That's the second worst. The worst is a hundred people, all convinced of the same thing, doing various arbitrary experiments until they get ones that kind-of su

That's the second worst. The worst is a hundred people, all convinced of the same thing, doing various arbitrary experiments until they get ones that kind-of support their claim, and then publishing ten of those to say they have "evidence".