Well, our model likes North Texas. I don’t know what to tell you guys. I hope it’s right. I don’t know that it’ll be right. We’ll get into that below.
The schedule for today goes as follows:
- 12:00 PM EDT: Games start.
- 5:35-ish PM EDT: Games end.
- 6:00 PM EDT: NCAAT Selection Show (CBS).
- 9:00 PM EDT: NIT Selection Show (ESPNU).
We’ll update our bracketologies (NCAAT, NIT) once the games are over, and we’ll update the NIT one again after the NCAAT Selection Show, when we know who the NIT 1-seeds are and who all is in the NCAAT. So that’s our schedule. After everything’s announced, we’ll have our NIT Bracket Challenge up for the entering, and we’ll have content on Twitter and we’ll be answering questions in the comments throughout the day. Now…the bubble picture:
Locks (automatic bids marked with an asterisk)
Conferences with locks in at-large territory:
- A-10: Davidson
- AAC: Houston*, Memphis
- ACC: Duke, UNC, Virginia Tech*
- Big 12: Kansas*, Baylor, Texas Tech, Texas, Iowa State, TCU
- Big East: Villanova*, Providence, UConn, Seton Hall, Creighton, Marquette
- Big Ten: Purdue, Wisconsin, Illinois, Iowa, Michigan State, Ohio State
- Missouri Valley: Loyola*
- Mountain West: Colorado State, Boise State*, San Diego State
- Ohio Valley: Murray State*
- Pac-12: Arizona*, UCLA, USC
- SEC: Kentucky, Tennessee, Auburn, Arkansas, LSU, Alabama
- WCC: Gonzaga*, Saint Mary’s, San Francisco
Conferences with automatic bids:
- America East: Vermont*
- ASUN: Jacksonville State*
- Big Sky: Montana State*
- Big South: Longwood*
- Big West: Cal State-Fullerton*
- CAA: Delaware*
- Conference USA: UAB*
- Horizon: Wright State*
- Ivy: TBD (Princeton or Yale)
- MAAC: Saint Peter’s*
- MAC: Akron*
- MEAC: Norfolk State*
- NEC: Bryant*
- Patriot: Colgate*
- Sun Belt: Georgia State*
- SoCon: Chattanoga*
- Southland: Texas A&M-Corpus Christi*
- Summit: South Dakota State*
- SWAC: Texas Southern*
- WAC: New Mexico State*
That’s 61 spots, leaving somewhere between five and seven open for the taking, with Richmond capable of stealing a bid and Texas A&M capable of locking themselves in with a win today.
The Bubble
For those last five to seven, there are probably twelve teams in contention. It’s possible the committee will throw us a curveball, but it’s really just twelve teams, to our eyes. Here are those twelve, in order of how our model ranks them, with words about each:
Michigan
The risk for Michigan is that their overall record (17-14) is too close to .500 for the committee to accept it. Beyond that, their résumé’s solid. Not outstanding—they lost fourteen games, after all—but solid.
Wyoming
Wyoming is similarly probably safe. Their KenPom ranking is only 58th, but that shouldn’t be low enough to make much of a difference, and their 4-5 record in Q1 games should be enough to offset their two Q3 losses.
Texas A&M
Texas A&M, as well, is likely safe after yesterday’s win. They’ve raised their KenPom ranking from 80th to 42nd in a matter of weeks, and while we don’t have data on where their NET ranking was at that time KenPom said 80, one would assume it was similar. Their SOR ranking—33rd—is great for a bubble team. They have few bad losses. They won four Q1 games, and while they’ll have lost ten, should they lose to Tennessee, a 4-10 record in that quadrant is still above the threshold at which our model expects punishment to be incurred.
SMU
SMU could have locked themselves in by beating Memphis. Instead, they’re squarely on the bubble, which they’ve occupied for months. Little jumps out about their résumé, which can be bad but can also be good. If there’s little objectionable, they can be a team everyone agrees on.
Xavier
Xavier has a very similar résumé to SMU in terms of how little it stands out. Xavier, however, finished the season on a 2-8 run. Does recency affect NCAA Tournament admittance? It’s unclear. Our model thinks the Musketeers’ five Q1 wins will be enough to pull them across the line, but our model can be wrong, and it’s deathly close.
