I. I am not able to reproduce figures for the Ace/Five count (discussed here: https://wizardofodds.com/games/black...ce-five-count/) with CVData. First, I attached the basic strategy for 6D H17 DAS LS, and then defined the value of each rank (A = -1, 5 = 1, others zero). Then, I defined a betting ramp based on running count as mentioned (1 for RC <= 1 and double until max bet as long as RC >= 2). Since this seems like a low-variance strategy, I ran 500 million rounds (thinking convergence would be quick).
When I went to look at some output (variance chart here), I got something not sane: a single row corresponding to a count of zero. Since I asked the software to double the bet at very modest counts, I'd expect rows for each value. Is there something I'm doing wrong?
II. How were the SCORE values obtained? Isn't
score.png
(a advantage, V variance)? Why then is (0.005)^2 / (1.1 or so) not even close? Thank you.
Bookmarks