Quote (Mastersam93 @ May 21 2017 10:10am)
Could run the simulation until you reach $0, over and over, and each time record which number hand had the highest bankroll and find the average.
How do you handle the fact the length of games may differ?
For example, you play a set of games that last 10 hands but had the highest amount at hand 5.
You play another set of games that last 20 hands but had the most at hand 6.
What's the math there? 5/2 probability for hand 5 and 6/2 for probability of hand 6?
Also, what happens if in the same game you match your highest amount multiple times before going bankrupt?
I think I am confusing myself.
Quote (Ideophobe @ May 21 2017 10:32am)
definately
so run a simulation of it until you go broke like a billion times (if it takes too long throw in an upper bound for your bank as well but this should be fairly high)
hold 2 running sums for the MaxBankAmount and MaxBankAmountSquared , ie what was the optimal cash out point?
then you can get the average and standard deviation for the cash out points from these two sums and whatever your n is.
because the odds are never changing you can expect bell curve rules of standard deviation to apply ie AT LEAST 75% of all games the optimal cash out point will be AT LEAST OptimalCashOutPointAVG - 2*OptimalCashOutPointSTDDEV
then you can run sims using your new cash out strategy as an upper bound and keep a running total of how much money you have spent/made
Not sure I understand. =p