Newsgroups: comp.ai.games,comp.ai,rec.games.bridge
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!news.mathworks.com!newsfeed.internetmci.com!in2.uu.net!news1.digital.com!decwrl!amd!netcomsv!uu4news.netcom.com!netcomsv!uu3news.netcom.com!ix.netcom.com!netcom.com!grampa
From: grampa@netcom.com (Thomas N. Schmied)
Subject: Re: How can we encourage AI interest in bridge?
Message-ID: <grampaDKF6Ko.GGt@netcom.com>
Followup-To: comp.ai.games,comp.ai,rec.games.bridge
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
X-Newsreader: TIN [version 1.2 PL1]
References: <4bursl$79f@pith.uoregon.edu> <4bvju3$gcn@news.paonline.com>
Date: Sat, 30 Dec 1995 22:02:48 GMT
Lines: 34
Sender: grampa@netcom4.netcom.com
Xref: glinda.oz.cs.cmu.edu comp.ai.games:3358 comp.ai:35560

Jerry Henderson (jhenderson@shivasys.com) wrote:

: I have a couple of commercial bridge programs. I always win, even
: though my game is not particularly good. The "state of the art" in
: bridge AI doesn't seem very good. I don't know much about the AI
: field, but I have worked as a professional programmer (13 years) and
: have done some simulations. I would think that bridge would be _much_
: easier to program than would be chess.
[clip]
: work. But it doesn't seem conceptually complex. Am I missing something
: here?

: --
: Jerry
Indeed you are.  Bridge much easier to program than chess?  The fact 
that there are now several chess-playing programs that compete at a 
world-class level, and none that are even good player to advanced, never 
mind expert, in bridge, should give you a clue.  The reason is not 
related to the amount of effort expended on each.  The overriding reason 
is that in chess, there is NO HISTORY.  The entire world of knowledge 
about any particular position is revealed in the current state of the 
board, with no concern to what has transpired before.  In bridge, every 
call, every card played to date, carries subtle implications about the 
probabilities for success of each current alternative.

IMHO, the best prospects for a bridge-playing computer program, able to 
compete with the best talent in the game (which in 
all likelihood will be based on AI and expert systems technology) are to 
have the effort funded by a truly wealthy patron of the game, much along 
the lines of what Ira Corn did with the Aces.  This is not an effort in 
which a few programmers, working in their spare time, are going to be 
successful.

Tom Schmied.
