Forum Archive : Programming

BKG beats world champion

From:   Marty Storer
Date:   10 September 1991
Subject:   Re: Computer Backgammon

> Several years ago, there was a backgammon program that defeated the world
> champion in a exhibition match.  This was described in Scientific
> American and I have been told that the program was briefly sold packaged
> on a PC for about $4000 around 1982.  Of course, the general
> marketability of a $4000 dedicated backgammon computer being what it is,
> the enterprise quickly failed. (One could also argue that $4000 is better
> spent playing against and learning from experts, but that's another
> story).
> At any rate, the computer won, but an analysis of the games revealed that
> 8-9 mistakes were actually made by the program, and were glossed over by
> the roll of the die.  We all know this from club or tournament play;
> there is no substitute for rolling well ;-)  The difference between an
> expert and a experience club player is perhaps 3% (feel free to argue
> this one!)...well, to me that translates into 48.5% of the time the
> lesser player will win.  That's why experts like to play rounds of 11 or
> 17 points and are very cautious (i.e. judicious) with the cube.  The more
> games, the greater their advantage over the lesser player.  I believe the
> exhibition match was to 9.

The match took place in '79, I believe, and I think it was a 7-point match.
A program developed by Hans Berliner of CMU defeated the winner of the
annual Monte Carlo tournament, Luigi Villa.  (There's no "World Champion"
in backgammon in the same sense as there is at chess; the winner of the
Monte Carlo tournament is labeled the World Champion for some reason,
despite the lack of central organization and all the other trappings of the
World Chess Championship that backgammon lacks.)  I remember seeing the
complete match score in some backgammon publication or other around
'79-'80; the machine didn't play as well as Villa did, but played pretty
well nevertheless.

The statement that there are programs that play better than humans was
probably made by someone who heard something about the computer-vs.-Villa
match.  It's false, anyway--the standard of play has improved considerably
since then, and there has been hardly any work on computer backgammon to
extend Berliner's very good work.  A program called Expert Backgammon,
recently on the market and available through various backgammon
organizations, seems to play a halfway-decent-but-not-that-great game.

I think Berliner made the "8-9 mistakes" claim (or a similar one) in his
subsequent Scientific American article on his program, his methodology, and
the match.  I'd bet the machine made more mistakes than that; I don't think
anyone but Berliner did the analysis you mention--at least, he didn't say
otherwise in his article.  I don't think Berliner's a recognized expert; I
know that he claimed the machine made a brilliant move with a roll of 22 in
the last game of the match, when in fact the move was a blunder.  (The
machine won a hopeless race by throwing multiple sets of doubles; its 22
move facilitated racing.  This caused some on-site expert commentators to
label the move a good one, and probably influenced Berliner's thinking
about it.  I'll dig up the article and post my analysis of the position, if
anyone wants me to.)

Anyway, Berliner's article (June 1980 Scientific American, I think) is
well worth reading, especially for people wishing to develop strong
backgammon programs.

    Play well,
Did you find the information in this article useful?          

Do you have any comments you'd like to add?     



Adjusting to a weaker opponent  (Brian Sheppard, July 1997) 
Anticomputer positions  (Bill Taylor+, June 1998) 
BKG 9.8 vs. Villa  (Raccoon+, Aug 2006) 
BKG 9.8 vs. Villa  (Andreas Schneider, June 1992) 
BKG beats world champion  (Marty Storer, Sept 1991) 
Backgames  (David Montgomery+, June 1998)  [Long message]
Blockading feature  (Sam Pottle+, Feb 1999)  [Long message]
Board encoding for neural network  (Brian Sheppard, Feb 1997) 
Bot weaknesses  (Douglas Zare, Mar 2003) 
Building and training a neural-net player  (Brian Sheppard, Aug 1998) 
How to count plies?  (Chuck Bower+, Jan 2004)  [GammOnLine forum]
How to count plies?  (tanglebear+, Mar 2003) 
Ideas for improving computer play  (David Montgomery, Feb 1994) 
Ideas on computer players  (Brian Sheppard, Feb 1997) 
Introduction  (Gareth McCaughan, Oct 1994) 
Measuring Difficulty  (John Robson+, Feb 2005)  [GammOnLine forum]
Methods of encoding positions  (Gary Wong, Jan 2001) 
N-ply algorithm  (eXtreme Gammon, Jan 2011) 
Neural net questions  (Brian Sheppard, Mar 1999) 
Pruning the list of moves  (David Montgomery+, Feb 1994) 
Search in Trees with Chance Nodes  (Thomas Hauk, Feb 2004) 
Source code  (Gary Wong, Dec 1999) 
TD-Gammon vs. Robertie  (David Escoffery, June 1992) 
Training for different gammon values  (Gerry Tesauro, Feb 1996) 
Training neural nets  (Walter Trice, Nov 2000) 
Variance reduction in races  (David Montgomery+, Dec 1998)  [Long message]
Variance reduction of rollouts  (Michael J. Zehr+, Aug 1998)  [Long message]
Variance reduction of rollouts  (Jim Williams, June 1997) 
What is a "neural net"?  (Gary Wong, Oct 1998) 
Writing a backgammon program  (Gary Wong, Jan 1999) 

[GammOnLine forum]  From GammOnLine       [Long message]  Long message       [Recommended reading]  Recommended reading       [Recent addition]  Recent addition

  Book Suggestions
Computer Dice
Cube Handling
Cube Handling in Races
Extreme Gammon
Fun and frustration
GNU Backgammon
Luck versus Skill
Magazines & E-zines
Match Archives
Match Equities
Match Play
Match Play at 2-away/2-away
Opening Rolls
Pip Counting
Play Sites
Probability and Statistics
Source Code
Strategy--Bearing Off
Strategy--Checker play


Return to:  Backgammon Galore : Forum Archive Main Page