In the spirit of open debate, here's a way to falsify my theory: the 1 year stats should start deviating from the all time stats in 9 months.