Your advantage at each TC goes up with deeper pen but so does the variance. Optimal bet is a function BR, RoR desired, advantage, and variance. Eliminate the former two and you have EV/variance. So for each TC bin, variance at shallower pen is reduced more than EV is reduced. More accurately the ratio between EV and variance decreases with deeper pen.
More 20 vs 20 pushes.
Last edited by Three; 02-12-2019 at 09:04 AM.
You are right. I started writing a different answer but got confused about a higher TC eventually decrease in advantage. My original answer was a guess that it had to do with decreased accuracy due to the exponentially increasing significance of deck estimation difference between the actual deck remaining and estimate deck remaining for data collection bins in the sim at deep pen. But I didn't have confidence that was right. I know that happens but don't see why it would reduce average advantage. I don't see the phenomena for Hiopt2/ASC so you are probably right about SE.
Thanks Don. The standard error conclusion makes complete sense, because I tried that exercise with multiple sims (my own and the archived ones) and it happened in all of them. So it can`t be an anomaly. It was a small difference, but I was curious as to what it could be.
Thank you for your response also Three, I didn`t know all that about optimal bet.
floating advantage is the reason for the same TC equaling a better advantage with deeper pen, but that doesn't really come into play until deep deep pen. The main reason you'd have a different ramp based on pen is because the lower the pen, the lower your overall edge is, so it would probably be recommended to have a bigger spread to overcome the smaller edge.
Basically lower pen needs a bigger spread to beat that game, hence bigger bets at lower counts
Last edited by CountinCanadian; 02-15-2019 at 11:17 AM.
Bookmarks