March Madness Over Data Analytics

Ken Pomeroy didn’t believe what the experts were telling him about college basketball. “Air Force was giving up the fewest points per game in the nation, and during every game the announcer would bring up their tremendous defense,” says Pomeroy, a National Weather Service meteorologist in Salt Lake City. So the self-confessed “stats kind of guy, math kind of guy” decided to challenge the conventional wisdom. He devised a formula to measure a team’s efficiency per possession, and proved his case: “It was Air Force’s offense that was driving the number of opportunities to score for each team. Their defense wasn’t very good.”

Four years and a Sports Illustrated article about his ground-breaking work later, Pomeroy runs one of the more influential sports blogs around (www.kenpom.com) and continues to redefine statistical analysis for hoops fans—and other interested parties. Last year, after George Mason University made its improbable run to the Final Four, Pomeroy got an appreciative e-mail from coach Jim Larranaga, who used analysis from the KenPom site to scout opponents.

As the NCAA tournament ramps up for another round of March Madness, Pomeroy keeps delving deeper into the numbers that help explain the game. Having added player efficiency ratings to his arsenal, he’s passionate about play-by-play data on the arcana of shot selection. So who does he like to take the title? The Virginia Tech grad doesn’t play favorites—but as the regular season neared its end, his top-rated four teams were North Carolina, Kansas, Texas A&M and UCLA.