in this post :
http://www.blackjackinfo.com/bb/show...6&postcount=15
i was alluding to something, that i'm not even sure what the heck i'm alluding to, lol.
errhhh, i think in a nutshell, what i'm trying to ask is:
essentially is it possible to have some positive EV for some play and concurrent to that is it possible to have concurrently some '
standard deviation', such that the worst negative standard deviation one might expect to possibly happen, can never really happen because of the fact that the worst case scenario is so much less of a disaster than the standard deviation (
derived from data) would lead one to expect?
if so, i'd think the bell curve would look sort of warped, errhh kind of skinny on the 'evil' side of the curve and fatter on the 'blessed' side of the curve. no?
possible?
edit: one thing that is happening data-wise is that relatively high variance has been happening of the 'good luck' nature, data-wise this drives up the standard deviation, but the thing is, i know the worst case scenario for a given play, and it can't even come close to the 'good luck' stuff that does happen.
so to me, it's as if the standard deviation derived from my data can't properly reflect the 'evil' side of the bell curve, sorta thing.