How to read the results
D×√n — the test statistic. It measures cumulative skew: your rolls are summed in order (face 1, then 1+2, then 1+2+3…) and compared against what a fair die would produce. The largest gap, scaled by √n, gives this number. Higher means more biased.
5% threshold — a fair die would exceed this value only 1 in 20 times. Crossing it is mild evidence of bias — suspicious, but not conclusive.
1% threshold — a fair die would exceed this only 1 in 100 times. Crossing it is strong evidence the die is genuinely biased.
Max dev. face — the face where the cumulative gap is largest. If it's low (e.g. 5), faces 1–5 are collectively over- or under-represented. If it's high (e.g. 17), the skew runs across most of the die.
Average roll — the mean of all your rolls compared to the expected mean of a fair die. A lower-than-expected mean suggests the die is disadvantageous to the player (damage rolls hit softer, skill checks fail more). A higher mean suggests an advantageous die. Note: this is a quick intuitive indicator, not a formal test — a die can have a perfect mean and still be biased in ways that matter.
Test methodology: Glen Barnett, 1992