Author: Ernst A. Heinz
Date: 17:32:37 01/26/00
Go up one level in this thread
On January 26, 2000 at 20:08:04, Dann Corbit wrote:
>
>On January 26, 2000 at 19:59:43, Dave Gomboc wrote:
>[snip]
>>IMO, in this domain, experimental results are of more value.
>
>That's a good idea. Do we have a single program for which we have accurate data
>for at least 4 different CPU speeds? By accurate data, I mean a huge number of
>games under controlled conditions. By single program, I mean a single version
>of a single program.
No, we do not because all previous experiments with this kind of
self-play featured not enough games per program version. I have
reanalyzed them all as part of my Ph.D. thesis.
Please see Chapter 9 "Self-Play Experiments Revisited" of my book
on "Scalable Search in Computer Chess" for more detailed information
(below I quote the entry of this chapter in the table of contents).
9 Self-Play Experiments Revisited ............................... 157
9.1 Introduction ............................................. 157
9.2 Statistical Analysis of Self-Play Experiments ............ 158
9.3 Self-Play Experiments in Computer Chess .................. 162
9.3.1 1982 -- Belle (Thompson) .......................... 162
9.3.2 1983 -- Belle (Condon and Thompson) ............... 163
9.3.3 1988 -- TechMate (Szabo and Szabo) ................ 164
9.3.4 1990 -- Hitech and Lotech (Berliner et al.) ....... 166
9.3.5 1994 -- Zugzwang (Mysliwietz) ..................... 171
9.3.6 1996 -- Phoenix (Schaeffer) ....................... 172
9.3.7 1997 -- The Turk (Junghanns et al.) ............... 173
9.4 Self-Play Experiments in Computer Checkers ............... 174
9.4.1 1993 -- Chinook (Schaeffer et al.) ................ 174
9.4.2 1995 -- Chinook (Schaeffer et al.) ................ 174
9.5 Self-Play Experiments in Computer Othello ................ 175
9.5.1 1990 -- Bill (Lee and Mahajan) .................... 175
9.5.2 1997 -- Keyano (Brockington) ...................... 176
9.6 Conclusion ............................................... 178
Some comments and descriptions regarding the slef-play experiments
in computer chess are also contained in my article "DarkThought Goes
Deep" as published in the ICCA Journal 21(4) in December 1999 (an
electronic preprint thereof is available from the WWW pages of
"DarkThought" at the URL http://supertech.lcs.mit.edu/~heinz/dt/).
But I am currently performing an extensive self-play experiment
with at least 1000 games per program version, the results of which
will certainly be submitted for publication some time this year.
=Ernst=
P.S.
"Scalable Search in Computer Chess" now at Amazon.de and MKP online.
http://www.amazon.de/exec/obidos/ASIN/3528057327/
http://www.mkp.com/books_catalog/3-52805-732-7.asp
Please visit http://supertech.lcs.mit.edu/~heinz/node1.html for more
information about the book.
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.