This year’s Golden Globes have come and gone (the glamour! the speeches! the surprise kisses!), but Oscar season is almost upon us. Whether you love the Oscars ceremony, love to hate-watch it, or prefer to pretend it doesn’t exist, one thing most of us can agree on is how fun it is to try and predict the winners. We’re a competitive bunch here at Luminoso, particularly when it comes to shamelessly using data to win.
So in advance of the Oscar nominations announcement on January 24th, we decided to use our natural language understanding software to predict which film will win Best Picture and win this year’s Oscar pool. As one does.
Predicting the Oscars: A messy affair
Of course, Oscars predictions are not the same as science and can be difficult to get right. Predictions are typically based on which films and performances people think the voting members of the Academy of Motion Picture Arts and Sciences will like best, based largely on past trends and reviews from critics with similar viewpoints.
Another element that is almost impossible to quantify is the impact of elaborate (and expensive) "For your consideration" events. These marketing events are hosted by the studios behind the nominated films to extol the virtues of their movie and to try and sway the Academy members’ votes. A nomination can often translate into increased viewership and tens of millions of dollars in additional box office revenues.
We decided to take a different approach in making out prediction... one based more on moviegoers’ reactions the films themselves rather than guesstimates about the Academy. We pulled over 84,000 reviews from average moviegoers - not critics - on IMDB covering the past four years (2013 - 2016), focusing on the top 50 most popular movies from each year. We then used Luminoso Analytics to analyze the text in those reviews and identify correlations between topics discussed in the reviews and the eventual Oscar nominees and winners.
Incidentally, we only analyzed reviews written before each year’s Oscar nominees were named -- why, you ask? Because once a movie is nominated, the language and overall sentiment contained in the reviews changes, and therefore can skew the results of the analysis. And since the 2016 Oscar nominees have not yet been named, we wanted to compare apples to apples.
Picking a winner
We found that certain terms, including “narrative,” "cinematography," "plot," “visuals,” “stunning,” "experience," and “masterpiece,” were more prevalent in reviews of movies that later went on to be nominated and/or win the Oscars. Not exactly surprising, of course.
What we found even more interesting was that the more detailed and sincere the reviews were, the more likely a movie was to ultimately win Best Picture. Surprisingly, considering a movie to be “Oscar-worthy” had exactly zero impact on whether that movie would win a nomination or the award for Best Picture.
So that was Part One finished! We’d developed an algorithm that could accurately predict the Best Picture winner of past Academy Award ceremonies, given how each movie was talked about in moviegoers’ reviews. Next, we started on the next phase - using that algorithm to predict the winner of this year's Oscars, given the reviews posted on IMDB prior to the announcement of the nominees.
And the Oscar goes to...
Given how moviegoers have talked about the most popular movies of 2016, we predict that the winner of the Academy Award for Best Picture will go to...
"Jackie," starring Natalie Portman.
Here’s our full list of 2017 Oscars contenders, ranked in order of likelihood to receive a nomination according to our algorithm:
“La La Land”
“Hell or High Water”
“Manchester by the Sea”
Of course, our combined prediction methodology, while extremely scientific, cannot completely account for the whims of marketing efforts and the impact on voting. If Jackie does not win Best Picture, that will indicate that the winner's marketing efforts outperformed the film's actual virtues and may not have won on its own merit.
Still, we’re penciling “Jackie” into the slot for Best Picture in this year’s Oscars pool - how about you?