Here's My Annual Push for NCAA Selection Committee to Consider "Wins Above Bubble" as a Tool

The NCAA selection committee uses the RPI as a primary tool in selecting teams. I’ve railed against the RPI in the past, but I don’t want to make this a breakdown of the measure. I attended the mock selection exercise last year and it is clear that it is not the RPI ranking, but rather the results against teams ranked at a certain level, that matters. How many Top 25 RPI wins did you get? What about Top 50?

On the summary sheet, these things are listed, and the individual would have to break down whether those Top 50 wins were home against the 50th team, or on the road at #10. They would have to look to see if wins against teams 51 and 53 are missing but basically similar to another team.

Ken Pomeroy wrote about this problem a few weeks ago, when he said:

"“It turns out those adjustments are important. Beating the 90th-ranked team on the road is about as difficult as beating the 50th-best team on a neutral floor, which is roughly as difficult as beating the 20th-best team on one’s home floor. (The exact relationship can vary by season.) So it’s poor form to ignore this when some teams in college never go on the road in non-conference and others rarely get to play a non-conference game at home.”"

I wrote about Wins Above Bubble last year, and I’ve written about a similar concept under a different name in the past. The philosophy is similar to what Pomeroy is laying out. If we are going to evaluate a resumé, let’s be fair and calculate how many wins a bubble team should have, once we account for venue and specific opponent.

Seth Burn is running the Wins Above Bubble numbers this year, so I will frequently mention them. The basic philosophy is this–figure out what a typical bubble team would be expected to do against each schedule, then see how many wins the team in question has.

Here are the teams that have the highest “wins above bubble” heading into tonight:

Oklahoma, +5.8
Villanova, +4.7
North Carolina, +4.5
Iowa, +4.2
Xavier, +4.1
Texas A&M, +4.1
Kansas, +3.8
SMU, +3.7
Providence, +3.4
Michigan State, +3.3
West Virginia, +3.2
Maryland, +3.2
Virginia, +3.2
Iowa State, +3.1
Dayton, +2.8
Oregon, +2.5

In comparison, here are the top 16 in the RPI. Now remember, just like RPI, WAB (Wins Above Bubble) isn’t taking margin of victory into account. Which one looks better?

Villanova
Oklahoma
Kansas
North Carolina
Xavier
SMU
Oregon
Texas A&M
Maryland
Dayton
Virginia
Iowa
Iowa State
Texas
Baylor
Kentucky

Who is the best team in the country? WAB has Oklahoma comfortably in front. Is Oregon more likely a 4-seed or 2-seed? Is Iowa worthy of a top 2 seed based on their resumé? They’ve beaten both Michigan State and Purdue twice, and have no bad losses (Notre Dame on a neutral court is the worst of the three). Is Texas a top 15 team? RPI has them there, playing against the toughest schedule in the country. Wins Above Bubble says that a bubble team would be somewhere between 11-8 and 12-7 against that difficult schedule. Texas is 12-7.

Wins Above Bubble would be a useful tool for comparison, as a baseline for whether a team has enough wins to be in consideration. It helps small schools that get limited opportunities but do well, have a decent RPI as well, but are told they don’t have the Top 50 win profile of tourney teams. Arkansas-Little Rock is 17-2, for example, with wins at San Diego State and Tulsa, and a loss at Texas Tech. Notice that all those games are on the road. A bubble team would have closer to a 15-4 record (and just go look at the “bad” losses bubble teams that will be boosted over UALR have). It would be a shame if a team like that missed out at 27-4, with a conference tourney loss. They could, by the way, even without a big upset. Conference rival UT-Arlington went to overtime on the road at the #14 team in the country (by RPI) and is ranked in the top 60 as well.

So I’m gonna track wins above bubble as the season continues. I’m under no delusion that the committee actually uses it, but it’s actually simpler and would be a more consistent application of looking at resumés.