This is a
so-called affront. An insult. End of the story and your tests... At least in
case of Kasparov. There are other chessplayers with colder blood...

6. New matches? Yes, you are welcome. But please qualify your machine in many
events before you ask again the actually best human player! Let us see what
history might tell us about the question who is the real loser of that event.
Kasparov or Hsu and his collegues... I fear that it is computerchess as a whole.


Not that I wanted to say that you really needed such extra lessons, but we talk
in front of many spectators...


Hans Gerber



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.