I'm talking about my idea of its implementation. There are plenty of ways to do it, but I will start with the easiest and the fastest one. I also feel that it will be enough in most cases. Technically speaking, "bluffing" is just raising your bet significantly above its statistically calculated best value. You can either try to develop a complex model or just use a Monte Carlo method to develop an approximation of this black box called "other players". The latter is way more easier and takes much less time. On top of that, the more experiments you make, the closer your results will get to a synthetical mathematical model if you will ever be able to develop it. If we are talking about real-world modelling, you will soon come to the point where there will be no noticeable differences at all due to natural limits of real-world objects. KISS