North Texas
Here’s the weird one. North Texas is 47th in NET, 46th in KPI, 55th in SOR, and 50th in KenPom. They went 1-1 in Q1 games, 5-3 in Q2 games, and 5-2 in Q3 games. They have a 55th-ranked nonconference strength of schedule. The Mean Green are sixteen games over .500 overall, past the point where our data indicates it starts to matter (this helped UNC-Greensboro contend for a bid a few years back).
The lack of Q1 games of any kind, the absence of any narrative boost, and the weakness of their closing argument (Louisiana Tech was only a Q2 loss, but losing to a team unlikely to make the NIT isn’t the best way to finish off your tournament quest) might do them in. But they’re probably closer than they’ve been made out to be, at least if we’re going off of precedent. They do, for what it’s worth, have a winning record in Q1 and Q2 games combined, which we’ve seen matter in the past.
Miami
Our model is low on Miami, who may be safely in the field. The things it doesn’t like are the 62nd ranking in NET and…that’s about it. They didn’t quite get to five Q1 wins, which our model views as an important threshold, but they do have a winning record in Q1 and Q2 games. Our model would have them lose their spot if Richmond were to win, but we understand if you think our model would be wrong. We think our model would probably be wrong.
VCU
Another team our model’s high on is VCU, but it does have the Rams out of the tournament. There’s little objectionable about VCU’s résumé, and they’re 29th in KPI, so if they get in, pin it on that.
Oklahoma
Oklahoma briefly surged into our projected field, but the corresponding surges of Virginia Tech and Texas A&M have knocked the Sooners out. Even having beaten Baylor, they won only 25% of their Q1 games on the year, which is right where our model starts pulling teams down (and hard). They have ten combined Q1 and Q2 wins, which lifts them back up, but our model doesn’t expect that to be enough.
Indiana
Here’s our other weird one. This and North Texas.
The deal with Indiana is that they’re only 59th in KPI, they didn’t get to five Q1 wins or ten Q1/Q2 wins, and if they have a narrative bump it’s one our model can’t see, because our model only looks at last week’s poll. If Indiana makes it, the likely fix for our model is to add some variables next year like “Q1 wins during championship week” or “top 25 wins during championship week” or “losses to non-tournament teams during championship week,” the last of which would pull North Texas down. Q1A wins might also be a worthwhile metric to look into adding.
Anyway, we understand that a lot of people think Indiana is safely in. We’re inclined to agree, but our model can highlight some things even in instances where it’s ultimately wrong. In this case, the takeaway is that Indiana’s résumé is not as strong as its narrative. Both might be strong enough, though!
Notre Dame
This is another one where our model disagrees with consensus, but in this case, we don’t understand what our model’s missing. Notre Dame went 2-8 in Q1 games. Their KPI ranking is only 64th, and their NET is just 53rd. Our model does give them a boost because they got at least one vote in the most recent AP Poll, reflecting the impact of either consensus, groupthink, or subliminal narrative influence, but even that’s not enough for the Irish in our formula. They’re behind UAB on our list. And South Dakota State. And Chattanooga. Can they make it? Definitely. But they could also not.
Rutgers
One thing lost in the Rutgers conversation is the absence of Geo Baker for a game and a half during that awful nonconference swoon. Another thing lost, though, is their 74th ranking in KenPom. The Scarlet Knights have six Q1 wins, and many of those were impressive, but they’re an inconsistent team that would be an underdog against Fresno ****ing State. Also, the wins weren’t that impressive. The Big Ten is strong as a whole, but it’s having a down year, and that’s most visible at the top. Wisconsin would be an underdog against Oklahoma.
The consensus seems to be coming around on the reality that Rutgers is not an NCAA Tournament-worthy team. But we don’t know with certainty. We rarely know with certainty with this stuff.
How We Expect Our Model to Perform
I’d be surprised if our model misses more than two NCAAT at-large teams. I’d also be surprised if it gets all of them right. It wouldn’t be stunning to us for any of North Texas, VCU, Oklahoma, Notre Dame, and Rutgers to make it, but it also wouldn’t be stunning if Texas A&M, SMU, Xavier, Miami, or Indiana were to miss the field. We’re fairly confident in Michigan and Wyoming gaining admittance, but there’s a large human element to this which makes it unpredictable, and when you’re comparing teams at the 85th percentile of a list (rather than the 97th percentile, as we do in college football), you’re splitting a lot of hairs. Lists are just tighter at that point in the bell curve.
In this sense, I don’t envy the committee. It’s a tough job, one that requires the arbitrary crushing of hopes and dreams. All we can hope for, really, is a consistent rubric. That’s what our model tries to reflect. We’ll see.
