Guynoire said:
I was trying to calculate the standard deviation of multihand play, when I realized that the two hands are not independent events, which complicates things.
Does anyone know the covariance between two hands at the same blackjack table or how to approximate it?
Recognize that the variance and co-variance will change based upon different rule sets.
That being said, for a standard 6D,S17,DAS,DA2 game the standard deviation for one hand is 1.155. For 2 hands, 1.35. For 3 hands, 1.52.
So, when betting $10 for one hand, the standard deviation is $10*1.155=$11.55
When playing 2 hands at the same time of $5 each, the total standard deviation would be $10*the square root of 2*1.35=$9.54
For 3 hands of 3.33, it would be 3.33*square root of 3*1.52=$8.76 etc
Hope that helps a little anyway.
To determine the variance and co-variance of a given game in the first place, I'd guess u need software and know what u r doing lol. I lack both requirements.
But if u know the variance and co-variance of a given game, there are formulae out there that you could apply. If u look hard enough, u may very likely be able to find these numbers for the game u have in mind.
I don't think using these numbers would be hugely off in most circumstances in that the variance for BJ doesn't vary that much. The co-variance might have a wider range of values depending on specific rule sets.