What is it?
It’s actually quite a neat rule. The rule states, for any , of the data lies standard deviations away from the mean. This is not specific to any shape of distribution. Some probability distributions have more specific formulae to get even tighter bounding.
Formally:
