Why we double-check bad news and accept good news

Bad result? Check the tracking, sample, seasonality and whether Mercury is in retrograde. Good result? Put it in the board deck before lunch.

The bathroom scale joke hurts because smart organisations do the same dance with better fonts. A campaign misses and analysts excavate subgroups until one smiles. A test wins and suddenly nobody asks whether the sample was adequate. One furious customer is “an outlier”; one compliment becomes “clear validation”. Scepticism changes shift depending on whether the result threatens somebody’s plan. Calling that rigorous doubt is generous; it is conclusion protection with statistical accessories.

Decide the rules before seeing the answer. Record the metric, threshold, sample, exclusions and what action follows. Then force favourable results through the same interrogation as disappointing ones. Ask what you would say if the number pointed the other way. Invite someone whose status is not tied to the outcome to review the exception. Pre-commitment leaves motivated reasoning alive while making the rule-changing visible—and embarrassment is a useful control. Stop granting pleasant evidence a diplomatic passport while unpleasant evidence spends three hours at customs; distrust is not the objective. Step on the scale once. If you dislike the number, change the behaviour rather than the floorboards.

Behavioural principles

Behavioural ideas at play in this post

Short, plain-English explanations of the principles behind this post, with links to related books and examples in the archive.