An average hides its evidence
Movie ratings explained properly require more than an average because 8.0 from a small fan audience is not the same signal as 8.0 from a broad population. Vote count changes confidence, while audience selection changes what the number represents.
A niche film may be watched mostly by people already interested in its subject. A major release attracts a more mixed audience and can receive a lower average despite broader satisfaction.
Separate quality, confidence and fit
Quality asks how strongly viewers responded. Confidence asks how stable that response is likely to be. Fit asks whether the film matches your mood, time and tolerance. Only fit belongs to you personally.
RankThatMovie applies Bayesian smoothing so tiny samples are pulled toward a wider baseline. It also shows a trust label and keeps momentum separate from time-neutral quality. The complete method and limitations are published on About the Radar.
Watch for false precision
The difference between 7.7 and 7.8 rarely justifies choosing one film over another when genres, runtimes and audiences differ. Treat close scores as the same neighbourhood, then use pace, premise and downside to decide.
A useful reading order
- Check vote volume and trust.
- Read the one-line editorial interpretation.
- Compare runtime and pace with tonight's context.
- Use the number as evidence, not permission.
A rating cannot promise that you will like a film. Its honest job is narrower: reduce uncertainty while leaving room for taste.