Definition
Regression to the mean means extreme results tend to move back toward a more normal average over time. In betting, a team, player, or stat that has been far above or below its usual level should not be projected as if that hot or cold stretch will continue unchanged.
There is no single betting formula for regression to the mean. The math comes from comparing a recent result to a longer-term baseline, then adjusting the projection toward that baseline.
Worked Example
A basketball player averages 20.0 points per game over the season but scores 34, 31, and 29 in three straight games. A sportsbook posts his next points prop at 27.5 with a −110 price on the over.
At −110 odds, a bettor risks $110 to win $100. The break-even probability is:
110 / (110 + 100) = 52.38%
If the bettor’s projection pulls the player back toward his 20.0-point season average and lands at 23.5 points, the over 27.5 does not clear the −110 break-even requirement.
Why It Matters
Regression helps bettors avoid overpaying for recent streaks and spot inflated lines after outlier performances. It is most useful when the market reacts strongly to a small sample while a larger, more reliable average points the other way.