Looking at the Other Metrics Involved in Building a Bracket

With their chance to make history against Kansas and solidify their case for an NCAA Tournament berth, the Cowboys fell short on Thursday. Now what’s left is a long weekend of waiting and hashing out the differences between the Pokes and the other teams “on the bubble.”

Joe Lunardi, ESPN’s Bracketology guru, has especially endeared himself to Oklahoma State fans over the last week, falling on the RPI sword, a noted antiquated and arbitrary stat. Will the other half dozen stats that favor the Cowboys’ case prevail, or will members of the committee hold an unfortunate and somewhat unlucky nonconference schedule against an OSU team that has very clearly looked the part of a Tournament team over the last month?

Earlier this year, ESPN’s Myron Medcalf wrote about some changes that we might expect to the selection process this season — the addition of five additional metrics that would be reviewed by each committee member along side the RPI rankings.

They are the following: ESPN’s BPI rating, ESPN’s strength of record, KPI, Jeff Sagarin‘s rating and Ken Pomeroy’s rating.

Each metric is linked for a more thorough explanation, but as briefly as possible, here are what they represent.

BPI or the “College Basketball Power Index” is ESPN’s stat that serves as a predictor of a team’s success moving forward. The SOR (strength of record) serves to quantify how difficult a team’s W-L record was to achieve.

KPI rankings are based on a mathematical formula that treats all games equally (unlike the weighted RPI) using a scale from -1.0 to +1.0 in tenths from worst loss up to best win.

Sagarin rankings were developed by Jeff Sagarin, a statistician that is way smarter than most of us, and his the recipe is somewhat of a secret. But they have been used by the committee for 35 years.

And KenPom.com is one of the more quoted references on this website. His ratings are based on his own adjusted efficiency, combing offense and defense.

So, using Lunardi’s most recent list of teams on the bubble, let’s look at each of those rankings beside the RPI. In bold are the rankings above 68. None represent an end-all, be-all of whether or not a team should be in or out. But I thought it would be interesting to see each of these team’s other metric side-by-side. It paints a more complete picture of these teams’s numbers, and shows just how difficult the committee members’ current task is.

These numbers are as of Thursday night and will change as the conference tournaments play out.

The last four at-large bids according to Lunardi, these teams should be safe in his mind save for any absolute crazy finishes.

Oklahoma did most of their damage in a 14-2 first half of their season but fell apart like few teams ever, finishing 4-11 and losing twice to you know who. These teams could still get knocked out by an upset or two according to Joey Brackets.

Oklahoma State is Lunardi’s first team out. They would seem like a solid lock for the field if not for that RPI-sized elephant in the room.

If a team like Louisville were to get smoked in the ACC Tournament quarterfinal, it might benefit OSU. But that remains to be seen.

These four are on the outside looking in and would need some series of miraculous events to unfold to get in according to Lunardi.

No one of these rankings are meant to be a strict qualifier (or disqualifier) but represent drawer full of tools to build the whole bracket. With the human element involved, there is always the likelihood of subjective outcomes and therefore the risk of mistakes. Let’s just hope the committee uses all of their tools, not just the ones they’re used to.