Among the best comebacks to criticism is to turn the tables on the doubter and say, "If you think you can do better, be my guest."

This, however, is a terrible plan if you are dealing with Luke Stanke.

Stanke, a University of Minnesota graduate student in the Department of Educational Psychology (Quantitative Methods in Education track), has devised a statistical model that he believes can more fairly select and seed the 68 teams that play in the NCAA men's basketball tournament.

Termed the "Win Index," Stanke's model uses weighted win-loss data to measure the relative strength of teams. Based on his research looking at the past eight years of NCAA tournament games, Stanke believes the Win Index does a better job parsing out teams than the Ratings Percentage Index, a formula for determining the strengths of teams that the selection committee relies heavily upon to pick the field.

Stanke's paper and research on the subject was presented this past weekend at the Sloan Sports Analytics Conference at MIT (and can be viewed on the conference web site, He also plans to submit his findings to the NCAA.

"The tournament committee is very highly correlative with the RPI ... but [the RPI] is obviously flawed in the way it's built," Stanke said. "It overemphasizes the strength of schedule of a team's opponents. The number they chose is very arbitrary."

Stanke, who describes himself as a "huge basketball fanatic," said he started doing research in hopes of finding a better way to predict the outcomes of NCAA tournament games but eventually started looking at different ways of picking the teams that made the field.

"It's taboo to say whether the committee is doing a good job," he said. "But this would be a better way to seed the field."

Stanke, a Green Bay native, noted that he also ran a program this past year involving NFL teams. He found the Vikings weren't as bad as their record suggested. As for the Packers?

"They were never the No. 1 team in the NFL," he said.

So you have to like his research, even if you might not fully understand it.