Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The Brick Wall

Author: Stuart Cracraft

Date: 11:38:47 09/19/04

Go up one level in this thread


On September 19, 2004 at 12:01:16, Robert Hyatt wrote:

>On September 19, 2004 at 11:10:14, Stuart Cracraft wrote:
>
>>Hi -- I am looking for 2 or 3 beta testers who would receive
>>(full) source code to my program and in return would provide
>>input and comments about improving the search. They would
>>simply agree not to redistribute it and in fact discard it after
>>a week or two of looking at it (and commenting.) The program is in C,
>>5000 lines. The search and quiescence routines are 600 lines total.
>>
>>The reason I am considering this is due to hitting a brick wall at 249/300
>>on WAC for several weeks now and knowing there are things I just cannot
>>find or go further with. The above score is at 1 second per position on a
>>1ghz P3 with a small transposition table. I am told that 270-300 is considered
>>"good" for this time control on this test. On this same machine
>>at the same time setting, with WAC, Crafty gets 270/300.
>>
>>The kind of beta testers I'm looking for are experienced programmers
>>who have written their own program and it has long since graduated
>>from Win-at-Chess as a test suite, perhaps scoring 270 or above at
>>1 second per move on a Pentium III 1ghz or above. To them, WAC has
>>become ho-hum and in fact they are currently just sitting on their
>>laurels without a lot of major advances. Their program has "matured."
>>They see themselves as senior chess programmers helping less
>>experienced authors.
>>
>>What I would favor
>>
>>   1) beta tester with solid program agrees to simultaneous exchange
>>      of source code
>>
>>   --and--
>>
>>   2) beta tester agrees to seriously review the quiesce(), search(),
>>      store(), retrieve(), and iterate() functions.
>>
>>I am fine to sign any non-disclosure agreement.
>>
>>This is just an attempt to break through a brick wall.
>>
>>Stuart
>
>What you are really looking for is a "quick fix" to a problem without one.
>There is no easy way to make a program revolutionarily better.  It is a
>continual evolutionary process that, for the main ingredient, needs _time_ in
>copious quantities.
>
>There are far better projects to tackle if the goal is "done in 180 days or
>less".  Chess doesn't fit that very well.  It is more like (for me) not done in
>36 years and counting...

I don't think a quick fix is what I am looking for. I am attempting not to
have to reinvent the wheel. I am attempting to build on the shoulders of
others without, like Sisyphus (http://en.wikipedia.org/wiki/Sisyphus),
having to roll the boulder up the hill and have it roll back on top of me.

I am attempting to avoid rethinking what has been thought before. That is why
we are humans and not machines. Writing and speech came about to share
experience so that individuals would not have to go through the exact same
experience again if we chose to listen to others and profit from their
experience perhaps in an even more direct manner as has been done on
this board between myself and the extensive general help from others.

But now the search has specific issues. General help may not help as
much. The curve is flattening.

If nobody takes me up on the offer, sobeit. But at least I tried.

I am curious about what are the changes that might take this program from
250/300 to 270/300 or 280/300 or 290/300 or 299/300. It has generated a life of
its own, that curiousity. It is not born out of commercial necessity nor to
attempt to impress others. It is simple, idle curiousity. I've been curious
about this subject for 36 years since age 10. It is not going to be shelved
or to go away for som other project. This *is* ***the*** project.

Assuming I got a result like 260/300 or 270/300 or something, would the
project be done? No -- obviously not. But perhaps I could go to the next
area knowing that the tactics seemed to be relatively more solid and clearly
not obviously weak. That would be a useful goal to have completed.

We all set our personal goals. For me, 260/300, then 270/300, then 280/300
are my goals. 250/300 fell today after a small extension improvement. Does
the program necessarily play better? Of course I would never say that. But
that's not my goal. I want only to ensure that on its first test suite it
has a decent record. Later test suites, endgame suites, etc. will come in
due time.

Stuart



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.