From our limited sampling, it happens with some measurable frequency.

Going from zero to one of something tends to be a lot less likely than going from one to two or more.