The Invisible Force That Fools Everyone
Regression to the mean is perhaps the most pervasive statistical phenomenon in everyday life — and the most consistently overlooked. Discovered by Francis Galton in 1886 while studying the heights of parents and children, it explains why the children of very tall parents tend to be shorter than their parents, why sophomore slumps follow rookie sensations, and why "miracle cures" seem to work when administered after symptoms peak.
Why Extremes Regress
Any measured outcome is a combination of a stable component (true ability, underlying condition) and a random component (luck, measurement error, daily variation). Extreme observations are extreme partly because both components aligned in the same direction. On remeasurement, the random component is just as likely to push in either direction — so the extreme total tends to moderate. This is not a causal force pulling things to the center; it is a simple consequence of randomness.
Real-World Consequences
Regression to the mean has profound implications for evaluating interventions. Speed cameras are installed at accident hotspots, and accidents decline — but they would likely have declined anyway, because the hotspot was partly a statistical anomaly. Students who score lowest on a pretest improve the most on a posttest — not necessarily because they learned the most, but because their low scores partly reflected bad luck. Without a control group, regression to the mean masquerades as a treatment effect.
Galton's Original Discovery
Galton noticed that while tall parents tended to have tall children, the children were on average less extreme than their parents. He initially called this "regression toward mediocrity." Crucially, the same phenomenon works in reverse — children of short parents tend to be taller than their parents. The population distribution stays stable across generations because regression works symmetrically. This simulator lets you witness regression in action across repeated measurements.