[sweetjazz]

Craig, did you know that the width of the bell curve is inversely proportional to the square root of the number of samples taken?

Basically, the width of the bell curve is measured by the quantity called standard deviation. If you graph the distribution of a player's observed winrate after N hands, the standard deviation will be equal to the standard deviation involved in playing 1 hand (which can be easily estimated, for example PokerTracker does this) divided by the square root of N. If you play enough hands (say hundreds of thousands), eventually the standard deviation becomes so small that the bell curve looks like a spike when you graph it on the same axis as you would graph the bell curve of, say, a 100 hand sample.

I've only skimmed your threads, but this seems to be what you're failing to grasp.