There's a enormous gulf between "occurred by chance" and "occurred because of tampering with the footballs".
The Colts' footballs and Patriots footballs were stored on opposite sides of the field, in different conditions, presuming. Moreover, the within-set variation between air pressures of the Patriots' balls is fairly large already.
I see nothing here that suggests that the calibration of this measurement process is fine enough or reliable enough to draw conclusions of tampering.
Please don't try to use a probabilistic argument to prove cheating must happened. You cannot use probability to prove something did happen. The most you can ever do is show that a result is either likely or unlikely according to a given model.