Definition
A confidence interval is a range that is likely to contain the true value you are trying to estimate. In betting, that value might be a team's true win probability, a player's average rebounds, or a model's expected edge.
A common 95% confidence interval for a sample average is:
sample average ± 1.96 × standard error
The interval does not prove the true value is inside the range. It shows how much uncertainty remains in the estimate.
Worked Example
A bettor tracks 100 similar -110 bets from a model. The model's average projected edge is 3.0 percentage points, with a standard error of 1.2 percentage points.
The 95% confidence interval is:
3.0% ± 1.96 × 1.2% = 3.0% ± 2.35%
So the interval is:
0.65% to 5.35%
At -110 odds, the breakeven probability is:
110 / (110 + 100) = 52.38%
If the model projects 55.38%, the estimated edge is 3.00 percentage points, but the confidence interval shows the true edge could be much smaller.
Why It Matters
Confidence intervals help a bettor separate a strong estimate from a noisy one. They are useful when sizing bets, reviewing model results, or deciding whether an apparent edge is large enough to trust.
