The mechanical rolling and computer vision is impressive, but wish he had noted if the results were statistically significant. Just a final touch to the article. At 100-150 rolls per face, my gut is saying no. I guess I'm going to have to whip up a couple loops in python after work to satisfy my curiosity.
You can't assign significance to a sample size alone.
A sample size of 10 can be significant (if it were 10 1's for example), and a sample size of 100,000 can be not significant, for example if you were to roll 4,953 1's.
One of the dangers, especially with statistical tests of uniformity, is that they often over detect departures from uniformity when using large samples.
