Scoring Policy

Why a composite score instead of one critic percentage?

No single source captures a film's reception. We normalise up to four clearly labelled inputs, average the available values, and label the result Celluloid Score so it is clearly ours — distinct from Tomatometer, Metascore, or IMDb averages alone. A worked example and a comparison table live in How the Celluloid Score works.

Which inputs go into Celluloid Score?

Four possible inputs, equal weight on the default path: the named editor’s 0–100 assessment; Metascore when Metacritic has one; Letterboxd stars ×20; IMDb ×10. The audience snapshot is displayed for context and excluded from the composite because coverage and sample provenance vary by market. If an input is missing, it is omitted — never replaced with zero.

• Celluloid Editorial Rating — the named editor's 0–100 assessment; it is not presented as a critic poll.
• Audience Snapshot — informational only; excluded from the ranking.
• Metascore — weighted critic average (0–100) when present.
• Letterboxd — star rating converted to 0–100 (×20).
• IMDb — user rating converted to 0–100 (×10).

How do Hindi and Bollywood titles get scored?

Industry reporting has documented paid influencer reviews and coordinated positive campaigns around Hindi theatrical releases, and platforms such as IMDb have flagged unusual voting on Indian titles. When we score Hindi / Bollywood films we therefore treat Letterboxd as the primary organic public-sentiment anchor, boost Metascore when present, down-weight IMDb when it runs hot relative to Letterboxd, and apply a modest haircut so inflated local channels cannot alone create a Celluloid Pick.

• Weights on that path: Letterboxd 1.4, Metascore 1.35, editorial 0.85, IMDb 0.45, then ×0.96.
• Festival and arthouse Indian titles with strong international Metascores are largely unaffected.
• Commercial Hindi releases with sky-high IMDb numbers and middling Letterboxd or Metascore land a few points lower than a naive average.

What do the tiers mean?

Celluloid Pick is 90%+. Recommended is 60–89%. Not Recommended is below 60%. A 61% and an 89% share a tier — read the inputs. These labels are ours, not Certified Fresh.

When is there no Celluloid Score yet?

Upcoming films and released films awaiting editorial verification display no Celluloid Score and no aggregate-rating markup. Publishing requires an editorial rating plus at least one independent input (two real inputs minimum). Missing data is stored as null. A zero is never used as a placeholder. Review Pending means we have not crossed that floor — not that the film scored zero.

How is this marked up for search engines?

When a film has published Celluloid critic reviews, schema uses that review count as ratingCount on our own reviews. We never treat score inputs as a review count, and we never emit supplier aggregateRating. The composite stays visible, dated, method-cited text.

How do we keep the formula honest over time?

Inputs are verified when we publish or refresh a title. We do not bump “Last reviewed” or sitemap lastmod unless the page actually changed. Worked examples on the complete score guide use numbers from live title pages, dated. If Metacritic never covered a film, that cell stays empty. If you find a stored input that does not match the source, that is a correction, not a taste argument.

Who should skip Celluloid Score?

Anyone who needs a binary Fresh/Rotten badge (use Rotten Tomatoes), a live box-office ticker we do not run, or a score for a film that has not screened. The title page is still useful for cast and where-to-watch. The composite is the wrong tool in those cases — and we say so.