How Oscar’s Preferential Ballot Works — And Could Produce a Best Picture Shocker

Predicting the best picture Oscar winner with any degree of confidence has become considerably harder over the past decade — and we can thank The Dark Knight for that.

In 2009, when the Oscar nominations were announced and Christopher Nolan‘s acclaimed 2008 comic book adaptation was not among the five nominees for best picture — bounced, in all likelihood, by a Holocaust movie, The Reader — it sparked outrage that convinced the Academy to take action. The organization’s board of governors voted to expand the best picture category from five nominees — at which it had been capped for 65 years, since the year after Casablanca won best picture — to 10 in the hopes of increasing the likelihood that a popular film like The Dark Knight would be nominated for the top prize in the future. (A few years later, the Academy decided to change its approach yet again so that anywhere from five to 10 nominees could end up nominated for best picture.)

With the expansion of the best picture category, the Academy also changed the voting method to determine the winner. The organization realized that a polarizing film could, in a year in which votes were really spread around, conceivably win with the support of only a small percentage of members, and that seemed wrong. A film like 2011’s The Tree of Life, for instance, was, in my judgment, beloved by some members, but strongly disliked by many more. It ended up losing to The Artist, but it could have won had the Academy not implemented a form of voting, for the best picture category only, that it had previously employed pre-1944, when it last had more than five nominees for best picture.

That system, brought back in 2009 and still employed to this day, is referred to as the “preferential ballot.” The point of it is to ensure that the best picture winner is the movie that is the most widely liked by the electorate.

How does that work? Members are asked to rank all of the best picture nominees from best to worst — and then the rest is handled by PwC, the Academy’s longtime accountants. PwC begins by sorting through the best picture ballots and creating piles for each film listed in the top spot on a ballot. This year, for example, there will be one pile in which Roma is listed as top film, another for Green Book, etc.

Unless a single film dominates by appearing in the No. 1 spot of more than 50 percent of all ballots — which can be difficult to do in a field of eight nominees, like there is this year — PwC then removes the film that has the smallest pile of No. 1 votes. But to ensure that that film’s supporters still have some influence on the outcome, PwC redistributes the ballots in that pile according to which film each ballot lists as its No. 2 choice. So if a ballot from that pile lists A Star Is Born as its No. 2 choice — that ballot is added to the A Star Is Born pile.

This process can continue for several rounds — the film with the smallest pile of ballots is eliminated and its ballots are redistributed according to its second-place choice, or its third-place choice, if the second-choice film has already been eliminated, and so on. That continues until one film’s pile accounts for more than 50 percent of all ballots.

Some voters think that they can avoid helping films they don’t want to win by only listing a few of the nominees on their best picture ballot. But that is incorrect. If a voter only lists three titles, but those three titles are eliminated before tallying is finished, then that ballot is discarded and has no influence on subsequent rounds of voting.

Is the preferential ballot causing films to win best picture that would have lost if the Academy were still using a traditional popular ballot, with only a single line for a single choice? We can never know for sure, since PwC doesn’t release vote totals, but there is one big reason to believe it is: a marked uptick in the number of splits between best picture (chosen using a preferential ballot) and best director (chosen using a popular ballot). Consider four of the last six years: Life of Pi‘s Ang Lee won best director, but Argo won best picture; Gravity‘s Alfonso Cuaron won best director, but 12 Years a Slave won best picture; The Revenant‘s Alejandro G. Inarritu won best director, but Spotlight won best picture; and La La Land‘s Damien Chazelle won best director, but Moonlight won best picture.

So how might the preferential ballot impact the 91st Oscars? Well, the two presumptive best picture frontrunners, Roma and Green Book, are actually rather divisive — people tend to either love them or hate them — which means they probably would have fared better with a popular ballot than with a preferential ballot.

That doesn’t mean they can’t win with a preferential ballot; it just makes winning harder. Films that probably would not have fared well if voting were conducted using a popular ballot — BlacKkKlansman, Black Panther, A Star Is Born, The Favourite and even Bohemian Rhapsody — actually stand a real shot at beating them thanks to the preferential ballot. Many voters like them a lot, even if they don’t love them enough to put them at No. 1, and might, therefore, list them at No. 2 or No. 3, which could be enough to put them over the top.

Even so, most pundits are sticking with Roma (me among them) or Green Book as their projected best picture winner because one simply cannot know, with any real confidence, which of the other nominees will wind up with the smallest pile after the first round and/or what supporters of that film — say, Vice — are likeliest to list on their ballot at No. 2. It’s all enough to drive one crazy — and to ruin one’s Oscar pool!