Analyzing Oscar

As I write this, the announcement of nominations for the 80th Academy Awards still lies in the future.

Nevertheless, I can make a make a few bold predictions: the actors nominated most likely appeared in dramas from major film distributors, and have either been nominated in the past or appeared in a film starring or directed by previous nominees.

Yes, for those of you who think that scientific research is too often focused on the needlessly esoteric, I give you “I’d Like to Thank the Academy, Complementary Productivity, and Social Networks,” a working paper by Nicole Esparza and Gabriel Rossman of the California Center for Population Research at UCLA.

As the abstract states, “This paper explores under what conditions a film actor will be nominated for an Academy Award.”

It explores three questions: whether “asymmetric centrality in the network of screen credits predicts Oscar nomination,” whether the skill of the other members of the filmmaking team has “spillovers” for the actor, and whether network ties to Academy members make a nomination more likely.

The authors helpfully explain all these questions in terms of Robert DeNiro (well, why not?).

“Asymmetric centrality in the network of screen credits” is defined as, “If an actor has outranked Robert DeNiro in a screen credit does this imply high status and the career rewards that come with it?” The second question translates as, “does having Robert DeNiro as a co-star make one more likely to be nominated for an Oscar?” and the third as, “Does your having worked with Robert DeNiro in the past make it more likely that he will nominate you for an award today?”

How do you go about answering these questions? Esparza and Rossman drew on every film geek’s go-to source for detailed information about movies: the Internet Movie Database. They looked at data contained therein on 171,539 performances by 39,518 actors in all 19,351 Oscar-eligible films made between the founding of the Academy of Motion Pictures Arts and Sciences in 1928 and 2005. A total of 1,394 of those actors were nominated for Academy Awards

Their conclusions? The single best thing an actor can do to boost his or her chances of being nominated is to avoid comedies: actors are nine times more likely to receive a nomination for work in a drama than in a non-drama.

Another good move is to avoid acting in a year in which a lot of movies are made. Not too surprisingly, actors were more likely to be nominated in years when fewer movies were screened.

Women are more than twice as likely to be nominated as men, simply because there are fewer female than male performers in films.

As for the “asymmetric centrality” effect, yep, it’s real: performers who regularly placed high in the credits of their movies were twice as likely to be nominated as those who didn’t.

Having a major distributor was the fifth most likely predictor for a nomination: it also almost doubled a performer’s odds, as did having been nominated for an Oscar in the past.

To answer the second question in the abstract, yes, performers get spillovers from high-powered creative teams. The researchers dubbed this the “Robert Forster Effect,” after Robert Forster, a character actor who performed in dozens of films but only received a nomination when he appeared in the 1997 movie “Jackie Brown,” written and directed by Quentin Tarantino and co-starring Samuel L. Jackson, and—yes—Robert DeNiro.

(However, performing with such high-powered costars can be a mixed blessing. It only improves the odds of being nominated as a supporting actor, not as a lead actor, which is what happened to Forster even though he had more screen time than either Jackson or DeNiro.)

The biggest surprise in the study turns out to be the answer to the third question in the abstract: industry ties had very little influence on nominations. People who had worked for years with a wide array of Academy members were no more likely to be nominated than those who hadn’t.

For you, oh readers in my future, the nominations for this year’s Oscar are now an open book.

You have my fearless, science-based predictions in hand.

How’d I do?

These weekly columns on science appear in the Regina (Saskatchewan) Leader Post and Red Deer (Alberta) Advocate.  They are available for one-time publication or regular syndication to any interested newspapers, magazines or on-line publications.  E-mail me for details.