desj@brahms.BERKELEY.EDU (David desJardins) (02/27/86)
In article <683@oliveb.UUCP> jerry@oliveb.UUCP (Jerry Aguirre) writes: > >The reason it was believable was in the nature of the game. Unlike >standard chess where all positions are visible this was more like >"battleship" or poker in that the "position" of many of the opponents >pieces were UNKNOWN. Thus a major portion of the game was in trying to >outguess an apponents stratigy. > >McCoy explained that he used information gleened from Spock's >(confidential) psychological medical records and exploited a weekness in >Spock's personality. > >This was entirely consistent with previous stories in which Kirk has >pointed out that, in a game like poker, logic isn't enough. It is also >consistent with McCoy's constant poking at Spock's psychology. Sorry for cross-posting to net.games, but I want to clear up a fairly common misconception. In *all* games satisfying certain minimal criteria (finiteness etc.), whether or not they contain hidden information, there is an "optimal" strategy. This includes battleship and poker, the examples given above. Obviously this strategy cannot be sufficient to win any individual game, but it is optimal in the sense that no strategy is statistically superior in a long run of games. And in particular, since all decisions in the optimal strategy are made randomly, no knowledge of the opponent's psychology can be helpful against a player who plays the optimal strategy. A note: the example of battleship is particularly interesting to me as I have spent some time trying to work out optimal strategy for simplified versions of this game. If there is any interest in discussing this on the net, or if anyone has interesting insights into this problem please let me know. (Maybe we need a net.games.theory? :-)) -- David desJardins P.S. Please direct followup articles to only the applicable newsgroups.