Browse all our box-office analysis in one place: the complete coverage index.
Last reviewed by the USA Times Data Desk. This page documents the definitions, thresholds, data sources, and limitations behind every chart and figure in our box-office coverage. It is a living reference; we update it when the method changes.
What the Star Index measures
The USA Times Star Index is a proprietary measure of box-office concentration. Instead of ranking films by how much money they make, it measures how much of the entire week’s domestic box office a single film commands. A film’s “share” in any given week is its gross divided by the sum of all films’ grosses that week. This turns a familiar question — which film was #1? — into a more revealing one: how much of the market did that film actually own? A #1 film that took 70% of a week dominated in a way a #1 film that took 28% did not, even if both held the top spot.
The three statuses
We classify every #1 film by its peak weekly share into one of three named statuses, with thresholds drawn from the golden ratio (0.382 and 0.618):
- Mini-geuse — 26.18% to 38.2% of the week. A clear leader, but one that leaves real room for the rest of the market.
- Betelgeuse — 38.2% to 61.8% of the week. The muscular middle tier: a film genuinely dominating without erasing the field.
- Geuse — 61.8% and above. The film is effectively the entire market for that week.
A film whose best week falls below 26.18% is not counted as a star movie; that week is classified as Dark (no film owned it). The lower bound, 26.18%, is 0.2618 — the golden ratio’s square — and marks the point at which a single film has taken a defining share of a week.
The qualification floor
A high weekly share can be misleading when the overall market is tiny. To prevent negligible films from registering as events simply because a given week was quiet, a film must also clear a lifetime box-office floor to qualify for the index: $26.1 million in U.S. domestic lifetime gross (0.2618 × $100M). Films that top a weak week but never reach that lifetime figure — a recurring problem during the 2020 shutdown, when tiny releases briefly posted large shares of a collapsed market — are screened out as noise rather than counted as genuine events.
Why 2020 is excluded
The 2020 pandemic closed most theaters for much of the year and distorted every share-based measure: with the market collapsed, small films posted enormous percentages of almost nothing. Because these figures describe an anomalous market rather than genuine box-office concentration, we exclude the 2020 calendar year from all aggregate analysis and from the historical charts. Individual 2020 weeks remain in the underlying data for completeness, but they are omitted from every trend, count, and comparison so that the pandemic does not contaminate the twenty-year picture.
Reading a whole week: the ten compositions
Beyond the single #1 film, we classify the shape of a whole week by the statuses of its top films. Because each status has a minimum share and a week only has 100% to give, only ten arrangements are mathematically possible — and a fourth star can never fit (four Mini-geuses alone would require 104.7% of the week). The compositions run from a lone Geuse owning everything, through pairs like Betelgeuse-Betelgeuse (two films both between 38.2% and 61.8% — a shape that has occurred only once in twenty years), down to a Dark week with no star at all. This is the basis of our “you can never have four” analysis.
Data sources
Weekly figures are drawn from Box Office Mojo’s weekly domestic charts, with The Numbers used as a cross-check and fallback where a week is unavailable. Each week’s total is the sum of all films reported that week; each film’s share is its gross as a percentage of that total. Worldwide and budget figures, where cited, are reported estimates from the same public sources and are labeled as estimates. Every figure in our coverage is traceable to a specific weekly chart; we do not model, interpolate, or estimate box-office numbers we cannot source.
How it is maintained
The index covers 2006 to the present and is updated as each week closes. The dataset, the classification code, and the charts are maintained by the USA Times Data Desk under the direction of editor-in-chief Achir Kalra. Analyses are reviewed and fact-checked against the source charts before publication, and this methodology page is revised whenever a definition or threshold changes — as it did when we adopted the current status names and the $26.1M qualification floor.
Limitations
We state our constraints plainly. Our weekly records track the top films precisely but do not capture the exact share of every minor release, so the two rarest three-film compositions cannot always be distinguished from two-film weeks. Studio, genre, and franchise attributions depend on secondary data that is not complete for every title, and any chart relying on them is labeled accordingly. The index measures concentration, not quality or profitability: a film can dominate a quiet week or make a fortune without ever leading one. What the Star Index captures is a single, specific thing — how the power of a given week was distributed — measured consistently across two decades.
Questions or corrections: contact the USA Times Data Desk. We publish corrections when we get something wrong.