NIT Bracketology: TCU Enters the Mix

Our NIT Bracketology is updated to account for yesterday’s games. We have an interesting second overall seed.

TCU’s loss yesterday was a bad loss. It came at home against a UCF team who finished in a tie for eleventh in the Big 12, one who at times this season has been reliant on the prospect of an automatic bid just to make the NIT. It was only a Q2 loss, and only TCU’s first of those (it was the Horned Frogs’ first non-Q1 loss all season), but it was a bad loss, and TCU is now on the bubble. I don’t think anyone, between now and TCU’s Big 12 Tournament’s tipoff on Wednesday, is going to have TCU entirely off the bubble.

Still, I’m not sure I trust our model when it comes to TCU.

We’ve talked a lot about how we haven’t found a good way to incorporate a team’s best wins into our model’s projection of the committee’s assessment of them. TCU’s best wins came at home against Houston, the best team in the country, and on the road against Baylor, most likely a 3-seed a week from today. Those are very good wins, and our model doesn’t look at them. Our model is probably, from that angle, underestimating TCU’s standing in the committee room.

At the same time, though, TCU has the 323^rd-ranked nonconference strength of schedule, and our model draws the line for that at 340. After 339, a bad NCSOS starts to really hurt you in the eyes of our model. Before 340, it has no effect at all. Is this really how it works? I think so. A bad NCSOS is not something the committee has treated with nuance in the past. Is the line at 340, though? That, I don’t know. That was the best number we could set it at during backtesting, but I doubt it’s actually a linear thing. Similarly, our model doesn’t consider how a team finishes its season, and while that’s not a formal variable on the team sheets, there’s evidence (Xavier, 2022) that it can have some effect. If TCU loses to Oklahoma on Wednesday, the Horned Frogs will finish the year on a 1–4 run, with the only win coming over sub-NIT West Virginia.

Ultimately, I’m guessing our model might push TCU back up out of the NIT field as early as tomorrow, with Drake ahead of TCU on its seed list and Drake likelier to lose today than to win. If not tomorrow, then a win over Oklahoma should save TCU if it happens. But then again, our model lines up median scenarios here, and the respective median scenarios involve TCU winning over OU and Drake losing to Indiana State. So…I’m not sure what to expect. At the moment, the model has Drake around 50/50 and TCU around 40/60 to finish above the cut line, a cut line that has plenty of time to move. In any case, our model’s right that TCU’s in trouble. The question is how big that trouble is.

No graphics again today, again in the interests of time. I’m curious how the Ripple Effect (our model learning how NET, KPI, and SOR reacted to yesterday’s games) will shake out tomorrow as well.

Last Four In, First Four Out, Etc.

On the low side:

• Last Four In (from the cut line going upwards): NC State, Indiana, VCU, San Francisco
• First Four Out (from the cut line going downwards): Duquesne, Cornell, Florida State, UMass
• Others in the mix (>10% NIT chance, below NIT field, not in our field or in the First Four Out): Xavier, Samford, Florida State, Rutgers, Georgia, Minnesota, George Mason, Boston College

On the high side:

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

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 82^nd, 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 82^nd seed would be the last NIT at-large.

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 82 and NC State is the 82^nd team on our seed list, 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: If we get JMU, we’ll probably lose App State. If we get Princeton or McNeese, we might lose South Florida or Richmond.

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

• Princeton
• Grand Canyon
• James Madison
• McNeese

Model Talk: NC State

Why didn’t NC State fall out of the field after yesterday’s game? Two things:

First, our model works with median expected results. NC State losing to Pitt was expected. Our model was mostly unshaken by the performance.

Second, our model doesn’t incorporate conference tiebreakers. Drawing Louisville and then most likely Syracuse is a very, very good draw for NC State in the ACC Tournament.

For the Archives