Most people ask whether medications and supplements are good or bad, but the right question is whether a specific intervention works for a specific person with a specific problem. Peter Atiyah explains how to define health problems clearly, evaluate evidence properly, understand baseline risk, and determine what's actually worth taking based on job classification—disease treatment, symptom relief, risk reduction, or optimization—each requiring different evidence standards and risk tolerance. Define problems with specific metrics, thresholds, and timelines before choosing interventions; vague goals like 'better health' almost guarantee false positives and wasted money.