The statistical phenomenon where extreme measurements tend to be closer to the average on subsequent measurements. This occurs due to natural variation and measurement error, not because of any real change in the underlying trait or ability.
First described by Francis Galton in 1886 when studying the heights of parents and children, originally called 'regression toward mediocrity.' The term comes from Latin regressus (a going back) as extreme values 'go back' toward the average.
This explains why the 'Sports Illustrated curse' seems real and why extremely tall parents often have shorter children! What looks like decline or improvement is often just statistics returning to normal.
Complete word intelligence in one call. Free tier — 50 lookups/day.