Calculate a confidence interval for a measured feature availability, freshness-pass, or validity rate. The Wilson interval remains useful near 0% or 100% and reports the observed rate, margin, and plausible lower and upper bounds.
x is the counted successes, n is the sample, and z is the selected confidence critical value.
Use the main result together with the supporting bounds, counts, or capacity figures. The estimate is only as reliable as the input definitions, sampling process, and operating assumptions.
Sampling uncertainty is only one source of uncertainty. Make sure checks cover representative entities, feature groups, times, and serving paths.
With 970 successes from 1000 observations at 95% confidence, the observed rate is 97.00%. The calculator applies the Wilson formula to report the corresponding lower and upper bounds.
It uses the Wilson score interval for a binomial proportion, which behaves better than a simple symmetric interval near 0% and 100%.
No. The counted successes must be between zero and the total number of observations.
A higher confidence level creates a wider interval. Use the level specified in your analysis plan; 95% is a common general-purpose choice.
No. A narrow interval indicates low sampling uncertainty but does not remove systematic tracking, selection, or modeling bias.
Interval overlap is not a complete significance test. Use a direct two-proportion comparison when the decision depends on a difference between groups.
| Input | Role |
|---|---|
| Successes | Numerator of the measured proportion |
| Total | Independent observations in the sample |
| Confidence | Coverage target for repeated samples |