That person might well think they can control the dice, but if we were to re-test them on 100 rolls they would likely not be so lucky. If 100,000 people were to roll the dice 100 times each, it's very likely we find at least one person that only rolls 2 or 3 sevens total. here).īeyond that, the more people that try, the higher the chance that we find someone with lots of evidence for being a non-random roller.

Even after 100 rolls with zero sevens, there's still a small chance this could happen for a random roller (i.e. The big problem is that we can never conclusively know that someone is influencing the dice. If that probability is low, that gives evidence someone is controlling the dice. If a shooter throws X sevens in N rolls, there's math to show what the probability of that happening is for a truly random roller (we call this a p-value). The simplest might be how many sevens a shooter throws in a specified number of rolls where they are trying to avoid the seven. In terms of what possible proof would be, that could fall under a variety of statistical tests.

I don't think anyone claims 100% control over the dice, it's usually a matter of implying they might have a lower seven-to-rolls ratio than random. This is basically a statistics question since it's about evidence under uncertainty.

Why do casino dice have sharp edges