Finding
The year has a mood
The darkest films of the year and the warmest films of the year come out six weeks apart. Most of that is genre — horror in October, holidays in December, which nobody needed measuring. About two-thirds of it is not, and that part is the same shape every year.
The obvious objection, first
Group every well-known film by release month, average the twenty readings, and a shape appears immediately: a dark autumn, a warm December, a flat January. It is also close to worthless, because it mostly says that horror films come out near Halloween and family films come out near Christmas. That is a fact about scheduling genres, not about the calendar having a mood.
So every number here is a film against its own primary genre. A horror film in October is compared with horror films, a drama in January with dramas. The question becomes whether films of the same kind differ by when in the year they arrive.
They do, by roughly two-thirds as much as the raw figures suggest.
Every reading shrinks once genre is held still, and the order changes: nostalgia, not beauty, travels furthest across the year.
Two peaks, six weeks apart
Each film against its own genre. September and October are the dark peak, November and December the warm one. The shaded gap is about six weeks.
A horror film released in October is measurably darker than a horror film released in June. A drama released in December is warmer and more nostalgic than a drama released in August. The peak is October at +0.17 on darkness and melancholy, and December at +0.22 on beauty and +0.16 on nostalgia — and the dark line has already fallen back to zero by then.
Ten months at roughly neutral, then two swings in eight weeks, in opposite directions.
And then there is January
The other extreme is on neither peak. January is the year’s floor: the lowest month for nostalgia (−0.25 against its own genre) and for beauty (−0.20), and not dark to compensate. It is flat.
That is a claim worth a second opinion, and there is one that has nothing to do with the twenty readings.
Audience rating against the film’s own genre average. January sits 0.16 below it; December 0.09 above. Raw, the gap is wider still — 6.44 against 6.79.
Two measures that know nothing about each other put January at the bottom and December at the top. The dumping ground is real, and it survives the same control.
What a release date is
A date is usually discussed as a competitive decision — which weekend, which tentpole to dodge. This says it is also an editorial one, made consistently enough to read off a calendar four decades long.
The two readings that travel furthest are nostalgia and beauty, which is not what you would guess from a chart dominated by a Halloween peak. Neither is about how frightening a film is. Both are about how much warmth and craft went into it — and beauty is the strongest single predictor of whether audiences will like a film.
Read that way the year is not shaped by mood. It is shaped by confidence: the best-looking films go where the attention is, the rest go in January, and the emotional shape of the year is largely a by-product of that one decision.
What this is not
Release date here is the first release anywhere, which for a festival film is months before anyone could buy a ticket — that pulls some autumn prestige releases earlier than their cinema date. It is also a northern-hemisphere calendar applied to a worldwide catalogue: the same October is spring in Buenos Aires and Melbourne.
Controlling for primary genre is a blunt instrument. A film has several genres and only the first is used here, so some of what survives the control is still genre mix moving underneath it. The direction of that error is knowable: it inflates what is left, which is why the raw figures are shown beside it rather than quietly dropped.
And every number is an average over four decades. A single year is noisier, and the December peak has grown as awards season consolidated. The shape is the claim; one year is an instance of it.
What each individual year felt like: every year in film.