Do the Awards Get It Right? Oscars vs Our Scores

Every year somebody posts the same complaint: the Academy got it wrong again. That complaint is about fifteen per cent true, which is not nothing, but it is nothing like the story people tell. Going back through the Best Picture winners of the past fifteen years and scoring each one on the same scale we use for everything else (no deference, no hindsight bonus, no adjustment for good intentions), a much stranger pattern appears. The Academy is not a body with bad taste. It is a body measuring something slightly different from the thing a score measures, and mostly measuring it well.

The disagreements cluster. They are almost never random, and once you can see the shape of them you can predict the next one a year out.

What an Award Actually Rewards

A Best Picture Oscar is not the output of a viewing process. It is the output of a campaign that happens to end in a ballot. The voters are working industry professionals: editors, publicists, cinematographers, actors between jobs. They are asked to rank ten films they have mostly seen on screener links in the four weeks before voting closes. Two thousand people, a short window, a lot of other obligations.

So the award ends up rewarding four things that have nothing to do with quality. Timing, because a September release has to survive five months of memory while a December release arrives fresh. Narrative, because “she has been due one for twenty years” is a genuinely persuasive argument in a room of peers. Mood, because the industry votes for the film that says something it wants to believe about itself that year. And broad likability, which is a real virtue in a film and also the single easiest thing to mistake for a great one.

A score rewards none of that. Our scale asks whether a film realised its own premise, whether the craft survives a second viewing, whether the performances do something a competent actor could not, and how much of it is still in your head a month later. A film can be an excellent campaign and an average piece of filmmaking. Those two facts are not in conflict; they are just answers to different questions.

A Plurality Vote Is Not a Graded Judgement

The deeper structural difference is the arithmetic. A score is a judgement with a magnitude attached: 7.8 and 8.8 are different opinions, and the gap between them means something. An Oscar is the residue of a preferential ballot, which means a winner can emerge with something like a fifth of first-choice support. The Academy does not publish the margins, and it is a safe bet that several recent winners would look considerably less decisive if it did.

Ranked-choice voting is also a compromise machine. It systematically rewards the film that nobody actively dislikes over the film that a passionate minority adores, because the second-choice pile decides everything. That is how you end up with a winner that a plurality of voters ranked fourth. It is not corruption and it is not stupidity; it is what the mechanism is built to produce. If you asked a thousand critics to run the same ballot, they would produce a similar kind of consensus winner, because the method, not the voters, does the work.

The “Right Film, Wrong Year” Problem

A large share of the Academy’s apparent mistakes are not mistakes at all. They are correct decisions delivered late. The institution does have a genuine nose for a major filmmaker; it just tends to find them one film after everyone else did.

The clearest version is a film winning for a body of work rather than for itself. When a veteran finally takes the trophy for a film that is merely good, the vote is not really about that film; it is a correction for two or three earlier omissions the voters feel bad about. That is generous and human and it wrecks any attempt to read the award as a quality signal in that particular year.

It cuts against newcomers hardest. A director’s best film usually arrives before the industry is ready to hand them the prize, which means their fifth-best film often collects it. Fifteen years of winners contain at least four of these delayed payments and maybe three genuine, unambiguous correct calls. The rest sit in a middle band: defensible choices that were probably the third or fourth best of their year, which is a perfectly respectable thing for a vote of two thousand busy people to produce.

Where the Academy Is Sharp, and Where It Is Soft

Split the ceremony by category and the picture flips completely. In the craft categories the Academy’s record is close to excellent. Editing, cinematography, sound, production design, costume, makeup: these are voted by, or heavily influenced by, the people who do those jobs, and they judge them with genuine expertise. An upset in sound design is rare, and when the wrong film wins it is usually because the right one was judged on volume rather than construction.

Comedy is the reliable blind spot, and the reason is mechanical rather than snobbish. Comedy is the only genre that gets measurably worse on a second viewing. The first watch is doing the work; the joke is spent. In a campaign where voters watch screeners in a tired evening slot, the comedy has to be funny twice: once to be noticed, once at the ceremony. Very few manage it, so the category gets read as lightweight even in years when it is the best thing on the slate.

Horror is not so much overlooked as quarantined. It gets its craft nominations and occasionally its screenplay nod, but the top category has treated the genre as a farm team for years. Animation has the mirror-image problem: a dedicated category that functions as a polite quarantine, so an animated film that is better than nine of the live-action nominees is never actually compared to them.

The Scoreboard: Winners Against Our Scale

Here is the comparison, side by side. The winner column is the film that took Best Picture; the score column is what we gave the same film on release, unchanged since. The most recent ceremony is deliberately absent. A winner needs a year of hindsight before we are willing to hold a verdict against it.

Three correct calls out of fourteen at the top of the scale, a comfortable middle of unobjectionable choices, and exactly one outright error. That is a better record than the internet gives the Academy credit for, and a worse one than the Academy gives itself.

Won It and Didn’t Earn It, Lost It and Did

The failures are fewer than the legend suggests, but they are concentrated and legible. Green Book is the worst call of the period by a wide margin: a film that resolves a country’s racial history through the personal growth of its white protagonist, made with competent craft and no interest whatsoever in the people it is nominally about. It won in a year that contained Roma, The Favourite and BlacKkKlansman. That is not a close call or a matter of taste. It is the mechanism selecting maximum offensiveness-to-nobody.

CODA is the softer version of the same mechanism: a likeable, conventional film that everyone could live with, beating a genuinely divisive one. Nobody who votes for it will defend it in ten years. The wins for Nomadland and The Shape of Water belong in a different drawer. Not errors, just films elevated a tier above their station by the year they landed in.

The losses are where the argument gets genuinely painful. The Social Network losing to The King’s Speech remains the clearest case of the period: a formally perfect, era-defining film beaten by a competent costume drama with a speech impediment at its centre. Roma losing to Green Book is a stranger result still, because the better film was also the more ambitious one in every measurable category. And the films that never entered the conversation at all (the horror, the comedy, the animation) are losses the ceremony does not even record as losses.

What Awards Are Good For, and Where They Mislead

The practical upshot is simple. If a film wins below-the-line awards, believe it. If a film wins Best Picture, treat it as a strong recommendation from a friend whose taste you like but whose reasoning you do not entirely trust.

The Verdict

Reader FAQ