You're thinking about extremes and not the general data. I get we talk about the interesting skews but that's bad science on your part.