the next level

I'm currently at a betting level where my base bet is $25. I have done well at this level over the last year + and want to double my base bet (and spread) to $50. I have the bankroll to support this but the problem is, I've been on a pretty good winning streak and fear an adjustment or losing streak as soon as I increase. It seems crazy to continue play and Hope for some loses before going up. Interested to know what others do?

Presumably you believe in the numbers, which is why card counting interests you. What makes you think that there is any connection between raising your unit and the onset of a losing streak?

I don't think i't superstition. I just know I have been doing very well lately and and that a correction or pullback or losing streak is due. Otherwise I would just continue winning indefinately at a rate several times higher than the math indicates. It seems like a bad idea to up your wager just before such a streak hits and lose twice as much as you would have. However the alternative is to contine playing at the current level, waiting, almost hoping for some losses to bring my winning rate more in line. That doesn't make much sense either.

Nothing is "due", and you have no idea when a streak is coming. To think that a negative streak is "due" or just around the corner is being superstitious.

Is the voodoo idea that mother nature is going to make your winnings disappear by future bad sessions. That isn't the way it works. The basic idea is that for this equation:

(X + N*Y) / (N+1)

where X is your large win, Y is your expected win rate, and N is the number of normal sessions you have in addition to that monster session, the limit of that equation is "Y", your normal EV, the X gets diluted.

A simple example, from my playing.

At a decent DD game, playing $25-$200 last summer, I won just under $10K in about 2 hours. I will add that I increased from $25-$200 to $100-$600 after the exceptional streak had tripled my session BR quickly. My actual EV for that game was around $100 or so for my initial spread. I've played a few more sessions and saw results much nearer to the normal expected EV, up and down as usual.

But suppose we take that $10K win, and that I go back for the next 500 times and _never_ have a losing session, I hit right on my EV each and every time. Notice what happens:

(10000 + 500 * 200) / (501)

that is, one 10K 2 hour session, 500 $200 dollar sessions, what is my overall win rate?

$220 per session, 110 per hour, about 10 bucks to the right of the mean of the bell curve.

Hopefully that explains how your big wins get "corrected". Not by corresonding bit losses, but just by large numbers of normal sessions which overwhelm that big session in the overall average...

That's how the god of statistics corrects our unexpected results. Not by offering up opposite results to cancel them out, but by offering up large numbers of normal sessions to wash the large +/- results away.

I doubt if there are any players who can attest to having their big winnings session diluted by averaging only, with no sessions that underperformed expectation just as often as the sessions that exceeded expectation. You may as well say that if you start off as a loser, you'll stay a loser, but it wont look so bad as you play more hands because it'll get lost in averaging. Probability doesn't suggest certainty.

either case _can_ happen.

You can have ups and downs. You can have more but smaller ups and fewer by bigger downs. Or anything in between. But the idea of "offsetting" is mistaken, as it implies a form of memory in the system, when there is none.

For example, just because you have one -10.0 SD session, does not imply that you will have a +10.0 SD session to get you back to the proper mean EV. Any more than flipping 20 heads in a row guarantees you that you will eventually flip 20 tails in a row. Both events are equally probable, but one happening doesn't guarantee the other, particularly when one is so improbable. Also note that having a single -10.0SD session could be followed by a +10.0SD session or another -10.0SD session, since the +10 and -10 sessions are _equally_ probable...

Note that I didn't say that there would be no offsetting sessions, there usually are. When you sample from a very large population, you get numbers all over the map, but the larger the sample, the closer to the overall mean that sample must be. I _did_ say that you could have one huge session, and be very lucky and have nothing but winning sessions after that, and that you _still_ end up right at your expected EV without having any losers at all. I didn't say that was very probable, I just used it as an example of how large numbers of sessions dwarf single sessions.