NIT Bracketology: Everybody Won

This is our NIT Bracketology rundown for the morning of Thursday, March 14^th. If you’re looking for the most current rundown, visit our homepage.

Entering yesterday’s games, there were 13 teams in the lower bubble section of our NIT Bracketology picture. Ten of them played. Here’s how they did:

• NC State: Won
• VCU: Won
• Duquesne: Won
• Florida State: Won
• Boston College: Won
• Xavier: Won
• Maryland: Won
• Georgia: Won
• Penn State: Won
• George Mason: Lost

So, we can say goodbye to George Mason, but that’s about it, and there were teams below these 13 who did some winning as well. Here’s our model’s latest projection of the eventual NIT field:

Mississippi State moving in is odd, but less so when you consider that New Mexico’s median scenario involves beating Boise State. The Lobos are a kenpom favorite. Our model uses kenpom as its starting point for game simulations. In other cut line news, we’re back to the scenario where we have 33 teams for 32 spots. Although: While San Francisco is 83^rd on our list of median NIT selection ranks, their median rank is only 85^th. 85^th is the 83^rd-best median NIT selection rank. Meaning: While our model doesn’t know who will pass the Dons, it expects someone to do that. So it’s not all that weird in practice for us to have San Francisco out of our field.

We do trust others more than our model around the NCAAT/NIT cut line, something we reflect over on our NCAAT Seed List. We think Mississippi State’s probably safer from the NIT than Texas A&M in reality, and we don’t dislike St. John’s chances of missing the NIT, though our model does know the Johnnies are favored today as well. More on A&M over at that seed list. Let’s talk NIT bubble.

Last Four In, First Four Out, Etc.

On the low side:

• Last Four In (from the cut line going upwards): Duquesne, Boston College, Florida State, Washington
• First Four Out (from the cut line going downwards): San Francisco, UMass, LSU, Cornell
• Others in the mix (>10% NIT chance, below NIT field, not in our field or in the First Four Out): Maryland, Xavier, Penn State, Georgia, Minnesota

On the high side:

• Last Four In (from the cut line going downwards): Mississippi State, St. John’s, Wake Forest, Villanova
• First Four Out (from the cut line going upwards): New Mexico, Seton Hall, Indiana State, Virginia
• Others in the mix (>10% NIT chance, above NIT field, not in our field or in the First Four Out): Oklahoma, Colorado, Colorado State, TCU, Texas A&M*

As always, we account for bid thieves by looking at the median cut lines in our model’s simulations. Today, those remain 45^th and 83^rd, meaning if there was a master seed list covering both tournaments, the 45^th overall seed would be the last NCAAT at-large and the 83^rd seed would be the last NIT at-large.

*Texas A&M is only 7.3% NIT-likely in our model’s eyes, but as we explain over at the Seed List, we’re skeptical of our model on this.

Conference Tournament Favorites (Reverse Bid Thieves)

Given where those cut lines stand and how many teams sit between them, we don’t have any reverse bid thieves today. All conference tournament favorites who sit in NIT at-large territory are, in today’s bracketology, not in the NIT. More likely than not, this isn’t what happens, but our cut line is 83 and Duquesne—our last team in—has a median NIT selection rank of 83 in our simulations, so there’s a good chance that whatever reverse bid thieves do happen are canceled out by current projected NIT at-larges playing their way into the NCAAT as automatic bids. For example: Princeton could lose in the Ivy League Tournament, but Richmond could win the A-10 Tournament.

The two remaining conference tournament favorites in NIT at-large bid territory if they lose:

• Princeton
• Grand Canyon

Model Talk: Appearances (Might) Matter

One of the most frustrating things about the NCAA Tournament’s selection process is how early it begins. The argument supporting this practice is that it gives the committee time to think things through in detail. The argument against it is that it gives the committee time to overthink, landing on a process in which different things matter more for different teams. The worst thing, though, about how early the committee gets to work is that it leads to situations like Texas A&M’s in 2022, in which the Aggies’ Saturday and possibly Friday games were clearly not considered. Every game matters more and more, until the results suddenly stop mattering at all.

