Functions that return random results do not fit into the most common testing paradigms. I have seen people try to verify the distribution of random results, an

Functions that return random results do not fit into the most common testing paradigms. I have seen people try to verify the distribution of random results, and even that could have concerns for theoretical flakiness. I wonder if the nature of the function contributed to the apparent lack of testing coverage here.