Feature Web Extra: Maggie Wigness’ Amazing Algorithm

When Pacific University computer science major Maggie Wigness '10 presented research at the 2009 New England Symposium on Statistics in Sports at Harvard University fall semester, she was one of only three students to do so.

Her presentation, "Using New Iterative Methods and Fine Grain Data to Rank College Football Teams," explores an algorithm that aims to more accurately rank NCAA Division 1 football teams. An algorithm is a precise rule or set of rules specifying how to solve a problem. Maggie's algorithm has greater than 70 percent accuracy. Compare that to the Bowl Championship Series' algorithm at 56.6 percent.

The idea to challenge the Bowl Championship Series (BCS) rankings came from Mike Rowell, professor of Mathematics, and Chadd Williams, professor of Computer Science at Pacific. Both professors describe themselves as football fanatics. Maggie, on the other hand, likes football but isn't obsessed. "Maggie came with a perfect background for a project like this," Rowell told The Oregonian. "We were interested in someone who knew football and could come in with a completely unbiased view."

"Mike and Chadd really just asked me if I wanted to do summer research with them," Maggie says. "I'm planning on attending graduate school next fall, so the thought of getting some research experience was intriguing."

In layman's terms, Maggie's ranking method uses play-by-play statistics that can be calculated throughout the football game. These statistics measure factors including team achievements, offensive and defensive statistics, the number of interceptions, sacks and safeties and average yards gained per play.

With articles in The Oregonian and online at Fox Sports, FitsNews.com and SBS Sport, the added publicity was certainly a surprising start to her senior year. Maggie says, "I'm trying to handle the publicity as best I can. I really didn't anticipate that my busy fall schedule would be in part due to the research I did this summer."

-- Jessica Cornwell '10

Related Images

You may click on any image above to see it at full size.