We have a highly unusual Oscar season on our hands. To start, the Best Picture frontrunner—the Iranian-hostage thriller Argo—doesn't make any sense. Argo is looking to become the first movie since Driving Miss Daisy (1989) to win Best Picture without even being nominated for Best Director. Stranger still, Argo could become the first film since Grand Hotel in 1932 to win only one Oscar—the one for Best Picture.
Indeed, it is an award season without much precedent. Last year, in my Academy Awards analysis that ended up getting every single category wrong, I tried to match up which categories track well with which other categories in terms of both wins and nominations. As Argo implies, that approach is pretty much out the window this year. Based on the other trick that Oscar prognosticators use—precursor awards like the Golden Globes and SAG Awards—a modest favorite has been identified in almost every category, but because of the lack of any historical support for many of the nominees, there is a sense that many of these favorites could be subject to upsets. (Or, after all this uncertainty, they could also all just win. That would be a bit of an anti-climax, though.)
So, this year, I figured I would take a look back at recent Oscar history and examine the circumstances that usually surround Oscar upsets—in essence, to see whether we can predict the unpredictable. Indeed, literally every single year, there is at least one category where the widely acknowledged favorite falls flat on its face.
According to my analysis of the past 10 Oscar ceremonies and pundit coverage thereof, the category that has ruined the most Oscar pools is Best Cinematography. A whopping six out of 10 times (in 2003, 2005, 2006, 2007, 2010, and 2011), the conventional wisdom has been wrong about this award—so much so that you might actually be better served betting against this year's favorite, Life of Pi. Interestingly, most of these upsets may be explained by an aversion to two specific cinematographers: Roger Deakins and Emmanuel Lubezki. Between them, Deakins and Lubezki share 15 Oscar nominations without a single win, despite being heavily favored to take home the gold in 2005 (Lubezki, Children of Men), 2007 (Deakins, No Country for Old Men), 2010 (Deakins, True Grit), and 2011 (Lubezki, Tree of Life). Deakins is nominated again this year for lighting the brilliantly shot Skyfall, but that film is seen as a distant second in the race. Furthermore, Life of Pi is shot in 3D, which has done well in this category recently (Avatar, Hugo). Although these are typically famous last words in this category, I don't expect an upset here.
Next, Best Original Song has produced four upsets in the same time frame, though none since 2006. In 2002, Eminem's "Lose Yourself" beat Bono's "The Hands That Built America"; in 2004, "Al Otro Lado del Río" triumphed amid a murky field of superstars; in 2005, "It's Hard Out Here for a Pimp" capitalized on a vote split between "In the Deep" and "Travelin' Thru"; and in 2006, "I Need To Wake Up" beat a triad of Dreamgirls melodies. What these upsets teach us is that the Academy isn't afraid to honor random films in this category; none of those films was nominated in more than one other category. It also demonstrates a perhaps surprising proclivity for hip hop. Neither of those criteria is particularly helpful in predicting this year's winner, but it's worth noting that Skyfall, the strong frontrunner here, has four other nominations this year. If Adele somehow loses for that title song (which I doubt), perhaps Ted's "Everybody Needs a Best Friend" best fits the upset bill?
Best Adapted Screenplay also has a 40% rate of surprises. The most recent came in 2009, when Precious upset Up in the Air; before that, you saw three straight surprises in 2002, 2003, and 2004. The common thread here may be that Adapted Screenplay can be more like Original Screenplay than it's given credit for. Often, the award for Best Original Screenplay goes to the year's funky, scrappy, original film as a sort of consolation prize—the Academy seems to be saying, "Well done. You deserve something, but you're not mainstream enough to be Best Picture." (Think Eternal Sunshine of the Spotless Mind, Juno, and Midnight in Paris.) In Best Original Screenplay, this attitude is usually foreseen by pundits, so it never turns out to be an upset. But in Best Adapted Screenplay, pundits continually choose one of the top two Best Picture contenders—yet this is wrong almost half the time. Sideways in 2004 (a comedy that fits the Original Screenplay profile perfectly) and 2002's The Pianist (which turned out to be beloved by the Academy, winning three major awards) stole this Oscar away from the eventual Best Picture winner, also nominated in the category. This year, Argo—and secondarily Lincoln—are the Best Picture favorites and thus the favorites to win Best Adapted Screenplay. However, recent history suggests that there's a coin flip's chance that the statue will go to comedy Silver Linings Playbook. There is a strong contingent of voters that love it, as evidenced by its eight nominations, and Argo and Lincoln could split the vote. Silver Linings also won this same award at the BAFTAs last weekend, a major precursor indication.
