IIRC, the Wikipedia article for Betteridge's law stated that, statistically, headlines in the form of a question were more often answered with a "yes."

I would argue that Betteridge's law breaks down when the point of the article is to invoke readers to ask themselves the question.