Comment by gradus_ad

I think it's working as intended. The year is simply a randomly chosen year not related to the density of births in that year. It would be a boring tool if it actually chose "random year based on random birth" because almost every choice would be from 1900 onward...