Method to March Madness

If it’s March, it must be time for March Madness, the annual college basketball lollapalooza that compels untold millions of people across the country to try and predict the outcome of a 64-team tournament (not including play-in games) in hopes of winning the office or online pool and all the glory and money that goes with it. The beauty (and perhaps irony) of March Madness is that no one – not the college basketball junkie, not the office nerd, not even the analytics guru well-versed in predictive analytics who studies “bracketology” – knows the outcome in advance, so anyone has an equal chance of winning . . . or do they?

For insight on an analytics topic that involves and impacts so many millions this time of year, we talked to Michael Magazine, a longtime member of INFORMS and a decorated professor of operations and business analytics at the University of Cincinnati who also happens to teach a class on “Sports by the Numbers” (think “Moneyball”), as well as a popular business analytics course each spring on “Bracketology” with Paul Bessire of PredictionMachine.com. Professor Magazine has also served as a reviewer, session chair and presenter at the prestigious MIT Sports Analytics Conference.

So what are Professor Magazine’s key Bracketology metrics?

“Percentage of offensive rebounds. How often a team gets to the foul line. Looking at a team’s depth.”

And what are the biggest mistakes the average office Madness pool player makes?

“No question. The biggest mistake is trying to pick upsets.”

Speaking of irony, Magazine’s Bracketology course includes an in-class competition that neither of the instructors won last year. Instead, the honor went to an international student who had never seen a basketball game. What does that tell the professor about analytics?

“It tells me that analytics really does a great job of prediction,” Magazine says. “He used the analytics he learned in the class. Though many of us have a keen interest in basketball, we also have our biases, which he did not have.”

And speaking of hoops analytics, we’ve already commented on former 11-time NBA All-Star Charles Barkley’s recent viral anti-analytics rant. So what does the professor have to say about Sir Charles’ tirade?

“My comment is it matches his golf swing,” Magazine says. “He still has lots of things he can learn.”