Big Data Predicts Big Screen's Oscar Winners
The 85th Academy Awards are set to take place Sunday, and as usual everyone from big-time film critics to your mailman has their predictions for who the winners will be.
Columbus, Ohio-based data sciences company Farsite thinks it has the answers, thanks to a sophisticated predictive model that taps the power of big data. Founded in 2007, the seven-employee company builds statistical models to help its clients, including the Dick's Sporting Goods retail chain and Ohio State University's Wexner Medical Center, predict revenue and determine customer patterns that can be incorporated into loyalty programs. If its Oscar picks prove correct, it may lend some pop-culture credibility to a technology trend that many still find bewildering.
Related: Branding Lessons From the Oscars
After the Golden Globes last month, Farsite began building an Oscar-prediction model based on past Academy Award winners, the results of awards shows leading up to the Oscars and the odds from online betting markets -- "part of the wisdom of the crowd," says Farsite co-founder and chief executive Michael Gold.
But it isn't enough simply to crunch the numbers, says Conor Gaughan, Farsite's chief strategy officer. "It requires a data scientist that has an understanding of the industry to look at correlations and then dissect what could be causing certain outcomes in order to make a smarter model over time," he says.
One goal of this stunt, Gold says, is to forecast the winners accurately. The other is to create an opportunity to discuss the promise of big data for the entertainment industry. "We figured what better way to do that than to tackle an example that 40 million Americans will tune in to watch?"
Farsite's Oscar Picks:
Best Picture: Argo (45.6 percent probability)
Best Director: Steven Spielberg (78.6 percent probability)
Best Actor: Daniel Day-Lewis (57.5 percent probability)
Best Actress: Jennifer Lawrence (65.6 percent probability)
Best Supporting Actor: Christoph Waltz and Tommy Lee Jones (split decision) (44 percent and 35.4 percent probability, respectively)
Best Supporting Actress: Anne Hathaway (94.3 percent probability)
Update: Farsite correctly predicted the Oscar for all major categories except Best Director and Best Supporting Actor. While there was no split decision in the Best Supporting Actor category, as Farsite had forecasted, it's worth noting that the winner, Christoph Waltz, was Farsite's top pick at 44 percent probability. And while Ang Lee won Best Director, beating Spielberg (Farsite's pick) and others for the award, Lee was Farsite's second choice for the award by probability.