Why Does Particle Physics Require Five Sigma for a Discovery?

Particle physicists often wait for a result to reach five sigma before calling it a discovery. The threshold is deliberately demanding because experiments examine enormous datasets, test many possible signals and depend on models of instruments and backgrounds that can be imperfect.

What sigma measures

Sigma expresses how far an observed result lies from what a background-only model predicts, measured using the model's statistical spread. A larger value means the data are harder to reconcile with ordinary background fluctuations. It does not directly measure how important a result is.

Five sigma corresponds to a background fluctuation probability of roughly one in 3.5 million for a properly specified test. That number is not the probability that the new explanation is true. It is the probability of obtaining data at least this unusual if the tested background model is correct.

Why smaller signals often disappear

Large experiments repeatedly divide data by time, energy, particle type and detector region. If researchers inspect enough combinations, an apparently unlikely pattern can appear somewhere by chance. Analyses therefore account for the look-elsewhere effect when they test many models or search windows.

Systematic errors create another risk. A calibration shift, an underestimated radioactive contaminant or an unmodelled detector behavior can mimic a signal. More data reduce random uncertainty, but they do not automatically correct a mistaken model.

Five sigma is a gate, not the finish line

Crossing the threshold does not make an interpretation unquestionable. Researchers still examine whether the result survives alternative selections, whether another instrument can see a compatible signal and whether the proposed particle or mechanism fits other measurements.

The convention is most strongly associated with particle physics. Other fields may use different thresholds because their experiments, sample sizes and error structures differ. The useful question is not only how many sigma a result reached, but also which backgrounds were modelled, how many hypotheses were tested and whether independent evidence supports the interpretation.

How a hint becomes a discovery

A credible progression usually begins with an interesting excess, followed by more data collected under predetermined analysis rules. Researchers test alternative background explanations, check calibration stability and publish enough detail for others to scrutinize the method. A separate detector or a complementary experiment can then provide especially strong confirmation.

Five sigma sharply reduces the chance of celebrating a random fluctuation. It cannot replace careful modelling, transparent methods or independent replication.