How a second choice can change the Best Picture race

Start with a question a ranking can answer

Imagine choosing one film for a shared movie night. Each person ranks the options rather than handing out scores. Someone may rank a demanding drama first and an easygoing comedy second. That second position does not mean the comedy received half a vote, or that the voter thinks it is mediocre. It identifies a preference if the first option is no longer available.

I find this a better entry point than calling a film a “consensus winner” before seeing any evidence. Consensus is a proposed explanation of an outcome, not a measurable property you can read from a poster. To use the term responsibly, you need to say whose preferences you mean and what makes you think they extend beyond first choices.

Try a twenty-person example

Here is an invented, deliberately simplified ranked-choice exercise with three films: Harbor, Lantern and Orchard. Twenty people participate. Harbor receives eight first preferences, Lantern seven, and Orchard five. No film has more than half of the first preferences. Harbor leads, but a lead is not yet a majority.

In this exercise we eliminate the smallest group, Orchard, and inspect those five ballots' next remaining preferences. Four choose Lantern and one chooses Harbor. The totals become Lantern eleven and Harbor nine. Lantern wins this demonstration even though it began second. All twenty people still contribute only one active vote. We have moved ballots, not created extra votes.

Change one assumption and the story changes

Now keep the first count exactly the same but reverse the transfer pattern: four Orchard voters prefer Harbor and one prefers Lantern. Harbor ends with twelve and Lantern eight. The opening totals have not changed, yet the outcome has. This is why a confident forecast based only on loud first-choice support leaves an important question unanswered.

The exercise also shows the limit of the phrase “second choices decide everything.” Transfers depend on the first count and the alternatives still available. Had Harbor already secured a majority in this demonstration, no transfer would be needed. The ranking order matters, but it matters within a counting procedure—not as a free-floating popularity score.

What Green Book and Parasite actually demonstrate

The 2019 split is a good antidote to treating Directing as a dress rehearsal for Best Picture. The two awards recognize different achievements, and their nomination fields are built through different processes. Green Book was not a Directing nominee that year. That fact makes an absolute rule—no directing nomination, no Best Picture victory—untenable. It does not establish the opposite rule that missing a directing nomination helps a film.

Parasite's paired wins in 2020 are equally important. A split is possible, not compulsory. We should resist explaining every result with whichever story fits afterward: broad affection when the prizes split, overwhelming passion when they align. Neither archive reveals voters' lower rankings. Historical results are valuable for disproving rigid claims, but explaining a particular count requires evidence the winner list does not contain.

The practical consequence is to evaluate a Best Picture candidate on its own terms. A directing nomination can be a meaningful sign of recognition. Its absence is a weakness to examine rather than a mathematical disqualification. If your forecast changes, say which assumption changed; do not turn a single precedent into a permanent law.

Keep the two Oscar rounds on different pages

The Academy's voting overview distinguishes preliminary, nominations and final voting. Best Picture nominations draw on eligible members across Academy branches. Other categories can have branch-specific or additional eligibility requirements. Those are institutional facts; the prediction you build on them is your own interpretation.

Reaching the lineup and winning from it are different tasks. A campaign that creates enough enthusiasm to enter a field has accomplished something important. The final comparison then takes place among a smaller set of choices. A nomination forecast therefore needs a fresh argument before it becomes a winner forecast: which competitors remain, and how might voters compare them?

Use the mechanism without inventing the ballots

For a real race, separate three kinds of evidence: evidence of enthusiastic support, evidence of wider acceptance, and what remains unknown. A public award result belongs in the first two only with a description of the electorate and procedure behind it. A social-media poll belongs in a separate audience-observation column. Neither gives access to the Academy's confidential ranking order.

A productive conversation can end with uncertainty. “I can see two plausible paths” is more useful than claiming a film must win because it seems least divisive. Write down what evidence would change your view. That turns the ballot from an excuse for a favorite prediction into a tool for asking better questions.

A ranking and a rating can disagree without either being wrong

A five-star review and a first-place ballot answer different questions. A reviewer may give three films the same high rating because all three clear a personal standard. A ballot asking for an order requires a comparison between them. That comparison can turn on something small: the film the voter most wants to revisit, the one whose ending remains unsettled, or the work that feels most complete. Equal admiration does not force an equal ranking.

The reverse also matters. Somebody can place a film second while feeling only moderately enthusiastic about it, because the remaining alternatives appeal less. Reading second place as an enthusiastic endorsement can therefore exaggerate what the rank tells us. A ranking contains relational information, not an absolute measurement of pleasure. The position is meaningful only alongside the alternatives the voter was asked to consider.

This is one reason public review scores should not be poured directly into an imagined Academy ballot. A collection of ratings comes from a particular audience using particular scales; an award ballot belongs to a defined electorate choosing among specified nominees. Scores can help describe reception, but converting them into rankings adds assumptions about people, comparison and participation. Name those assumptions before treating the result as a forecast.

Try the system at a watch party without losing the point

For a small group, use films everybody has actually seen and agree on the counting method before collecting choices. Ask each participant to rank the complete set in order. Count first preferences, look for a majority and, in this simplified exercise, transfer ballots from the lowest remaining option until one title has a majority. Publish a tie procedure in advance; a coin toss agreed beforehand is less contentious than inventing a favorable rule after the count.

Then discuss both the result and what the count conceals. How many people began with the eventual winner? Which transfer changed the lead? Did a film that divided the room still have a plausible path? Keep the ballots anonymous if participants prefer that, and avoid turning the exercise into a demand that people justify their taste. The interesting object is the collective mechanism, not a trial of the least popular voter.

Finally, compare the winner with a simple first-preference result and with the group's favorite film to debate. Those can be three different titles. That difference reveals why awards are only one way to organize a conversation about art. A ballot is useful when a group needs to choose one option. It becomes less useful when it is expected to compress every kind of merit, influence and emotional attachment into the same answer.

A majority does not mean everybody agreed

Return to Lantern's eleven votes in our example. It has crossed the required majority in that simplified exercise, yet nine voters still prefer Harbor at the end. Some of Lantern's supporters originally wanted Orchard. Calling the result unanimous would erase both groups. Calling Lantern nobody's favorite would be equally wrong: it started with seven first preferences.

This matters when discussing an unpopular result among your friends. A winner can emerge from a valid process without being the choice of every community, critic or viewer. The ballot answers a collective selection question. It cannot settle whether the film is the most adventurous, most moving or most likely to last. Those remain worthwhile critical arguments, and they become better arguments when we stop asking a trophy to resolve all of them.