Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: New Test Suite of "Positional" Positions

Author: Bob Durrett

Date: 12:34:58 12/06/02

Go up one level in this thread


On December 06, 2002 at 13:18:31, Mike S. wrote:

>On December 06, 2002 at 09:07:30, Bob Durrett wrote:
>
>>There is a book, published recently, which is devoted to "positional" positions.
>>One hundred and fifty positions are given with the solutions.
>>I would publish the positions here, but that would probably be illegal since
>>the book is copywritten.
>
>I think the "naked" positions themselves, without the book's comment (or the
>games they come from) could be published. There's a website which contains such
>data for many chess books for download (the webmaser even respects, when
>publishers do not want the pure chess data to be offered there).
>
>http://www.uni-klu.ac.at/~gossimit/c/book.htm
>
>There, I found a file for "Positional Sacrifices" by Neil McDonald (see near the
>top of the page, if you scroll down a bit).
>
>I thought, if the positional ideas are related to a *sacrifice*, these could be
>especially suitable as computer test positions, because they may have the
>necessary "testing character" of the solution move.

A true positional sacrifice [a sacrifice intended only to obtain positional
advantage] is a very interesting special case.  True sacrifices to obtain
non-positional advantages, i.e. tactical, might also be considered a special
case, too.  How chess engines handle positions involving either kind of true
sacrifice is a very interesting question.  What is needed is a quantitative
measurement of the performance of the top chess engines when faced with such
positions.  This measurement would have to be done using a large diverse sample
of true sacrifice positions, each of which would be quite different from the
others.  Perhaps a suite of 100 to 300 positions might suffice.  Perhaps the
test could be run first with positional sacrifices and then with tactical
sacrifices.  The sacrifices should be true sacrifices, where the outcome is
beyond the normal search horizon of the computers.  This would assure that the
test would measure performance of the position evaluation software.

>I'm afraid I can't explain
>that properly... but I'm convinced that "normal" moves are not good for test
>positions, neither for tactics nor for positional tests.

Why not good?  The non-sacrificial positions need not be positions having
several equally good moves.  I suspect that positions of this type could be
found, where one move was much better than the rest.

Note that I am thinking that design of engines to do well with positional
positions would *primarily* be a problem for design of the "position evaluation
software," as opposed to the "searching software." That may not be a valid
assumption.  I don't know.

>Sacrifices usually
>provide a strong "hint" that the engine has seen the pointe of the variation
>"requested" (the same goes for situations when pieces which are en prise, are
>not taken, IOW when material sacs are refuted for the reason hidden in the
>solution).
>
>But I didn't analyse the positions of that book (yet), or see if they are useful
>for tests... Maybe somebody has read it and can tell his impressions about it?
>
>(IIRC, I had difficulties to understand the point of some positions without
>having the book's text.)
>
>Regards,
>M.Scheidl



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.