The big four: discrete choice modelling to predict the four major Oscar categories

The present study formulates regression models that predict the four major Oscar categories (Picture, Director, Actor and Actress). A database was created, collecting publicly available information from 2005 to 2016. The approach taken was to apply discrete choice modelling. A remarkable predictive accuracy was achieved, as every single Oscar winner was correctly predicted. The study found evidence of the crucial role of directors, the predictive power of box office, gender discrepancies in the film industry and the Academy’s biases in the selection of winners related to the film genre, nominees’ body of work and the portrayal of actual events