I understand the value of entertainment. I too love watching a dozen conference championships on Saturday, five more on Sunday, and then perching on the edge of my seat for a Selection Show (two shows if you are among our little NIT cult). I do not at all support getting rid of this. There should be no gap between Champ Week and Selection Sunday. But couldn’t the field be set in a day or two? Bracketing according to principles is automatable. Seeding hews close to a set of formulas. Most of the field is obviously in. There’s little to argue over by the time this weekend gets here. Let the data materialize. Use something like WAB or SOR to build an initial seed list. Tinker with it until a majority approves the final product. Do the whole thing faster. Make more games matter more.

The College Football Playoff conference room, with games on the big TVs, has become a meme. But when they cut to the selection committee, it doesn’t seem like a whole lot of basketball is being watched by the time the bubble is being decided. As for what is watched…

How does the committee view Cincinnati’s win over Kansas, given Kevin McCullar and Hunter Dickinson didn’t play? How does the committee view Kansas after its loss to Cincinnati, given the Jayhawks have media waterboys proclaiming that at least McCullar will be back next week? Should injuries matter? How much? Should those answers be the same every year? Are they the same every year?

To answer all these questions: It’s discounted; they’re probably uneasy but taking the company line; I don’t know; only a little if anything; yes please make them the same every year; no of course they are never exactly the same. In the end, Cincinnati’s win does little for it, because the committee is well aware of the injuries to KU. Why are they well aware? Because it’s a major story in a way it would not be a major story in January. It’s timing. Sure, Cincinnati might have caught a break playing Kansas without two strong players in a situation in which the Jayhawks were clearly looking ahead to next week. But given how much it’s likely to be discounted, you could argue the bad luck of timing outweighs any good luck the result might be. Similarly, even if Cincinnati upsets Baylor tonight, what they do tomorrow might not make a dent in the committee’s final decision, despite a late-season single-elimination game against a quality opponent being exactly the kind of thing the committee always hints it overvalues, seeding teams based on how they’re expected to perform in the NCAA Tournament.

Cincinnati dug their grave. They played themselves into this situation. This is how it works, and this is what they did, and because the “rules” would be applied this way to anybody, it’s perfectly fair. But for as silly as the notion is that March games should matter more than November games in the committee room, an even sillier one is that March games should matter more until Thursday, at which point they should start tapering off, to a degree where the Big Ten and SEC Tournament Championships are often exhibitions for the teams involved, especially if those teams take the view that the regular season title is the real crown.

Anyway, a refreshing thing about the NIT committee is that they work quickly. And thankfully for us, it makes it easier to statistically model their work, so long as they’re consistent. Our model is much more accurate on the lower bubble than the upper bubble. On the upper bubble, our model might be in love with Cincinnati by morning. I was worried it’d be fawning over them already today.

The Graveyard

As always, we want to talk about who’s losing what because of the change in the automatic bid rules. We think it’s important to track the real impact of the changes, or at least our closest estimate of that impact. So:

• Losing an automatic bid (teams who would have made it in last year’s format): High Point, Eastern Washington, Little Rock, Eastern Kentucky, Central Connecticut State
• Gaining a bid (teams we think are receiving those bids instead): Duquesne, Boston College, Florida State, Washington, VCU
• Losing a home game (teams who would have been seeded and had the option to host last year): Richmond, Butler
• Gaining a home game (hosting the first round in our projection only because of their automatic bid): UCF, Mississippi

Not all non-power conference teams are created equal. We believe strongly in the high-major/mid-major/low-major dichotomy (as opposed to calling teams only power conference schools and mid-majors). In total, then, what we have is this:

• Five NIT spots shifting away from low-majors. Three of those go to mid-majors. Two go to high-majors.
• One NIT home game shifting away from a mid-major. It goes to a high-major.

For the Archives