Math could predict Oscar award winners
The Academy of Motion Picture Arts and Sciences has published the 2005 nominees for its prestigious Oscar film awards. The list has attracted criticism with regards to absent names as well as rampant speculation about who is the worthy winner in each category.
Decision scientist Iain Pardoe of the Lundquist College of Business at the University of Oregon explains how to select candidates in the latest edition of Chance. According to Science News, Pardoe tackles the sometimes complex decision of selecting winners in the four major Oscar awards - Best Picture, Best Director, Best Actor in a leading role and Best Actress in a leading role - from each year’s nominations.
Pardoe believes that the Oscar Awards panel do not fully exploit formal statistical analysis and that therefore the result is not always empirical.
Several factors could facilitate the decision making process; Other Oscar category nominations, previous nominations and wins and previous movie awards can all serve as predictors. In order to pick out the most significant of these, Pardoe made use of a technique known as discrete choice modeling.
With this technique an outcome is determined by several decisions. It involves a sequence of choices—made from a finite set of alternatives by individuals in the population under consideration. The probabilities are calculated using a ‘multinomial logit’ model. To demonstrate his proposed selection technique, Pardoe used data from 1928 to 1938 to make predictions for 1939, then cumulative data for each succeeding year.
Pardoe's results go some way to explaining the disagreements that usually arise after the Oscars. His study put actors who were somehow downplayed into favorable positions for winning awards in several years.
Pardoe’s study could send a positive message about the role of mathematics to people working in the film industry and media.
Copyright 2006 PhysOrg.com
