Keith wrote:For what it's worth, the standard deviation for a 307-roll sample of a six-sided die is just under 0.1 (about 0.0976 if I've done the math correctly). Using a normal approximation, therefore, we expect to see something in the range 3.4-3.6 about 2/3 of the time. And we expect something in the range 3.3-3.7 about 95% of the time.

Fantassin's result is a bit more than 1 1/2 standard deviations below the expected theoretical mean. Such an outcome is certainly bad luck. But it's far from completely out of the norm or even weird. It's worth a complaint or two, but probably not worth trying to "fix" your computer.

And certainly worth "bitching" to your opponent about the unfair bad luck that ruined your perfect plan. I've certainly had that happen to me.