Oscars: See How Microsoft’s AI Predictions Stacked Up

Courtesy of Microsoft Al

The 91st Academy Awards on Sunday was full of unpredictable moments — from Melissa McCarthy and Brian Tyree Henry livening the presentation for best costume design to Green Book winning the most important award, best picture. But in a night full of twists and turns, there were some moments that played out exactly as forecasted.

The Hollywood Reporter and Microsoft teamed up to project who would take home an Oscar at the 2019 Academy Awards through an online prediction poll called Awards Predictor powered by Microsoft AI.

Microsoft’s Bing predictions technology has projected winners of the Super Bowl, World Cup and many other entertainment and sporting events. Using the same Microsoft AI technology, Awards Predictor used a ranking system to predict each category's winner.

THR then brought the conversation to Twitter so film fans could vote for their favorite nominees in five categories: best director, best actor, best actress, best picture and best original song.

Microsoft’s AI predicted Alfonso Cuarón for best director, Rami Malek for best actor, Glenn Close for best actress, Roma for best picture and “Shallow” for best original song.

Similarly, Twitter predicted Cuarón for best director, Malek for best actor, Roma for best picture and “Shallow” for best original song, but Lady Gaga for best actress.

Bohemian Rhapsody nabbed early victories during the ceremony and four wins overall, for best sound editing, best sound mixing, best film editing and, as predicted, best actor for Malek’s portrayal of Queen frontman Freddie Mercury. Despite a great awards campaign, Roma fell short on best picture but took home three Oscars for Cuarón’s vision in cinematography, direction and best foreign-language film.

Another stunning upset occurred in the best actress category, where Olivia Colman won for her role as Queen Anne in The Favourite. Unsurprisingly, after an intimate performance between Bradley Cooper and Gaga, “Shallow” won best original song.

To see all the winners and predictions, visit the Awards Predictor site at

This is branded content, produced by our marketing department in partnership with our advertisers—not by editorial.