Improbability Principle

The milk bottle opening behavior may serve as the All-or-None example I referenced earlier, since there isn't ordinary tit behavior which gets in the way of observation. That is, we wouldn't expect to see tits attempting to pierce or remove bottle caps in the absence of learning/imitation/modeling about how to get cream. This becomes relevant to the OP, as All-or-None examples don't rely on our poor intuitions about probability, and so may be more reliable.

"All-or-None" is a term which refers to a particular kind of evidence in medicine. Either historically "all" subjects experienced the same outcome (commonly, death) and now some do not, or historically some subjects experienced a particular outcome and now "none" do. This potentially offers high level evidence. An example would be the initial treatment of infections which historically were often fatal, with penicillin.

A number of the claims made for psi would be amenable to the All-or-None idea - those claims which don't involve any sort of muddling with ordinary events/experiences - such as levitation. Of course, the difficult part comes with establishing that it's the extraordinary aspect which is taking place, rather than various combinations of trickery (inadvertent or otherwise).

Linda