Expanding parameters as we move to Middle Position

Before my week long hiatus due to personal matters, I ran initial trials in middle position for the Conan Librarian TAG strategy and a version of the strategy where every hand was limped.

I collected considerable data, but intermittently during my week's personal trials, I thought of the process and eventually realized that a factor would create several more variables for middle position play: unlike Early Position, chances were likely that a player in middle position would face action in front. An Early Position player could elect to limp or raise knowing he was usually 1st in, and comparing the profitability of either play was simply a matter of simulating 100,000 trials of each starting hand for each play.

However, a middle position player could face being 1st in, or face one limper, or face several limpers, or face a raise in front, or a raise and one or more callers in front. Each of these situations creates separate sets of variables that affect the profitability of each hand differently, since players will limp certain hands, and raise certain hands, and call raises with certain hands. Obviously, if there's a raise and re-raise, the decision becomes far easier: you're facing a monster and shouldn't play unless you hold AA, KK, QQ or AK. But most action falls into the many former categories.

Therefore, you need to conduct several different forms of trials for each starting hand to get an accurate measure of the hand's profitability. You will need to factor in how a hand plays when raising limpers, when limping behind, when open raising, when open limping, and when calling raises. Add in the fact that more hands are playable from middle position than early position, and you have the making of a long, arduous set of trials, all to determine an optimal middle position starting hand strategy. Plus, all of these hands are played with a static solid postflop strategy that itself has yet to be experimented with. It's enough for an average man to go bald tearing his hair out with.

But these answers aren't going to find themselves, and little in life worth doing is easy. Once an optimal strategy is devised, if you can comfortably profit from a 3/6 game considered by most to be too sub-optimal to play for profit (3-4 players to the flop), it will all be worth it.

The next step will be to determine the exact control parameters to set in the simulator for each trial, to best determine the profitability of each hand. I will outline this process once I devise suitable controls.