Mark Fickett Rolls Lots of Dice

Mark Fickett got really serious about checking dice experimentally for fairness. He constructed a mechanism that automaticallyy a die every 4½ seconds and ran that on different dice, around 8000 times each, until he could analyze all the different dice from major manufacturers. There are charts, videos, statistics, and construction details.

One interesting finding that Mark makes for this experiment, regarding number of rolls per die, is: "This leads me to a heuristic of 100 rolls per side (800 for a d8, 2000 for a d20). It is of course a subjective heuristic." This broadly lines up with my own prior recommendation from looking at power curves for chi-square tests on these dice: "For a d6, I wouldn't want to use any less than n = 100 as a minimum (and ideally something like n = 500 if you're serious about it). For a d20, n = 500 would be a useful minimum (and at least several thousand to find reasonably small variations)."

Mark Fickett Art: Dice Roller