Regression to the mean is one of the most important and most ignored concepts in sports betting. It describes the tendency for extreme performances to be followed by more average ones.
The Core Concept
When a team or player performs much better or worse than their underlying ability, future performance will tend to be closer to their true average. Extreme outcomes include luck — and luck doesn't persist.
Betting Applications
Shooting efficiency in basketball: A player or team shooting 60% from three over 3 games is likely to regress. Their "true" three-point percentage is probably 35-40%.
Turnover-driven wins/losses: A team that won 3 games in a row with 8 forced turnovers will likely see fewer next week. Betting on them to continue doesn't account for regression.
Quarterback passer rating: Elite QB performances in one game rarely repeat in the next at the same level.
Using Regression to Find Value
When public bettors chase hot streaks, lines move toward the streaking team. Fading these teams — when the streak is driven by variance rather than true quality improvement — is a systematic edge.
Look for:
- Shooting/completion percentage well above career norms
- Turnover differential well above sustainable rates
- Special teams touchdowns inflating a team's point differential
[Track your betting record on regression-based picks with Oddible →]
