So I was reading this paper this that talks about clustering different poker hands based on their performance against opponent clusters. However, I am not sure how to do this efficiently. For the flop, there are 1326 * 19600 = 2.6 million combinations, and exponentially more for the flop and turn. How do I reduce this sample space so I can still calculate all of these values while also making sure to account for all the combinations?
Asked
Active
Viewed 22 times
