The Random Noise Rule in Statistics

There is a filtering out of things that are, you know, that have no evidence to be anything but random noise. But there's no rule that you have to pick the strongest of the remaining correlations. I would say almost, I mean, 5% has become such a standard physics,. by the way, physics uses a much, much lower threshold.