you could start by using actual data

your 80:1 stat seems pretty dubious -- do you have a source for it?