Different disciplines accept different probabilities of Type I error. 8000000 to one against would not be considered worth reporting in particle physics. Any co

Different disciplines accept different probabilities of Type I error. 8000000 to one against would not be considered worth reporting in particle physics. Any confirmed supernatural phenomenon would be a change in our understanding of the universe more dramatic than any particle physical discovery, so really it should be held to even stricter standards. Randi is being generous.