Yeah, I messed that one up. The case where the upper bound is larger than 10x the lower bound is weird.

I like the randomness increasing slightly slower though.