[Irieguy]

Your question gets to the heart of a common gambling/statistical misconception invovling an apparent paradox between two concepts:

1. Independent Trials: "The cards have no memory." The historical results of a random event have no bearing on the results of the event the next time it occurs. It doesn't matter how hot or cold you have been running for X hands. For the next X hands you rate to win at your win rate +/- your standard deviation

2. Regression to the mean: "Everything evens out in the end." Regression to the mean is real, but it isn't a phenomenon... it's just what happens because of how "mean" is defined. The more trials you have, the closer your actual mean result will come to your expected mean result. But if your actual results are skewed, you don't suddenly become more likely to have results that weight your average in the opposite direction.

These two concepts should be easy to understand, but I have seen several 2+2ers employ the casino-betting strategy of waiting until the roulette wheel has come up red 5 times in a row and then betting their whole bankroll on black.

Variance isn't an entity. It has no power, no volition, and no meaning. It is just concept. A concept so hard to grasp that we give it ultimate power, malignant volition, and supreme meaning in our little pea brains.

Irieguy