The bracket below is our best current prediction of the final NCAA Tournament bracket. It is predictive of where things will end up. It is not a reflection of where things currently stand. (Update: With selection happening today, it is now a reflection of where things stand.)
The bracket below is also objective. We are not making choices based on our opinions of various teams. Our college basketball model simulates the remainder of the season 10,000 times, including the selection process and postseason tournaments. It gives us the likeliest automatic bid in each conference as well as each team’s median position on the committee’s seed list. With that information, we build the bracket based on our best understanding of the bracketing principles. Here is how the model works, in full detail.
If you notice any irregularities or have questions of any sort, please contact me through the information available on our about page or on twitter: @joestunardi.
- The winner of the first region listed plays the winner of the second in the Final Four.
- Automatic bids are marked with a single asterisk (*).
- If you want NIT Bracketology, we have it.
- If you want our model’s probabilities concerning each team’s tournament chances, we have those as well.
- For an alternate, more subjective (and therefore conventional) look at the field, check out our Seed List.
Last Updated: Sunday 3/17 – FINAL
SOUTH
Second Weekend: Dallas; First Weekend: Memphis (1/8), Salt Lake City (2/7), Indianapolis (3/6), Spokane (4/5)
| 1. Houston |
| 16. Grambling State*/Montana State* |
| 8. Clemson |
| 9. Mississippi State |
| 4. Alabama |
| 13. Samford* |
| 5. Utah State |
| 12. McNeese* |
| 3. Creighton |
| 14. Morehead State* |
| 6. Wisconsin |
| 11. Grand Canyon* |
| 7. Nevada |
| 10. Texas A&M |
| 2. Arizona |
| 15. Long Beach State* |
WEST
Second Weekend: Los Angeles; First Weekend: Charlotte (1/8), Memphis (2/7), Pittsburgh (3/6), Spokane (4/5)
| 1. North Carolina |
| 16. Stetson* |
| 8. Colorado |
| 9. Michigan State |
| 4. Baylor |
| 13. College of Charleston* |
| 5. Gonzaga |
| 12. UAB* |
| 3. Illinois* |
| 14. Akron* |
| 6. Saint Mary’s* |
| 11. NC State* |
| 7. San Diego State |
| 10. Northwestern |
| 2. Tennessee |
| 15. Western Kentucky* |
EAST
Second Weekend: Boston; First Weekend: Brooklyn (1/8), Charlotte (2/7), Pittsburgh (3/6), Salt Lake City (4/5)
| 1. UConn* |
| 16. Saint Peter’s* |
| 8. Nebraska |
| 9. TCU |
| 4. Kansas |
| 13. Vermont* |
| 5. BYU |
| 12. James Madison* |
| 3. Duke |
| 14. Oakland* |
| 6. Dayton |
| 11. Oregon* |
| 7. Texas Tech |
| 10. Florida Atlantic/Boise State |
| 2. Auburn* |
| 15. Longwood* |
MIDWEST
Second Weekend: Boston; First Weekend: Indianapolis (1/8), Omaha (2/7), Omaha (3/6), Brooklyn (4/5)
| 1. Purdue |
| 16. Howard*/Wagner* |
| 8. Texas |
| 9. New Mexico* |
| 4. Kentucky |
| 13. Yale* |
| 5. South Carolina |
| 12. Duquesne* |
| 3. Marquette |
| 14. Colgate* |
| 6. Florida |
| 11. Drake* |
| 7. Washington State |
| 10. Virginia/Colorado State |
| 2. Iowa State* |
| 15. South Dakota State* |
Hi there, any thoughts about Ohio State’s chances at an at-large spot? I know their conference record and having a Quad 3 loss (with only 3 Quad 1 wins) are hurting them.
Our model’s pretty high on them, actually. Here’s its latest probabilities: https://thebarkingcrow.com/college-basketball-probabilities/
I’m surprised more brackets don’t have Ohio State in their Next Four Out right now (i.e., 5–8 spots behind the cut line). They have a stronger team sheet in almost every relevant category than Iowa and Memphis, and those teams are at least getting courtesy mentions. The one thing I can find that clearly hurts is combined Q1/Q2 record, and their SOR isn’t great either. But they definitely have a chance if they can make some noise the first couple days in Minneapolis.
As a Colorado fan, they seem to be ranked too high here. How much do Kenpom/Net factor into this model vs others?
In its light version, our model does heavily lean on kenpom as a stand-in for future results, so you’re right: That’s a lot of what this is. That said, even if we were running a fuller version of the model (simulating individual games), we’d be higher on Colorado than the consensus. The consensus is more reflective of where things stand, while we’re always trying to point towards where things will land.
