POLL: Yale student uses computer program to predict Oscar winners

(WTNH) –The odds favor a Yale Chemistry PhD student and her methodology. She figured out a way to pick winners and losers with the Oscars and her results were pretty impressive.

In a chemistry lab on Yale’s campus, Allison Walker found new data to plug into a computer code she developed. And it’s all about the Academy Awards, with their online reviews.

“What the program does is it goes online and downloads reviews of movies that normal people have written,” Walker said.

Normally, Allison Walker spends her days studying ribosomes, molecules and proteins. But after she read a blog request, posted on the popular website Fivethirtyeight.com, hosted by famed statistician Nate Silver, she thought maybe her methodology could help.

“They had a method for predicting the Oscars, but they didn’t think it was that accurate,” Walker said. “And they were wondering what others could come up with.”

They liked her idea and featured it on the website along with several other submissions. By using keywords like the phrase”all-time,” as in “all-time favorite movie” or the “best movie of all-time,” Walker’s program analyzes winners and losers for six categories the website picked. With the results now in, Walker’s method proved the most reliable, with over 80 percent accuracy.

“I got five of six categories correct. Only one I missed was Best Picture, which is probably the biggest and the one people are most interested in,” Walker said.

Some may be wondering if her methods can also be used to predict gambling outcomes or things like Powerball numbers. Walker’s advice; don’t get your hopes up just yet.

“Theoretically there should not be a way to predict the results of that,” Walker said.

Walker said her methodology could be used to possibly predict voting behavior and the outcomes of political elections. It’s too late this year, but she wants to give it a try for the next election season.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s