NIT Bracketology: Final Thoughts On The Bubble and the Black Box

The NIT bracket was released (after it was corrected) last night around 10 pm ET. The biggest surprise (to me) was the inclusion of UCF.

The biggest takeaway for me from the teams that were selected (and the teams I assume were close) was that the NIT Selection Committee relied heavily on the NCAA's own NET ranking. As you can see below the only outlier in terms of the NET for which teams were selected around the bubble was Santa Clara. The Broncos were clearly the third best team in the West Coast Conference this season. I think they ultimately got in because of that and because of their strong resume metrics. Santa Clara was 64th in ESPN's Strength of Record. It's also possible based on the seeding that either Villanova, Seton Hall, or Santa Clara was the final team into the field after North Carolina declined a bid. (And I would guess it was Santa Clara.)

Team	Selected	Seed	NET	KenPom	Strength of Record	Q1/Q2 Wins
UAB	Yes	4	57	62	75	3
Florida	Yes	4	62	63	77	3
Colorado	Yes	3	69	59	94	7
Cincinnati	Yes	4	70	54	73	3
UCF	Yes	5	71	65	107	5
Washington St.	Yes	4	73	57	92	7
Villanova	Yes	6	75	60	78	5
Virginia Tech	Yes	5	76	81	86	6
Seton Hall	Yes	6	77	70	72	6
Dayton	No		78	78	106	2
Utah	No		79	72	100	5
Marshall	No		83	84	96	2
Santa Clara	Yes	6	84	82	64	4
Wake Forest	No		90	88	83	5
Nebraska	No		92	95	65	7
San Jose St.	No		95	93	81	6

Considering the NET was originally introduced as a sorting tool that is supposed to place teams into quadrants, it's odd that the selection committee leaned so hard into it as a selection tool. (Though it is worth noting that Wisconsin and Vanderbilt, which were 80th and 81st in NET received high seeds. On the other hand, both those teams had a plethora of strong wins and were considered potential NCAA Tournament bubble teams for a long time.)

In addition, it could've been that the committee leaned heavily on KenPom and not NET and they're just so closely aligned that it's hard to see the difference, but that would actually make Santa Clara and Virginia Tech outliers.

I've tended to use 100 in NET as my cutoff heuristic for delineating potential at-large teams for the NIT. That still seems like a reasonable line to draw, but it does seem like a team needs something rather extraordinary to make it even with a NET above 80 or so.

Some of these questions and concerns could've also been alleviated if the NIT Selection Show on ESPNU had given literally any insight into how the bracket was selected and seeded. I would've loved to learn what other teams were considered and which team entered the bracket after the Tar Heels declined. The NIT could also publish their s-curve like the NCAA Tournament does, but that has never happened either.

So we are left with this black box. We'll try to figure it out again next season.