The Chaperone said:
FWIW, I don't think his numbers are anywhere near accurate.
I agree. He oversimplifies everything way too much.
He uses a standard deviation of 1 unit. Which makes the rebate much worse. The higher the stdev the better the rebate is. We know stdev is higher than 1 unit per round.
He also assumes that your win rate per round is what your advantage is. This is not true. You lose more hands than you win.
To get an accurate answer you have to find the probability of results at various points and find your expectation at each point. To find the probability of a result you can use the normsdist function in excel. What I did was find my expectation at each point from 15mm to -15mm in 10k increments. For any point less than 0 your expectation = result *0.8. You sum up the expectation at all points and then you get your total expectation.
So if we assume the following:
Rounds per Hour 60
Hours 12
Unit 100,000
Stdev per Round 1.5
Edge -0.41% (6D S17 DAS)
Rebate % 20.00%
I get the following results:
Expectation 52,357
Total Bet 72,000,000
Edge 0.07%
Now he possibly could get the casinos to offer him LS and/or RSA to reduce the house advantage. He may also be playing less than 12 hours (eating, bathroom breaks, etc). Throw in some index plays and maybe a 1-2 spread and he could work up a mediocre advantage.