You're making strong assumptions about his methodology vs a guy that is only looking at public ally available exits. I would be remiss to say that his data most definitely supports his position.