Segment width is the weight. The green fill is a real film: Coco, scored 95 of 100, tier S. The biggest question the rubric asks is the widest one: does the protagonist matter because of what they do, or what they were born into?
Audiences love both films almost equally. The rubric does not, because it is not measuring whether a film is good. It is measuring what the story's structure endorses: who gets to matter, who solves the problem, and what counts as the prize. Where the two rankings disagree is exactly where the interesting films live.
Scores view
Every film plotted: audience score against rubric total, tier bands S through F behind the points, bubble size is inflation-adjusted gross. Search any film to zoom straight to it.
Rank view
Audience rank against rubric rank, dots colored by tier, with the disagreement mapped into four named quadrants.
Each film is scored by Claude against a single rubric file, six dimensions with anchored scoring bands and named reference films, and every score ships with a written justification you can read on the chart.
The rubric measures narrative structure only. A film is never penalized for its rating, violence, or language. An R-rated film with an ordinary protagonist who earns their ending can outscore a G-rated film about a chosen one inheriting a throne.
This is a discovery tool, not a verdict. The weights above are one parent's values written down. Disagree with them, write your own, and the same machinery will score every film against yours instead. The capability, not any particular ranking, is the story.