The nominations for the Academy Awards were announced this morning, and based on the data going in — from gambling sites like Paddy Power, which published early odds for top category winners, and Gold Derby, which aggregates support among Oscar experts — there were some pretty significant snubs and upsets. Is it weird that the best pre-Oscars odds come from Irish gambling sites? Sort of! But following the crowd’s money has proved to be a good strategy in the past.
One of the biggest upsets of the morning was in the supporting actor category. While Mark Ruffalo (“Spotlight”), Tom Hardy (“The Revenant”), Mark Rylance (“Bridge of Spies”) and Sylvester Stallone (“Creed”) were all favored by Paddy Power (which set odds for winning an award, not for being nominated),1 Christian Bale’s nomination for “The Big Short” was totally remote, as of data pulled midnight before the nominations were announced. That means Idris Elba (“Beasts of No Nation”) and Michael Keaton (“Spotlight”) lost out big. Elba had 6-to-1 odds of winning it all, according to the bookmakers, which gave him the third-likeliest score behind Stallone and Rylance. And although Keaton was a relative long shot to win, at 8-to-1 odds, he was still considered a more probable candidate than Ruffalo, Hardy or Bale.
Among directors, Ridley Scott, whose film “The Martian” just won a Golden Globe for best comedy or musical, can’t be thrilled that he was snubbed in the directing category — neither can the gamblers who bet on him with 2-to-1 odds to win the category. Instead, Lenny Abrahamson’s direction of “Room” made the cut; Paddy Power oddsmakers had him as an absurd 100-to-1 shot to win the category.
Rachel McAdams (“Spotlight”) was likewise a remote possibility in the supporting actress category — again with 100-to-1 odds of winning — but managed to get a nod in a strong group alongside front-runner Rooney Mara (“Carol”), Kate Winslet (“Steve Jobs”), Alicia Vikander (“The Danish Girl,” though her performance in “Ex Machina” was also a potential nominee), and Jennifer Jason Leigh (“The Hateful Eight”).
Paddy Power didn’t have a betting section for screenplays, so to assess those nominations, I turned to Gold Derby, a site that finds people who are preternaturally gifted at picking the winners and polls them. Although most of the nominees were in line with Gold Derby’s predictions the night before the nominations were announced, the biggest snub has to be of “Steve Jobs,” which didn’t receive a nod in the adapted screenplay category.
Not all the categories included shockers, though. The nominations for best actress and best actor came in exactly as the gamblers expected. Brie Larson (“Room”) and Leonardo DiCaprio (“The Revenant”) were strong front-runners, according to the pre-Oscar odds, in their respective categories.
Of the eight nominees for best picture, only “Brooklyn” was something of a surprise. Bookmakers at Paddy Power gave it 66-to-1 odds, which means that it beat out other on-the-bubble movies like “Star Wars: The Force Awakens” (50-to-1 odds), “Joy” (50-to-1 odds), and “Inside Out” (66-to-1 odds). All the other nominees — “The Big Short,” “Bridge of Spies,” “Mad Max: Fury Road,” “The Martian,” “The Revenant,” “Room” and “Spotlight” — were largely in line with expectations.
Still, we won’t be flying using only the knowledge of Irish gamblers much longer. Now that we have the nominations, the FiveThirtyEight Oscar model is slowly gurgling to life. We will look at how films perform at other award shows — weighted by the accuracy of those awards in predicting outcomes over previous years — to determine the state of the race. While it’s still very early in the cycle and we have far from enough data to make a true estimate, we’ll have an update from the model next week.
Even more exciting is the bake-off we’re organizing. We’re almost done recruiting eight amateur modelers who are taking a stab at developing an innovative statistical model to predict the Oscars. Over the course of the awards season, we’ll talk to these Oscar nuts about their take on the biggest races of the year, and hopefully in the end, we’ll find cool new ways to approach this tough prediction problem. The first post in that series — which looks at how the Internet can help us predict the Oscar winners — drops early next week, so stay tuned!