Science is used to predict the Oscar winners. The data scientists at Farsite, the advanced analytics division of ICC, has the data to show who and what will win on Sunday night. In the days leading up to the March 2nd Oscars telecast, Farsite will detail the science behind its predictions and provide commentary at the Farsite Forecast and on Twitter at http://www.Twitter.com/FarsiteForecast.
“Our predictions are far more than lucky guesses. Most people are surprised to hear that the same sophisticated, predictive modeling we use in industries like retail and healthcare can predict Oscar winners quite accurately,” says Ryan McClarren, Chief Science Officer at ICC. “And while social media buzz may be high this week for Leo DiCaprio, sadly, our data shows he is not going home with a statue on Sunday.”
And the winners are…
The data shows that Matthew McConaughey will win best actor for his role in the movie Dallas Buyers Guide; Alfonso Cuaron will win best director for the movie Gravity; and 12 Months a Slave will win the coveted prize for best picture – which is the closest among all the races. The awards will not be a clean sweep for any particular picture, although the other award winners are expected to be Jared Leto for best supporting actor in Dallas Buyers Club; Cate Blanchet for best actress in Blue Jasmine; and Lupita Nyong’o for best supporting actress in 12 Years a Slave.
Last year, Farsite—largely alone among prognosticators—correctly predicted Christoph Waltz would win best supporting actor for Django Unchained. According to some, Farsite also went out on a limb with its pick of Argo for best picture, but they were correct again, despite the age-old rule that the best picture Oscar typically goes to the best director. While Argo director Ben Affleck was not nominated for best director, in what many considered a snub, his movie did win.
Farsite uses a first-of-its kind data-modeling tool to predict Oscar winners. The model analyzes more than 40 years of film industry and Academy Award related information to forecast probabilities for the winners. This information includes real-time data and an array of variables, including total nominations, other Guild nominations and wins, buzz and nominees’ previous winning performances.
According to Farsite data scientists, there are three factors to the model for predicting Oscar winners: current awards, buzz and prior performance. The first factor to consider is that during awards season there are other award winners that can provide insight into likely Oscar winners. The second factor to consider is the momentum or buzz behind particular nominees. Since Oscars are at the end of awards season, if the results of previous awards and other buzz by the glitterati are strong enough to sway the votes of some members of the Academy, then signals that historically point to the “correct” nominee could be hijacked. The third factor is the history or prior performance of the nominees. Some nominees may have an edge given their past.