On Wed, Feb 17, 2010 at 12:41 AM, Elmer Fittery <[email protected]> wrote: > So, how do you calculate a number that gives a measure of how > good your position is?
Well... first you would have to define what "good" means. For example, hypothetically you could give each possible combinations of cards dealt a score and you could assign weights to those scores, and you could do a weighted sum of them (multiplying their score by their probability). Of course, to do this you would need to nail down some issues (like, which set of poker rules are you playing? as best I can tell poker has an infinite variety of rules -- it seems like i can always find another set of rules). Once you have a basic "goodness" rating system you can then do another pass where players will probably fold when their positions are not strong enough. If this radically changes the goodness rating you can build up from there basing your model on more informed players. But, anyways the first thing you need to do is break down your ratings into something tangible (such as cards dealt). However, I would not worry about players missing from the table. I think that that would be equivalent to a smaller table. FYI, -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
