Actually using Effects of Removal and Correlation Coefficients with the tag values of the derived counts is more accurate than simulations and three is no "random chaos" in my approach.
The Effects of Removal were calculated to six decimal places with combinatorial analysis of all possible combinations for that particular situation. The EoR for each card in a given situation are a least squares line estimate but they are still very, very good. These EoR for each situation are listed in Don Schlesinger's Blackjack Attack 3rd edition. When you do simulations you are going through many, many trials hitting on all of these combinations that have already been taken into account in the EoR. It required a lot of computer power to calculate the EoR to six decimal places for each situation that was only published a few years ago by Don Schlesinger. In Peter Griffins time the EoR were just approximate and to only a few decimal places.
So to double check that my calculations with the EoR were correct, what I did, as I explained earlier, was that I calculated the indices for HL with my spreadsheet. I tested almost every index I could find for the HL. Over 100 indices as I also checked negative indices. In EVERY SINGLE CASE, the calculated indices from the EoR agreed 100% with the published indices generated from simulations.
The EoR calculation are very quick and have zero variance as opposed to simulations that require millions and millions of hands. There is no "random chance" here.
Bookmarks