Loose evidence and no evidence are different, and people overcorrect in both directions.

Imprint databases are mostly US-centric. A confident hit for another market is still a guess.

Half of what people describe as tolerance is a change in the underlying condition, and the two need different answers.