Grouping the last 229k nostr events by their pubkey gives me 7539 that ever only sent one event and 1 that sent 7932 events.

The following is a scatter chart with each spot (x, y) representing the number of pubkeys x that posted y events.

While this chart looks very orderly, I suspect it's the result of these frequencies being almost too perfectly random. The regression line probably has something to do with random numbers repeating exactly in the range >1000 being far less likely than <10. In the first range they have to be equal +-0.05% and in the second range only +-5%.