Getty Images
You’re not making NewsFeed feel very good about its bracket, Jay.
The NCAA Selection Committee revealed its college basketball seedings Sunday evening, and soon after, one of ESPN’s most prominent analysts publicly berated the committee’s decisions. “I wonder if some people on the committee know whether the ball is round,” Jay Bilas said on ESPN Sunday night.
Bilas was specifically referring to the committee including UAB and Virginia Commonwealth in the tourney while leaving out quality teams like Colorado, Virginia Tech, Boston College and St. Mary’s.
(More on TIME.com: See a brief history of bracketology)
In an interview with NewsFeed the Friday before Selection Sunday, Bilas questioned anyone who thinks they can actually predict the tournament, saying he wouldn’t even fill out a bracket if he wasn’t on ESPN. “We make picks in these things because it’s good for television and people seem to enjoy it,” Bilas said.
But he also took issue with the selection committee for the kinds of statistics that it uses to seed teams.
“I’ve said repeatedly over the last several years that the metrics used by the selection committee in college basketball are not helpful,” Bilas told NewsFeed. “The RPI is an acronym that is best described as ‘really poor indicator.’”
RPI, actually known as Ratings Percentage Index, is based on wins and losses and a team’s strength of schedule. It’s been in use since the early 1980s but has been heavily criticized over the years for not being a precise system to accurately rate teams.
“The committee has said over the last several years that the RPI is a blunt instrument, not a precision tool,” Bilas said. “If you’re going to organize information, why wouldn’t you want it to be as precise as possible?”
So, what’s a better alternative? Well, there are sets of statistics – some old, some new – that appear to be much more accurate in judging the quality of teams. One of those is a system designed by Ken Pomeroy, who has been rating teams since 2003 largely on offensive and defense efficiency. His system has even been compared to baseball’s in-depth sabremetrics stats that revolutionized America’s past-time. And when you compare his past ratings with how well teams performed in the tournament, the accuracy can almost be unsettling. For more on Ken Pom’s stats and others, check out Time.com later in the week for a full story.
(More on TIME.com: Read about BYU’s premarital sex controversy that got one of its top players booted from the team)