Nate Silver can't predict the Oscars

The awards show may not lend itself to statistics the way polls do -- which could be a problem for Silver's new gig

By Daniel D'Addario
Published July 23, 2013 12:00AM (EDT)
Nate Silver                              (AP/Nam Y. Huh)
Nate Silver (AP/Nam Y. Huh)

Nate Silver's portfolio has expanded quite a bit.

The journalist has left behind The New York Times, where his FiveThirtyEight blog was focused largely on electoral politics, for ESPN and ABC, where he'll do everything from commenting on sports to offering on-air predictions as to who will win the Academy Awards (ABC's most important and consistently popular non-sports franchise).

There's a problem with this plan, though: Silver isn't particularly good at picking the Oscars, as has been noted elsewhere. His 2009 predictions, most notably, saw Silver state that Taraji P. Henson was most likely to win the best supporting actress prize; she appeared in "The Curious Case of Benjamin Button," for those who've forgotten. She had never been a big part of the Oscar conversation that year, but for Silver, her film's best picture nomination was a more meaningful data point than it had been for Oscar voters. (Indeed, many acting nominees, like Jacki Weaver in this year's "Silver Linings Playbook," appear to be dragged into the race by enthusiasm for their film in general but are never considered front-runners -- but then, we have no real polling data.)

It turns out there are some things that data-driven journalism can't help decode. Silver's particular bugbear in the period after his rise to ultra-fame following the 2012 Presidential election has been critiquing the journalism of "narrative," the day-to-day tracking of airy political currents that only pundits can detect. He's gone after the Wall Street Journal's Peggy Noonan and Politico for their focus on elements of the race that drew attention away from the cut-and-dried numbers.

But the Oscars are literally all narrative. "The thing that's so fascinating about the Oscar race is how much emotion plays into it," said Nathaniel Rogers, a film and Oscar-predictions blogger whose site is The Film Experience. "It depends on feeling the moods in the air -- which is much more volatile with a group of 6,000 than an entire nation. Statistics --" among them, Oscar nomination tallies as well as the awards given by organizations less exclusive than the Academy (e.g., the Screen Actors Guild) or made up of entirely different people, like critics or the Hollywood Foreign Press -- "don't account for what people say to one another at parties, the hermetically sealed industry stuff."

In Silver's defense, he takes into account how irrelevant, e.g., the Golden Globes are, but the whole thing is alchemy. His calling "Argo" as the most likely best picture winner this year was prescient. But a film like "Argo" winning the best picture prize without a best director nomination is historically quite rare, another set of statistics that Affleck overcame with eager campaigning (Silver's method hinged not on observing Affleck's glad-handing but on weighting precursor awards based on their historical accuracy). "Narrative is why Ben Affleck won [for producing 'Argo'] when statistics would tell you that movie would not win best picture," Rogers said. Silver also acknowledged the lack of meaningful data in the best director race, which he called wrong, saying, "You can’t claim to have a data-driven prediction when you don’t have any data."

"How you can really predict is by talking to voters," said Sasha Stone, editor of Awards Daily. "If you were on the ground [in 2006], the buzz was around 'Crash,'" a film that was largely passed over for many precursor awards and had fewer Oscar nominations than frontrunner "Brokeback Mountain." "You can't do that with Presidential elections -- which we saw, because, for a time, the buzz was with Romney." In other words, the sort of talk-to-a-voter reporting that created so many skewed narratives during the 2012 presidential campaign actually helps the Oscar predictor.

Some of Silver's other writing at FiveThirtyEight has used statistics in outright bizarre ways -- for instance, using "Bayesian reasoning" to suggest that Julian Assange's accusers in his alleged rape case were politically motivated. But this may change at ESPN and ABC: Silver's time at the nation's most prominent newspaper has been marked by an odd sort of division where what goes on at FiveThirtyEight has little to do with what goes on at the rest of the paper, as documented by public editor Margaret Sullivan's column indicating Silver did not fit in there. The best sort of data journalism, said WNYC data journalist John Keefe, supports narrative journalism rather than existing separately from it. "Almost everything we do on the data news team comes from editors in the newsroom. It's almost entirely been in support of narrative journalism that the rest of the team is doing. [Data journalism] helps you find the story -- usually it's not the story itself."

In other words, it's not quite the narrative-spinners-versus-number-crunchers divide Silver has portrayed in his public remarks in most modern newsrooms. "I think anybody who thinks there's going to be a demise in narrative journalism or story journalism is going to find themselves on the wrong end of a prediction," Keefe said.

Silver's role at ESPN and ABC, whatever happens with the Oscars, is likely to draw attention to the sort of rational, data-based thinking he has advocated throughout his career. "I only see it as a positive," said Paul Ford, a writer, programmer and news analyst. "He's somebody who cares about statistics and cares about rigor and doesn't belong to some horrible think tank and isn't a robot. Anything that tries to be analytical and scientific can only help."

When it comes to much of Silver's portfolio, it's hard to disagree; given that he's working at ESPN, he'll likely be able to deflate some of the more irritating woo-woo thinking about teams' "momentum" that disregards hard-and-fast statistics. But the rapid rise of "Argo" this past year after statistics were, at first, very much not in its favor; or the "Crash" shocker; or voters finally tiring of Meryl Streep being an Oscar also-ran in 2012 and not a moment before; are more difficult to reduce in such a manner.

And that's to say nothing of the fact that the existence of bloggers like Stone and Rogers means that, on the Academy Awards beat, Silver is for the first time in his career not a trail-blazer. "Everybody I know who's in the Oscar business," said Stone, "says they're the Nate Silver of the Oscar race."

Daniel D'Addario

MORE FROM Daniel D'AddarioFOLLOW dpd_

Related Topics ------------------------------------------

Abc News Ben Affleck Espn Nate Silver Sandra Bullock The New York Times The Oscars