Statistics is based on a model. It becomes invalid when the model becomes invalid. Here, we had a new phenomena that was not included in the existing model.

Statistics is based on a model. It becomes invalid when the model becomes invalid. Here, we had a new phenomena that was not included in the existing model. It was not "statistically impossible". Rather, it was "not covered by existing statistical model".