Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: M-Chess Pro7 : strength ??

Author: Don Dailey

Date: 13:18:07 12/26/97

Go up one level in this thread


On December 26, 1997 at 14:40:50, Jeroen Noomen wrote:

>On December 26, 1997 at 13:26:09, Don Dailey wrote:
>>
>>I also find this quite distasteful.  I think it's becoming a war that
>>they feel they have to fight just to stay on even ground.
>
>As the openingbook maker of Rebel I can only say that I would be
>delighted when we would turn back to normal theory and let the programs
>find out what happens. More and more I am detecting anti-computer lines
>that have nothing to do with playing strength. The only point is to cook
>the opponent, leading to a reaction from that opponent etc. Personally
>I don't like this development at all. The programs are strong enough to
>deal with various openinglines, so let THEM play and not the opening
>experts.
>>
>>It seems to me that the large variety style of book has some advantages
>>and also (maybe) some drawbacks.   First of all, it doesn't prevent
>>book cooking, but I believe it will require significantly more effort
>>on the part of the book "cooker."  But eventually the cooker will
>>traverse every book end node and will have decided where to vary.
>
>Correct. But don't we all agree that this should come to an end? I fear
>that in a few years the main part of work of the openingbook expert is
>to detect 'cooks' and to repair those lines. I find this disgusting and
>this has NOTHING to do with opening theory, playing strength or
>whatever.
>
>Best regards, Jeroen Noomen


Hi Jeroen,

I completely agree with you in everything you said.   I see this as
really being 2 different problems:

  1) How strong is the engine really?
  2) How good is the opening book?

Most of want the best program they can get (without considering the
opening book.)   We feel a little cheated when we thought we were
buying a strong chess program but we were really buying a strong
opening book database!   We want to play our own games, experiment
and  try out endings etc,  in short we want the best calculating
engine we can get.

But the "book cookers" have a point of view also.  It has always
been a part of Master chess to prepare hard for your opponents.
They are doing nothing that all strong humans do not do.  Or are
they?

In principle I agree with them but NOT in practice.  I know they
are only defending themselves against others who do this too and
it's difficult to blast them too much for it.  But this is a grey
area thing and you must look at their motives.  Are they legitimately
trying to improve their chess program or are they trying to get
as many wins as possible?  The two things are different I believe.

Of course if you ask them, they will say of course they want to
improve the overall chess play.   And it's hard to defend this.
But if they had a choice between getting cheap wins and getting
to the top of the list, or actually making the program stronger
but not getting there, which would they choose?   I think this
is the real crux of the matter, how genuine is the improvement?
Certainly a well tuned book makes a program legitimately stronger
as it does with humans but a fudged book designed to take unfair
advantage does not.

This is really a moral issue because they have legitimate arguments
that can not be easily refuted.  And also they are in the difficult
position of being forced to compete against others who may also
do this which puts them in a very difficult situation.

I like the idea of eliminating the "book factor" in some way.
But this is a real difficult problem too since it's certainly
legitimate to tune your book to your programs style.   In some
ways the book really is an integral part of your program just
as it is with people.   I do not think there is an easy solution
to this although there are some interesting possibilities.

-- Don













This page took 0.03 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.