NCAA office pools once served a special societal purpose. Before the advent of tweeting, they were the one communal activity in which the entire populace could weigh in on things it knew absolutely nothing about. (Well, that, and elections.)
Picking some upsets and a national champion for an entry fee of a few “M&Ms," I've heard it called, has always been a fun workplace trifle: First, for water cooler talk on Friday and Monday morning, then for chatting across cubicle walls while everyone watches the games live on their computers, and, soon, for experiencing the competition five seconds before the actual players do. Modern culture can't get its information fast enough, so we're one crash course in Heptapod from being able to tell the world that State College already lost the game on a buzzer-beater.
But the diversion of selecting brackets has evolved into an industry, however niche and seasonal. It spawned a science—and I'm unsure just one set of quotation marks does it justice—called "bracketology," in which forecasters spend the college basketball season projecting the 68 teams that will participate in March Madness, and where they'll be seeded. ESPN's Joe Lunardi is the most famous practitioner, though his success rate ranks fortieth among all bracketologists in an average of the last five years.
And yes, there are 40 recognized bracketologists. According to bracketmatrix.com, which does its own seeding of the guys who predict the seeds, there were 175 of them this year. The most accurate one was Kevin Pulsifer, an ESPN statistics editor and former manager for the defending national champion Villanova University Wildcats. That's how it goes at Nova: Your team wins a title one season, and your onetime student aide wins one the next. Said Pulsifer, who's only been at this for two years: "I'd like to thank the academy, my parents, my fiancé for dealing with me staying up [until] 3 AM poring over KenPom data (even if the [tournament selection] committee still refuses to admit that metrics are important), Joe Lunardi for teaching a 'basics of bracketology' class that sparked my interest in the first place, and ESPN for allowing me the opportunity to watch college hoops for a living." Sergeant Mahoney couldn't have said it any better.
Pulsifer mentioned KenPom, named after Ken Pomeroy, which is a statistical analysis website that's supposed to help evaluate teams throughout the season and brackets once "Selection Sunday" has passed. Pomeroy uses tempo-adjusted numbers—some clubs play quicker than others, and thus have more possessions in a given 40-minute game—as well as the relative quality of schools' opponents in a year to grade all Division I programs from one to 351. These rankings favor efficiency on both ends of the floor (points scored and surrendered per possession), making a group like the University of Virginia a recent mainstay in the top 10, despite its 10 losses in 2016-17. A team that lost six fewer games, and plays against similar "power conference" foes and with a more aesthetic style, is run, gun, and fun UCLA. But despite the Bruins' offensive prowess, their defense ranks only eighty-second in the nation. They're 11 spots lower overall than Virginia in the KenPom order. And yet Virginia is seeded fifth in one of the tournament's four 16-team regions, whereas UCLA is seeded third. This gets confusing quickly.
Get the Exclusive
04.28.2024