Should Exist IndexResearch publications that show the cost of being wrongScientific recommendations should state the consequences of false positives, false negatives, delay, and overconfidence.