What's the real rate?

You ran a batch of trials and counted the failures — so the failure rate is kn, right? Only on average. The true rate could be lower (you got unlucky) or higher (you got lucky), and Clopper–PearsonThe exact confidence interval for a binomial proportion: built by inverting the binomial test, so the stated coverage is guaranteed for every k and n. The Rule of Three is its k=0, 95% upper-bound corner. pins down the range it plausibly sits in. Wider than the textbook ± formula, never wrong at the edges, and unlike the rule of three it works for any k, not just zero.

We sawfailures in.
0.31%4.32%
95% confidence intervalThe range the true rate falls inside the stated fraction of the time, in repeated runs. Clopper–Pearson is conservative: actual coverage is at least the stated confidence, sometimes a touch more.

Observed rate 1.5% (3 of 200 trials) — the true rate is plausibly anywhere from 0.31% to 4.32%, a width of 4.01%.

Clopper–Pearsonexact
0.31%4.32%
Wilson scoretighter, almost always fine
0.51%4.32%
Wald (normal)textbook ±, breaks at the edges
0%3.18%

The dot is the observed rate. Wald is the familiar p̂ ± z·√(p̂q̂ ⁄ n) — clean, but its coverage collapses as k approaches 0 or n (the interval can fall off the edge, or shrink to a point with no failures). Wilson is the practical default — close to Clopper–Pearson but a touch tighter. Clopper–Pearson is the conservative one: a hair wider, but coverage is guaranteed.

Hold the observed rate near 1.5% and watch the width shrink as the run grows. To halve the width you need roughly four times the trials — the √n curse.

Sample sizek (≈ p̂·N)95% interval
100
0%30.8%
width 30.8%
300
0%11.6%
width 11.6%
1002
0.24%7.04%
width 6.8%
3005
0.54%3.85%
width 3.3%
1,00015
0.84%2.46%
width 1.62%
3,00045
1.1%2%
width 0.91%
the Clopper–Pearson interval — the exact CI for a proportion, the rule of three's bigger sibling.