Something we do a decent amount around here is approximate KenPom win probabilities. We have a normalized curve that converts his adjusted efficiencies to win probability, and it’s a little different from his actual curves, but it gets the job done, so we use it quite a bit for things like figuring out which conferences are likeliest to send automatic bids to the NIT (we live an exciting life). Here’s what it says about Final Four probabilities out of each of the four regionals this weekend, plus championship probabilities out of the NIT Final Four this coming week (not all will round to 100%, due to rounding error):
- East Region: Tennessee 42%, Kansas State 20%, Michigan State 19%, Florida Atlantic 18%
- West Region: UConn 36%, UCLA 34%, Gonzaga 20%, Arkansas 10%
- South Region: Alabama 50%, Creighton 31%, San Diego State 17%, Princeton 3%
- Midwest Region: Houston 51%, Texas 28%, Xavier 13%, Miami 7%
- NIT Final Four: North Texas 33%, UAB 28%, Wisconsin 20%, Utah Valley 19%
The question, then, is what to take away from this.
Probability isn’t linear. The gap between Princeton at 3% and Arkansas at 10% is a lot bigger than the gap between Arkansas at 10% and San Diego State at 17%. Arkansas is more than three times as likely as Princeton to make this Final Four. San Diego State isn’t even twice as likely as Arkansas. A better way to put this is that San Diego State has roughly a one in six chance, Arkansas has roughly a one in ten chance, and Princeton has roughly a one in 33 chance.
Princeton, then, is in another category from the rest of the field, and Miami’s in a bit of its own category as well, and there are big gaps as you work upwards until suddenly there aren’t anymore. Let’s do some categorizing:
The Coin Flips: Houston, Alabama
Each checking in at about 50%, Houston and Alabama are good teams with good draws. By KenPom, Houston’s region includes the 5th, 10th, and 15th-best team left in this. Alabama’s includes the 8th, 9th, and 16th. Alabama’s probably going to have two challenging but manageable matchups. Houston should theoretically have little trouble with Miami, and then they’ve got a decent chance of drawing Xavier instead of Texas. Plenty can go wrong—it’s hard to win two games in a row, and these probabilities are just coin flips—but at this stage, this is close to the best a team can ask.
The Factors: Tennessee, UConn, UCLA, Creighton, Texas
All of these teams fall roughly between two in five likely and two in seven likely to make the Final Four. For Creighton, it’s almost all draw: Creighton gets to play Princeton tomorrow, which shouldn’t be easy but should be the least difficult game for anyone from here out. The other four are simply good. They’re all between 3rd and 6th in KenPom, both overall and among those remaining. We’ll have more on the teams in a minute, and their risks and consistencies, but in terms of sheer quality, these four—beyond Creighton—are near the top.
In the Mix: Gonzaga, Kansas State, Michigan State, Florida Atlantic, San Diego State
These are all between one in six and one in five likely to advance past Sunday. They’re a wide range—Gonzaga is right after Tennessee in terms of quality, San Diego State’s stuck around Creighton and Xavier, the other three are in a pack down past Arkansas—but the East Region is so open, and Gonzaga’s path is so difficult, that they end up in the same spot.
Longshots: Arkansas, Miami
Maybe these shouldn’t be lumped together—Arkansas’s one in ten likely, Miami’s at one in 14—but they’re comparable enough for our purposes. Arkansas seems like they belong, narrative-wise, at this stage. But. They have the worst path of any team left in this, and they merely beat a shorthanded Kansas who was only 9th in KenPom to begin with. Why are we referencing KenPom so much? This is a KenPom section. More thoughts that require thought in a minute. Miami, meanwhile, just isn’t that good. Besides Princeton, they’re the worst team left in this, and they might be an underdog right now against North Texas, the NIT favorite. They would be a KenPom underdog in that matchup. Their path is tough, too, drawing Houston and theoretically Texas.
Princeton: Princeton
Fun to have a 15-seed here, but again: This is roulette wheel probability.
**
As for the games themselves, some thoughts on each of today’s:
I’m in knots over the Michigan State/Kansas State game. Not emotionally, but mentally. Kansas State seemed so overvalued for so long. Michigan State now seems potentially overvalued themselves. With each, there’s a question of whether they’re happy to be here, but each head coach is deservedly highly thought-of. They’re both rather well-rounded, they’re both led by their guards…in the end, the matchup seems to boil down to whether Kansas State’s firepower can outgun Michigan State’s clean play. Kansas State is the variable here. We have a pretty good idea what we’ll get from MSU. With shooting a question and with Kansas State so adept at moving the basketball and not relying on the three, I do think Kansas State should probably be favored, but damn, is this one close.
In the later New York game, it seems like people are sleeping on FAU after spending a little while sleeping on Tennessee. FAU is so deep, so clean, so well-rounded, and so good. They’re also still a little bit of an unknown. They were 40th in KenPom a month ago this morning, and they’re now 21st. That’s the kind of swing that can happen when you’re either playing really well or running into better than usual competition. Tennessee, meanwhile, has certainly dropped in the absence of Zakai Zeigler, but the question is by how much. He was not their best player, but he was a big part of what they did. Still, FAU’s the more one-dimensional team here. They have depth, but only Vladislav Goldin really grades out all that well. Compared to the betting odds, I’m high on the Owls, but Tennessee is a really good team, and deservedly a big favorite here. Duke would have been a better draw for FAU.
Out in the West, Arkansas and UConn should be electric, to repeat what everyone else is saying. Each team has a high chaos factor. It doesn’t seem like UConn gets enough credit for Donovan Clingan, their stand-in for Adama Sanogo when the big man needs a breather. The 7’2” freshman from Bristol plays huge minutes every game and is usually highly effective. EvanMiya has him as the fourth-most valuable player in the country when he’s on the floor. Small sample, yes, but that says a lot. It’s things like this—UConn has five players above Arkansas’s best on EvanMiya—that show the real difference here. UConn is flawed. They turn the ball over too much, they don’t deal all that well with adversity. But they haven’t lost by more than a possession since January 15th, and their individual players are all so good. Which is wild to say about the team playing Arkansas rather than saying it about the Hogs themselves.
In the late game there, injuries are the key story. David Singleton and Adem Bona are role players, but they play big roles, and at less than 100%, their presence and/or absence is going to be a big deal. Gonzaga’s strength is in its forwards. Drew Timme is a wizard offensively. Anton Watson is a more impactful overall player. UCLA’s strength is more, right now especially, in its guards. Tyger Campbell is trustworthy, and that tends to matter in single-elimination basketball. EvanMiya doesn’t really back up the theory that UCLA is deeper than it’s showed—Kenneth Nwuba doesn’t grade out all that well, even if his scoring numbers are positive—but of everyone in the tournament, Jaime Jaquez is a strong candidate for the guy you’d most want in the last minute of a tie game, and that should matter.
If making picks for tonight, I’ll stick with the FAU value, Kansas State, UConn, and UCLA. Those aren’t bets—though we do have futures outstanding on both FAU and Kansas State to get through the weekend—but they’re the thoughts. I think Tennessee will win, but I’d pick FAU if I was on one of those picking shows. You have to take your shots. I think their consistency will go a long way against an inconsistent team like Tennessee, in a twist on the usual good/consistent vs. medium/inconsistent dichotomy.
1 vs. 0
One thought from the Bulls game last night: I think I’d rather have 0 points than 1 in a situation like that one. When it was 17–0, it was ugly but it was early. When it was 23–1, it was pitiful. Also, it was evident that someone had missed a free throw.
