A decade ago, when baseball GMs started using statistical analysis to evaluate their teams, it revolutionized the game (see 'Moneyball'). Now, the numbers-based approach is coming to basketball. Muthu Alagappan, a 22-year-old Stanford undergrad, was crunching cancer-related numbers for a data analysis firm when he got the idea to turn his attention to hoops. Last summer, at MIT's Sloan Sports Analytics Conference in Boston, Alagappan presented his potentially groundbreaking model for studying the pro game – a software system that takes traditional stats (rebounds, points, blocks, etc.) and maps them spatially, revealing how seemingly disparate players actually have similar skill sets and styles of play. "I found similarities between players that I hadn't thought of before," Alagappan says, "like [Spurs journeyman point guard] Patty Mills next to [Bulls MVP point guard] Derrick Rose. The software tells me that, analytically, they play in a similar fashion." Word trickled to the NBA and now Ayasdi, the firm where Alagappan works, has been recruited by a handful of teams to analyze their rosters. Alagappan has also started breaking down college hoops data, hoping to find even more surprises. Here, for NCAA and NBA basketball fans, are six counterintuitive ways to look at the game.
In the tournament, seeding isn't everything.
There are two rules you should keep in mind when filling out your bracket, according to Alagappan. "The first is that there's no formula to filling out the bracket," he says. "They call it 'March Madness' for a reason." The second is a little less obvious: A team's ranking is less important than its style of play. Just because a team is a number one seed doesn't mean they'll run over a lower seed. To pick a winner, compare teams' playing styles. "I'd identify teams that are prolific rebounders and others that are very good at three pointers," Alagappan says. "I might give the advantage in that matchup to the team that shoots the three, simply because they're able to space the floor on what probably are bigger, slower opponents." Then, weigh each squad's record, the difficulty of their schedule, and the average age of their players, which Alagappan says is often an indicator of success. Finally, take a look at how each team has performed in the 10 games leading up to the tournament. Momentum counts.
Credit: Harry How / Getty Images