Four categories experienced upsets in three of the past 10 years: Best Foreign Language Film, Best Original Score, Best Sound Editing, and Best Editing. Given its reputation for upsets, it will actually probably surprise many close Oscar watchers that Best Foreign Language Film is, at best, the fourth-most volatile category. It has this reputation thanks to an unusual voting requirement that winnows the voting pool to a much more select few than in most categories—making it more unpredictable. Another category that has used the same system over the past 10 years (although this year it is opening voting to the entire Academy membership), Best Documentary Feature, also had only two upsets during that time. While there have been high-profile losses in Best Foreign Language Film before (The White Ribbon in 2009; Waltz with Bashir in 2008; Pan's Labyrinth in 2006), they're just not that much more common than in "normal" categories. Therefore, I fully expect Amour and especially, with the Documentary democratization, Searching for Sugar Man to conquer their respective categories.
The other three three-fers aren't particularly noteworthy, as we are getting into the zone where categories aren't particularly unique (or prolific) in their upset tendencies—it's just random variation and luck of the draw. In Best Sound Editing, two of the three times when the winner came as a shock (2007 and 2009), it was when pundits were predicting a split with Best Sound Mixing—but instead the same film won both awards. This shouldn't be surprising, since the Academy as a whole tends to be pretty ignorant about the sound categories, especially the difference between mixing and editing. Upsets can therefore occur when they confuse the two and just mark the same film on their ballot twice. This year, Les Misérables is leading the field in Sound Mixing, but that's no help for predicting Sound Editing—Les Misérables isn't nominated there!
It is also worth a mention that Academy voters seem to dislike Alexandre Desplat, whose two regal go-arounds as the frontrunner for Best Original Score (The King's Speech and The Queen) both saw him go home empty-handed; he was beaten by more experimental fare (The Social Network and Babel, respectively) both times. Desplat is nominated this year for Argo, so don't count on a win here for the Affleck film; the haunting Life of Pi is the favorite anyway.
As for Best Editing, action films tend to win this category overall, and two-thirds of the upsets were indeed action films (The Girl With the Dragon Tattoo in 2011 and The Departed in 2006). However, the other third was The Aviator defeating the more action-y Million Dollar Baby in 2004. With the action-packed Argo the frontrunner in this category as well, it seems safe, unless Zero Dark Thirty can knock it off its perch.
All other categories (excepting the shorts, for which not enough information was available to analyze) have had only one or two upsets over the years—a pretty normal tally, since pundits can't be right all the time. (An interesting footnote is that, technically, there has not been an upset victory in Best Costume Design for the past 10 years. However, all you have to do is go back to last year, when the race was too close to call, to see the limitations of this stat; there can't be an upset if there is no favorite!)
When we get to this level, perhaps we're better served looking at the circumstances when an upset has occurred rather than the categories it has occurred in. One easily discernible pattern is when a candidate garners a last-minute surge in support. Therefore, it can be instructive to see who appears to be gaining momentum in the last days before the ceremony; if someone goes from a long-shot to a dark horse in the last week, they may peak at just the right time and end up taking home the prize. (It's the exact same phenomenon as last year's Iowa caucuses, when Rick Santorum entered the final weekend surging, but still behind, in the polls but ended up winning the state by 34 votes.) The classic example of this comes from the 2005 awards, when Crash was perceived to have pulled into a tie with Brokeback Mountain for Best Picture; it ultimately triumphed, of course. It's also thought that Eddie Murphy's extremely salient role in Norbit turned off Academy voters so much that they turned around and gave his 2006 Best Supporting Actor trophy (for Dreamgirls; it would have been well deserved) to Little Miss Sunshine's Alan Arkin. And, going back farther, Russell Crowe's boorish behavior probably created a backlash that caused the heavy 2001 Best Actor favorite to succumb to Denzel Washington. Unfortunately, however, these trends are much harder to identify in the technical categories, which are seldom subject to such trackable, day-by-day gossip.
Surprises also often happen when a category is a total tossup, allowing a third candidate to take advantage of the opening created by so many split voters (again, you're probably familiar with this from politics). This happened most memorably at the 2002 ceremony, when an undercurrent of support for The Pianist was able to break through in the Best Actor category (split down the middle between Daniel Day-Lewis and Jack Nicholson) and Best Director category (Rob Marshall and Martin Scorsese). Likewise, controversial director Tim Burton may never have enough mainstream support to win a majority of Oscar votes, but a big enough chunk of the Academy loves him that he has been able to capitalize when the mainstream vote has been split. In 2007, his Sweeney Todd snuck in between co-favorites Atonement and There Will Be Blood for Best Art Direction, and Alice in Wonderland pulled off the same trick, in the same category, in 2010 (victimizing Inception and The King's Speech). The Girl with the Dragon Tattoo was the most recent beneficiary, snagging a Best Editing award last year despite 12-to-1 odds (!). As you may recall, that was one of the categories I analyzed in last year's piece, where I found it to be a two-way race between The Artist and Hugo.