The Oscars explained: how the voting actually works

Every year the Academy Awards produce at least one result that seems to contradict the season that preceded it — a film that swept the precursors loses the top prize, or a category splits in a way nobody predicted. These outcomes are usually attributed to taste or politics. More often they are arithmetic, because Best Picture is counted differently from everything else.

Who votes

The Academy is a membership organisation of several thousand film professionals, divided into branches by craft — directors, actors, editors, sound, and so on.

Two rules follow from that structure. In the nomination round, members generally nominate within their own branch: directors nominate directors, editors nominate editors. In the final round, the whole membership votes on every category.

This produces a well-known asymmetry. Nominations reflect the judgement of practitioners in that craft. Winners reflect the judgement of the entire membership, most of whom do not practise it. A sound mixer’s nomination is chosen by sound mixers and awarded by everyone.

The counting method that changes everything

Most categories are simple: the nominee with the most votes wins. A film with intense support from thirty per cent of voters takes the award even if the remaining seventy per cent preferred something else.

Best Picture is not counted this way. It uses preferential counting, where voters rank the nominees in order.

If no film has a majority of first-place votes, the film with fewest is eliminated, and its ballots are redistributed to whichever remaining film each voter ranked next. This repeats until something crosses fifty per cent.

Why that produces different winners

Under preferential counting, being ranked second or third by a great many voters is worth more than being ranked first by a passionate minority.

A polarising film — adored by a quarter of the membership, placed last by another quarter — performs poorly. Its first-place votes are not enough to win outright, and when other films are eliminated their ballots redistribute elsewhere, because voters who disliked it ranked it low.

Meanwhile a film that few voters name as their absolute favourite, but that almost everyone places second or third, accumulates ballots through every round and wins.

This is the single most useful fact for predicting Best Picture. The question is not which film is most loved; it is which film is least objected to. That is why the winner sometimes feels like a compromise — structurally, it often is one.

The nomination round has its own quirk

Nominations also use a preferential system, designed so that a film with a dedicated constituency can secure a slot even without broad support.

This is why the nominee list frequently includes films that clearly have no chance of winning. They were nominated by a mechanism rewarding concentrated enthusiasm, and will then be judged by one rewarding broad acceptability. The two rounds select for opposite qualities.

What campaigns are actually doing

Awards campaigns are often described as buying votes, which misunderstands the constraint. With thousands of members and dozens of eligible films, the binding problem is that most voters will not watch most nominees.

Campaigning is largely an effort to ensure a film is seen, and seen in a favourable context — screenings with discussions, accessible viewing links, a narrative that makes the film feel significant.

Under preferential counting this has a specific target. A campaign does not need to convince voters that a film is the best of the year. It needs enough of them to place it third rather than eighth, which is a far lower bar and explains why campaigns emphasise craft, sincerity and importance rather than superiority.

Why precursor awards mislead

Guild and critics’ awards are widely treated as predictors, and they are useful, but most use plurality counting. They answer “which film has the most first-place support” — precisely the question Best Picture does not ask.

A film can win several precursors on concentrated enthusiasm and lose Best Picture to something with wider, milder approval. When that happens the narrative is usually about a late surge or a backlash. Frequently nothing changed at all; two different questions were asked and produced two different answers.

The takeaway

Nothing here suggests the Oscars are objective, and the counting method does not remove taste, campaigning or bias. What it does is explain a recurring pattern that is otherwise attributed to mystery: the film that wins Best Picture is frequently not the one with the most admirers, but the one with the fewest detractors — and that is a deliberate design choice, not an accident.