Thank You
for demonstrating once again your complete ignorance of the methods of calculating Basic Strategy. Baldwin, Thorpe, Braun, et. al did not make any assumptions about the results of splits doubles, hitting or standing. They simply started with a full deck, and removed the players cards an the dealers up cards, then calculated the probability of drawing each card, then repeated the process until each chain led to a bust or 21, and also did this process for the dealer (stopping of course at 17 or higher), calculating the probabilities of the dealer making 17, 18, 19 ,20, 21 or bust. It is all based on the proportion of cards remaining in the deck, or decks for multi-deck. This process is called combinatorial analysis. There is nothing terribly complicated about it, but since blackjack has open-ended drawing, there are a lot of branches to follow. What you have is a whole lot of repetitive, simle calculations to make. This the sort of task for which computers are ideally suited. Baldwin and his collaborators (I don't remember their names off hand), the first group to make the effort to calculate Basic Strategy in the mid '50's, didn't even use a computer. They calculated it manually, using mechanical calculators. Their results were pretty much the same as those of Thorpe, Braun and later workers. Once you know the probability of drawing each hand, and the probability of each dealer result, it is eady to calculate the "Expected Value" This is simply the net monetary result for each outcome multiplied by the probability of that outcome. Basic Strategy is the play that give the highest Expected Value. The only assumption made in any of this is that the order of the cards is random. This is often not strictly true (otherwise shuffle tracking and ace location qwouldn't work), but no one using rigorous methods has shown that actually casino shuffles will produce results much different from a truly random shuffle if the cards are played on the assumption that they are random.
You obviously believe that a player's recent results for a particular play (or more correctly, his (probably fualty) memory of them, are of greater value in predicting his future results than probabilities casculated by sound mathematical methods. I doubt that I or anyone else will change your mind, but I don't understand why you believe this to be true. Do you believe that the presence of a particular player at the table somehow influences the shuffle? How? A small sample of random events will very often differ considerably from the theoretical results. The discipline of statistics can accurately predict the probability of this. Also, one's memory can often be influenced by frustration. When you say that you draw 10 to a 12 90% of the time, or that you lose 75% of your doubles against high cards, I strongly suspect that your perception of these numbers has been warped by the frustration of losing. If these numbers are accurate, I suspect that they are based on a sample too small to be statistically significant. Note that a lot of Basic Strategy is a matter of choosing the least bad alternative. A lot of splits are defensive. A great example of this is splitting eights against a 10. This play has a terrible EV. It just happens to be a little less bad than hitting, standing, or surrendering. You will frequently lose both hands. However, I would submit that if this play is killing your bankroll, your bankroll is much too small, and you would be much better off not playing blackjack than altering your strategy because you are afraid of the double loss.