Definition
Regression to the Mean is a statistical phenomenon where, if a variable is extreme on its first measurement, it will tend to be closer to the average on its second measurement. It occurs in any process where outcomes are determined by a combination of skill/merit and Luck/Randomness.
Why It Matters
We often mistake a natural return to the average for the result of our own interventions. If we ignore this statistical inevitability, we waste resources chasing outliers and build false narratives of cause-and-effect (e.g., punishing a student after a rare bad grade).
Core Concepts
- Softening of Extremes: The best will always appear to get worse and the worst will appear to get better over time, independent of any additional action.
- The “Kahneman Headline” Trap: Misinterpreting natural regression as a causal effect (e.g., thinking “hugging a cat” cured depression because an extremely depressed child improved after three months).
- Control Group Necessity: The only way to tell a real improvement from regression is to compare the treated group against a control group expected to regress by randomness alone.
- Feedback Fallacy: If you praise a pilot for an exceptionally good landing and their next landing is worse, you may conclude that “praise doesn’t work,” when in fact the pilot was simply regressing to their average performance.