The Science of Awards Ceremonies – Can We Predict Our Winners? by Freya Wood

Millions of viewers tune in to the Academy Awards each year: a spectacle of Hollywood glamour, couture fashion, excitement and surprise. As the last big ceremony of awards season, with the same few actors, directors and films nominated as in most award shows—are the Oscars really that unpredictable? And if we can foresee the winners, why do we avidly tune in year on year?

Mathematician Iain Pardoe has developed a statistical model capable of predicting Oscar winners with an impressive 75% accuracy. Since 2005, Iain has correctly predicted all top four award winners (best picture, best director, leading actor and leading actress) in 2008, 2009, 2013 and 2014. Data science company Farsite have also developed a dynamic model which considers static data such as a film’s genre, critical reception and cast as well as real-time award wins and nominations. This model has a slightly higher success rate at 83%. So what factors decide who wins and who goes home empty-handed?

The most significant predictive factor for Oscar winners is their success at preceding awards shows. 25 films have achieved the ‘big five’ since 1996, winning the top prize at the Golden Globes, BAFTAs, Critic’s Choice and Screen Actors Guild (SAG) awards before going on to clinch best picture at the Oscars. The SAG awards are most predictive of the academy awards due to similar voting procedures. Both the SAG and Oscars are voted for by a nominee’s peers in the film industry. Many SAG members are also members of the academy and therefore the voting pool overlaps, causing significant similarity between winners.

To win best picture, nominees generally need a best director nomination. This rule was broken by Green Book this year as Peter Farrelly failed to receive a best director nod. Best picture hopefuls also require a best screenplay nomination in order to win. However, Titanic sunk this rule in 1998 being the only film in 50 years to win best picture without a screenplay nomination.  Curiously, the number of nominations a person has received over their career is a significant predictor for best leading actor but not leading actress. Total number of nominations for a particular film is an important predictor for both best picture and director but interestingly not for either acting accolade. When it comes to best director, winning the Director’s Guild Award almost guarantees an Oscar win, recently demonstrated by Alfonso Cuarόn for Roma. Perhaps unsurprisingly, box office success is not of huge importance when predicting best picture. If it was, 2019’s winner could have been Black Panther, with winning film Green Book placing fourth.


Alfonso Cuarón with his Director’s Guild Award for Roma

A piece of advice to filmmakers—if you want to win an Oscar, don’t make an animation, foreign language film, remake or sequel. 93% of all best picture winners have been dramas, with actors nine times more likely to be nominated for a dramatic role. Furthermore, US actors dominate the Oscars, winning 4/5 prizes each year, while only comprising 2/3 nominations. The same trend is seen for British actors at the BAFTAs. But where does this bias come from? Dr Niklas Steffens from the University of Queensland explains that actors are more likely to win awards if they belong to the same social group as voters. Sharing a social group makes the audience more likely to perceive a performance as exceptional regardless of its objective quality. The existence of this bias allows us to predict that Green Book, a drama about Americans in America would win the Academy Award over Black Panther, a superhero action feature about Africans in Africa. Compiling all this information into a single model makes it relatively simple to predict which films and performers are in with a realistic shot of winning.

unnamed (1)

Olivia Colman’s charming reception of the Best Actress Oscar

If the Oscars are so predictable, why do we continue to tune in with such excitement and anticipation? Stuart Fischoff, Professor of Psychology at California State University suggests it’s down to humankind’s natural sociality. Our attention is drawn by the ‘alpha’ male and females in the group, otherwise known as celebrities. We form emotional attachments to characters and stories as we watch films and by extension, the actors that play them. Perhaps the biggest reason to watch the Oscars is the fear of missing out if we don’t, commonly known as FOMO. As social animals, we crave the approval of others and the feeling of fitting in with a group. Being unable to join in global conversations surrounding a large-scale community event is therefore undesirable. Of course, there is the possibility of surprise, and everyone loves an underdog. Take Olivia Colman for example, Iain Pardoe’s model had her lagging way behind Glenn Close in the race for best leading actress, her win was unexpected but by no means disappointing.

So, no matter how predictable the Oscars may become FOMO will make sure we continue to tune in. Perhaps recent efforts to diversify the Academy Awards will succeed in returning the ceremony to the unpredictable celebration of film and creativity it is supposed to be.

And the Oscar goes to… an American in an American drama about Americans in America. Probably.


5 views0 comments