A test performed by Jason Mills for
White Dwarf magazine concluded that the frequency distribution of the Zocchihedron was substantially uneven, 5,164 rolls (binomial distribution dictates that to be confident of an average result on a fair dice the number of trials required (
n) is equal to
μ ÷
p, where
μ = mean, or die average, and
p = probability of a single result.
n = 5050 for a d100) led Mills to conclude that results greater than 93 or less than 8 are significantly rarer than results between these figures. Mills attributed this to the especially high, and especially low numbers being situated at the poles of the sphere — and thus being closer together. Numbers near the equator are more widely spaced.
Later versions of the Zocchihedron have been designed with a different pattern of number distribution, resulting in more even results overall. Individual numbers still suffer from the bias.