Anomalies can arise any time you deal with statistics of this sort. That's why statistical calculations come with confidence ranges; e.g. 95% confidence means there's a one in 20 chance the result falls outside the mathematical / statistical expectation.
In poker, there's a compounding factor in that the amount available to win can differ significantly. Again, this will average out over a large sample size, but a given individual's sample may be too small, which can lead to results both above and below what the math predicts.


Reply With Quote
