Every year since 2017, I have run a statistical model to predict the winner of the Oscars. The predicted probability of each movie or actor in the following categories for this year’s awards are below. Percentages don’t add up to 100% because the models calculate the independent probability of each person or movie to win. The model is based on data collected since 1995.
My models use variables like whether an actor or movie won a Golden Globe, BAFTA, or other award; the age of an actor; and how many previous nominations or wins an actor has to figure out which variables are the most predictive of winning an Oscar. In each category, it turns out that different variables are highly predictive of winning. For example, for Best Actress, winning the SAG Award the same year is one of the best predictors for winning the Oscar, but for Best Director, winning the Director’s Guild Award is a good sign for winning the Oscar. Then I generate predicted probabilities for each actor and movie in the major Oscars categories, plugging in data like whether they just won or were nominated for awards this cycle and other information. The models that guess at the percentage likelihood of each movie or actor winning are below.
Ryan J. Williams 