Texas Holdem' AI Ideas
-
Guys I'm tired of the casino games you pick up at wal-mart... So I'm writing something better. I'm looking for ideas on ways to organize the AI. My experience with AI in college was pretty much limited to the Chess game we programmed. Computer suggests a move and then plays out an artificial game for about 4 moves ahead, repeats this for 3 or 4 possible moves, picks the best choice, and executes.Its a brute method and I'm wondering if anyone out there has any suggestions or could point me to some references.
-
Guys I'm tired of the casino games you pick up at wal-mart... So I'm writing something better. I'm looking for ideas on ways to organize the AI. My experience with AI in college was pretty much limited to the Chess game we programmed. Computer suggests a move and then plays out an artificial game for about 4 moves ahead, repeats this for 3 or 4 possible moves, picks the best choice, and executes.Its a brute method and I'm wondering if anyone out there has any suggestions or could point me to some references.
I don't think the chess analogy will work with Texas Holdem'. There is to much luck involved. I think I would develop a point scheme: 1) What are the odds that I currently have the best hand? 2) What are the odds that I could make it better? 3) How does the known betting history of my 'in' opponents effect these odds? 4) What is my position in the bettting sequence? 5) How much money do I have in relation to my 'in' opponents? I would give points to these factors and decide to bet accordingly. I would then adjust the point system until I have achieved an adequate winning ratio. You can even adjust points to emulate different playing styles (beginner, aggresive, conservative, etc.). Thats just off the top of my head, of course.