when a person tells you they are very confident—“100%”, “95% sure”—of a conclusion on a disputed matter about which they’ve no special information, the rational response is to downgrade your estimate of the reliability of the source, not upgrade your estimate of the reliability of the conclusion.