It seams that the best indicator of a winning or losing streak are wins or losses and that betting more when winning or less when losing would provide a distinct player advantage.
Brilliant! I never thought of it. Sort of like, the best indicator of winning a baseball game is scoring more runs than the other team, all you have to do is have more men on when you jack one out.
Sorry for the sarcasm, but here's the deal:
Imagine everyone's favorite coin toss game, or any "50/50" game with outcomes x and y: the sequences xxxxxxxxxx and yyxxyyyxyx are equally as likely to occur. Which one is going to stick in your memory though? How do you predict the first one, and bet accordingly? If you recognize you are in the "middle" of a hot streak, how do you know how long it will last? Maybe the BJ Gods owe you one, and subsequently deal nothing but stiffs to compensate...
If you are going to bet more, do it after a loss: removal of the average subset of cards leading to your loss makes the remaining pack ever-so-slightly more favorable to you - ON AVERAGE. But it's a tiny amount. Very tiny. If I remember correctly, the strongest indicator is a push, and it's a negative one - more likely when the count is high, you'll push on 20's, count drops like a rock.
That's the only mathematical (and therefore, correct) connection between previous hand results and future hand prediction I know of. The gain is absolutely infinetesimal relative to actual knowledge of remaing deck composition (i.e., counting), however, so don't give up your sense of logic for voodoo! If you insist on streak-ology, I'm sure Walter will be happy to entertain you on the proggy page. :)
p.s. You'll never see a pure progressionist barred - so it IS worth checking out for camouflage - use a crappy progression during neutral advantage situations and gain some cost-free playing longevity!