A data-driven map of film history

Why another "best movies" site?

Most rankings on the internet are one of two things: a single publication's editorial opinion, or a user-vote aggregate that rewards recency and franchise loyalty. Both are useful; neither tells you what the whole conversation of film history has concluded. That conversation happens across many rooms at once — the once-a-decade Sight & Sound poll where critics and directors vote separately, the American Film Institute's periodic canon-making, festival juries at Cannes, Venice and Berlin, the Library of Congress choosing which films the United States will physically preserve, Japan's Kinema Junpo critics voting every year since 1924. Each of these lists is a fact you can check. What has been missing is a transparent way to read them together.

That is what this site does. We collect the placements — "Vertigo was #2 in the 2022 Sight & Sound critics' poll," "Casablanca won Best Picture at the 1943 ceremony," "Stagecoach entered the National Film Registry in 1995" — weight them by the authority and structure of each source, and sum them into one composite score per film. The score is not an opinion about whether you will enjoy a film. It is a measurement of canonical standing: how firmly a film is held by the institutions, polls and communities that decide what gets remembered.

How to read the rankings

Every list page on this site is ordered by that composite score, never by any single source's order. Under each title you'll see the strongest individual facts behind its position, so a ranking is always auditable at a glance: click through to any film's page and you get the complete accounting — every list appearance, every award result, and exactly how many points each contributed. If you disagree with a placement, you can see precisely which sources put it there. The full formula, including tier weights, rank decay and how we handle winners versus nominees, is documented on the methodology page.

The library grows in waves — new lists, new festival years, new registries — and every page recomputes automatically when the underlying data changes. Decades, genres, national cinemas, directors and film-history movements are all cut from the same single table, which means the 1950s page, the film noir page and the Billy Wilder facts will never contradict each other. One measurement, many views: that is the whole idea.