## The Iron Law of Statistics
Regression to the mean is the statistical phenomenon where extreme performance tends to be followed by performance closer to the average. It occurs because extreme observations are partly caused by genuine quality and partly by random variation — and random variation does not persist.
## Sports Examples
**The striker on a purple patch:**
A striker has scored in 6 consecutive matches. His form is celebrated as exceptional. But: some of the goals were from outside his normal xG range (low probability shots that happened to go in). As the random component of scoring regresses, his goal rate normalises. The question is whether the underlying xG rate (the genuine component) is better than baseline.
**The goalkeeper with 3 clean sheets:**
A goalkeeper who has kept 3 consecutive clean sheets is praised as "in form." But clean sheets are heavily team-dependent. If the defence in front of him has not genuinely improved, the clean sheet run likely reflects favourable opponents and random variation — not a step change in goalkeeper quality.
**The team that conceded 0 goals in 5 matches:**
Investigate the xGA (expected goals against) over those 5 matches. If xGA was high but actual goals was zero: the defence benefited from exceptional goalkeeper performance and/or opponent misses. Regression toward the xGA trend is likely.
## The Market Opportunity
Markets often fail to adequately account for regression. After an extreme positive run, prices tighten excessively — overvaluing the team. This creates lay value or opposition backing value for the analytically rigorous bettor.
Rule: always check the underlying statistical rate alongside the headline result. If results and underlying rates diverge, regression of results toward rates is the base expectation.
Create a free account to track your progress and save bookmarks.
