The movie "Crash" defied not only critical expectations, but also mathematical data when it won the Academy Award for best picture.

To predict certain Oscar winners, assistant professor of Decision Sciences Iain Pardoe created a statistical formula based on a pre-existing mathematical model.

Using the "discrete choice model," which predicts the likelihood that a certain choice will be made, Pardoe made predictions on who would win best actor, best actress, best director and best motion picture.

Three of Pardoe's four predictions were accurate, with the exception of best picture.

Pardoe said the discrete choice model is a formula that weighs differing factors against each other. For his Oscar predictions, he weighted candidates based on their previous performance in film competitions, such as victories at the Golden Globe Awards and Screen Actors Guild Awards.

The analysis also took into account other Oscar nominations in key categories, total number of Oscar nominations for the nominated movie and previous Oscar nominations and wins.

He then factored these variables into a mathematical formula that calculated a probability of the actor, director or movie winning an Academy Award.

The result was a percentage, and Pardoe chose the highest percentages as his picks.

According to his predictions, it was 94.8 percent probable that Philip Seymour Hoffman would win as best actor in a leading role for "Capote," and Reese Witherspoon, who portrayed June Carter Cash in "Walk The Line" had a 75.9 percent probability of winning best actress in a leading role.

"Brokeback Mountain" had a 91.1 percent probability to win best picture, and there was a 90.5 percent probability that the film's director, Ang Lee, would win best director.

Winners of an academy award are chosen by the 6,000 members of the Academy of Motion Pictures Arts and Sciences who vote to decide who most deserves an award, according to the AMPAS Web site.

English professor Kathleen Karlyn, who also teaches film studies classes, said that although she did not make any predictions, she was not surprised that "Crash" won best picture. Even though "Brokeback Mountain" was considered the favorite, she said the actual winner was an easier movie for the academy members to pick.

Karlyn said that she liked all the movies that were up for awards, and that she was pleased with the winners and nominees.

"I'm happy to see this kind of swing towards more thoughtful provocative movies," she said.

Pardoe said he started working on his formula, which he developed because he wanted to find a different application for the discrete choice mode, last summer, and this is the first year he has had the chance to use it.

In his research report Pardoe said that using discrete choice model to predict who wins an Academy Award has more importance than just satisfying an academic interest.

Winning an academy award increases the chances actors or directors have of getting hired and how large a salary they can command, said Pardoe.

"This knowledge might lead to recommendations about how to improve the Academy's selection and voting process," said Pardoe in his report.

Contact the business, science and technology reporter at bsaxton@dailyemerald.com

Article Tools