Sport keeps a public score of almost everything, repeats the same contest thousands of times a year, and records who did what to whom. That makes it the cleanest data anyone has on human performance, and the entries in this subject show what the data have done to the stories sport tells about itself. Several of those stories turned out to be luck, regression or selection.
Stories that turned out to be statistics
The Sacking Bounce is the clearest case. Results improve after a manager is dismissed, and they improve about as much at comparable clubs that kept their manager, because clubs sack at a low point and low points tend to be followed by recovery. Momentum in Sport and Icing the Kicker face the same question, whether the effect is real or the ordinary clustering of random sequences, and the answer so far is small or unclear.
The Relative Age Effect is a story about selection rather than luck. Children born just after a cut-off date are heavily over-represented in elite squads, because a few months' extra growth is mistaken for talent at the age when coaches choose. The Peak Age Curve has a selection problem of its own: players who decline early leave the sample, which makes the decline look gentler than it is.
Separating skill from luck
A second group consists of tools for measuring what the score hides. The Elo Rating System updates a rating by how surprising a result was. Pythagorean Expectation estimates a team's true strength from points scored and conceded, which predicts future results better than the win record does. The Paradox of Skill adds the uncomfortable corollary: as competitors become uniformly excellent, luck decides more of the difference between them.
The Moneyball Effect closes the loop. Oakland's statistical edge in baseball disappeared once rivals copied the method, which is the same pattern economists describe in markets: an advantage that can be seen can be competed away.
Leagues as small economies
The modelling entries treat a league as a market with unusual rules. The Uncertainty of Outcome Hypothesis holds that fans want close contests, which is the standard justification for drafts and salary caps. Rottenberg's Invariance Proposition, four years before Coase, argued that talent ends up where it is most valued whichever way the rules assign it. Tanking shows the cost of rewarding the worst teams with the best draft picks, and Rosen's Superstar Effect explains why small differences in ability produce enormous differences in pay.
Where it applies
Most of these entries have direct counterparts off the field. A board that replaces its chief executive after a bad year should expect a recovery whether or not the new one is better. A school that streams children at five meets the relative age effect in its own classrooms. An investor meets the paradox of skill, since a professional market full of capable managers leaves luck deciding most of the ranking, and the Moneyball effect, since a public edge does not last. Any contest designer who rewards failure with advantage meets tanking.
Where it is argued over
Sport's own data have unsettled several claims here. The hot hand, long treated as an illusion, turned out after a 2018 correction to be small but real. Marginal Gains, credited with British cycling's success, is hard to separate from the large funding increase that arrived at the same time. The uncertainty-of-outcome hypothesis has weak and inconsistent support, since dominant teams often coincide with record revenue. Matches played behind closed doors in 2020 gave a natural experiment on Home Advantage: studies found it shrinking in empty stadiums, and the tilt in referees' decisions shrinking with it, which supports the crowd as a cause.
Why it is worth knowing
Sport is where people most readily explain outcomes by character, momentum and leadership, and it is also where those explanations have been tested most thoroughly. The subject teaches a habit that is hard to learn elsewhere: before crediting a person or a method with a result, ask how much of it luck, regression and selection would have produced on their own. The answer is usually more than anyone watching assumed.
The entries behind this essay: Sport & Performance. Every subject has its essays on the essays page.