My guideline was that the probability that they would share a class with someone with the same name was below some threshold. This kept things in perspective si

My guideline was that the probability that they would share a class with someone with the same name was below some threshold. This kept things in perspective since "in the top x in popularity" wasn't very meaningful to me. And if it was on the high side, I wanted to see it trending away from popularity over the last decade.