A Model For Folding
When analyzing the last hand, I realized that I need a model to predict whether my opponent (given a certain hand) will fold. Basically a way for me to say “if he has X-X cards, given the board and this bet size, he’s Y% likely to fold”. At the moment I don’t have that model yet. But can it be done? I certainly think so.
Off the top of my head, there are a few variables to consider. First, definitely, would be the pot size and how much he needs to call (essentially the pot odds). Then the value of his hand given the board. I’m thinking that as a rough guide, you can use the percentage of hitting top pair or higher by the end of the hand to do this. I say a top pair or higher because that’s often good enough to win. You can tweak the number of outs he has according to various situations, but I think that’s a decent general guide. Those are the main variables – number of outs he has, and the pot odds he’s getting. Add in a variation for bluffing or semi-bluffing, and you have a rough model.
It’d be a crude model, for sure, but for analysis purposes, I think it could work decently enough to determine the best play in a situation. It doesn’t need to be extremely precise to be able to help in decision making. And I am going to get started working on the model.
What do you think? Can such a model be done? Or do you think it can be done, but that I’m missing some variables – if so, what?