Was busy doing my taxes. Some more explanation:
CVData calculates variance by round (not by hand) by TC. Independently, it calculates variance by round for all counts. These calculations are separate. Since calculations are by round, no co-variance correction is required. Any skew, kurtosis, change in actual penetration, or any other aberration would be included in these calcs.
CVCX is a single-hand simulator. The two-hand estimates are rather complex as they allow for changing from one to two hands at a specified count, which significantly alters TC frequencies. These calcs are based on a set of tables that were created running many billions of hands. It must also hold the data to recalc results instantly when switching the penetration card by card, or rules, decks, etc. But, the results are estimates for two hands as stated in the manual.
If you want to see accurate multi-hand results in CVCX, you would need to run a CVData sim with two to seven hands, and then click Call CVCX. The data will be transferred to CVCX and allow modification of the bets there. In this case, results should be accurate.
Reply With Quote
Bookmarks