I am not, nor pretend to be a "math guy", and you (Don) will probably prove me wrong, but I would think in the very short term, there would be some sort of increased risk for playing two hands based on the co-dependence of both played against the same dealer cards. I mean you run into a dealer that is pulling an obnoxious number of 20's and 21's (and we all have), and both hands are going to get squashed repeatedly. A very short-term variance type of thing and while we normally don't place much consideration on such a small sample size, someone with limited funds maybe should consider this...no?