I checked against my notes, I see about the same numbers using the `n**2` taylor series approximation. I missed that the probability of `>=1` collision is about

I checked against my notes, I see about the same numbers using the `n**2` taylor series approximation. I missed that the probability of `>=1` collision is about the same as exactly one collision, but I suspect that's quite reasonable as this scale.