As you may remember, I’m working on a game of Kaiser for Facebook. (Yes, I’m still working on it, it’s just a really slow process!) I got a comment on the post just this morning and it points to exactly the problem I’ve run into: how to make the artificial intelligence (i.e., the computer) know how and when to play its cards. How will it know whether its hand is any good? And how will it know, based on cards that have been played, what card to play next and whether its hand has gotten better?