Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site rti-sel.UUCP Path: utzoo!linus!philabs!cmcl2!seismo!harvard!talcott!panda!genrad!decvax!mcnc!rti-sel!trt From: trt@rti-sel.UUCP (Tom Truscott) Newsgroups: net.chess Subject: Re: Why can't a machine be World's Checkers Champ? Message-ID: <332@rti-sel.UUCP> Date: Tue, 6-Aug-85 19:14:16 EDT Article-I.D.: rti-sel.332 Posted: Tue Aug 6 19:14:16 1985 Date-Received: Sat, 10-Aug-85 23:29:57 EDT References: <474@oakhill.UUCP> Organization: Research Triangle Institute, NC Lines: 23 I feel that if the effort and expertise spent in creating any one of the current top five computer chess programs were spent instead on a computer checkers program we would have a computer as World Checkers Champion (not officially, but it would be generally suspected). For example, back in 1977 a checkers program written by Eric Jensen, Bruce Wright, and myself had a credible score (1 win, 2 draws, two losses) against Elbert Lowder who was at that time considered #2 or so among human players. The program had an excellent search, was tolerably efficient, had no opening book, and the authors knew nothing of checkers but the rules. (In checkers, many tournament games *end* within the book.) The same authors spent an order of magnitude more effort on the "Duchess" chess program, which never achieved an Expert rating. I think the problem with checkers is that, whereas it is much "easier" for a computer than chess, it is sufficiently harder than, say, reversi (which is dominated by computers) that few are willing to put in the effort needed to produce a world beater. Particularly since checkers is not thought of as an intellectual game. But it would be an interesting milestone. Tom Truscott