A four bid Mountain West! Just how it was intended to be!
The Mountain West is going to be a league that could possibly steal some bids from power conference, especially if New Mexico and Boise State play well all season. Nevada could be the fourth program that could advance into March Madness, especially if they win the Mountain West tournament.
So now that my Tarheels have a win over UVA and have avoided any bad losses, what gets them in? IF they beat Florida State but lose to Duke do they still have a path other than winning the ACC tournament, or do they have to win these last two AND get a couple of wins in the ACC tournament. I hate being on the bubble…lol.
They’re definitely close. Depends so much what teams do around them, but seems like they should really beat Duke. Don’t know the ACC Tournament seeding yet, but if they can beat Duke and avoid any more bad losses, I’d guess they narrowly get in. That’s just a guess, though. Working on better projections every day.
Any chance the Tarheels sneak in? I know it would take nothing short of a miracle, but what, other than winning the ACC tournament, would that miracle be?
I really don’t think they’re terribly far off, to be honest. May just need to split with UVA/Duke and then win two in Greensboro (or sweep UVA/Duke and win their ACC Tournament opener against VT or whoever). It helps them a lot that teams like Wisconsin and New Mexico are underwhelming and teams like Utah State and USC have to dodge a lot of potentially bad losses. Keep watching Bracket Matrix! Plenty of season left.
let me be the first to comment on 2023. Why so much love for West Virginia over OU?
If OU sweeps the series this weekend does that change the narative?
Boomer Sooner
That’s a fair comparison. We have West Virginia so much higher because our model’s looking at what each team’s expected to do from here, using KenPom as the best predictor of that. West Virginia has a solid chance of finishing the season without a Q2/Q3/Q4 loss, which is a big deal. Oklahoma’s current median expectation is to finish the season 16-16, which is traditionally no-go territory for the NCAAT committee. The nice thing about where they’re at, though, is that if they steal even one or two they shouldn’t win, they’ll rise really fast, because that expected W/L record is the primary thing holding them back. So yes, if they finish off the sweep it’ll change in our eyes.
thanks for the insight. Bedlam tonite home with OSU and FREE tickets. Not impressed giving free tix to OSU fans. LOL
Alright, Joe. I have to ask. Is there any path for South Carolina to make the NCAAT at large? Would beating Auburn (unlikely, I know) and winning two SECT games at least put us on the bubble? I think our resume is competitive, but obviously NET is terrible. Looking forward to hearing your thoughts.
There’s a path, but it’s very small. The nice thing about NET is that it’s possible to make a huge swing with one game, which isn’t true of the others. If they can get their NET into the 60’s and grab a Q1 win in the process, they’ll probably be sniffing the bubble. My best guess for what would get them there would be either A) beating Auburn and making the SEC Semifinals or B) losing to Auburn but keeping it close, then making the SEC Championship. Margin matters a lot for NET, so in all their games, winning big or losing close is important.
Good to hear. I guess I will get my hopes up again only to get disappointed this weekend (haha). Thanks!
No Furman anywhere? I notice Chattanooga seeded in both NCAA and NIT?
We had Furman in our NIT projection a few weeks ago, but the UNCG loss was bad, and they’ve done worse than expected following that, too.
With NIT automatic bids, right now our model looks at the most likely number of auto bids (currently 12) and then assigns those to the current regular season favorites outside of NCAAT at-large bid territory who are most likely to lose in their conference tournament. Currently, Chattanooga is one of those 12. For NCAAT automatic bids, the model just takes the current conference tournament favorite. Chattanooga also fits that description. It’s confusing, but it’s the best solution we’ve found that reflects the prominence of the NIT automatic bids.
Chatt is the favorite to win the Socon Tourney but they are one of the regular season champs most likely to lose in the tourney. Got it.
Yes, roughly 40% likely to win the SoCon tournament—better than any other team—but therefore about 60% likely to lose, which would give them an NIT automatic bid—one of the highest NIT auto-bid probabilities in the country.
criminal that you have michigan instead of rutgers
Michigan’s now 13-10 (7-6 and eighth in the Big Ten standings; 2-6 in Quad 1 games and 30th in NET rankings before today). Five of Michigan’s final seven games are against teams above them in the Big Ten standings with four Top 20 matchups (Illinois and Michigan State at home, at Wisconsin and Ohio State) plus a home game with Rutgers and home-and-home with Iowa.
Although Rutgers has a much better overall, conference and Quad 1 record (15-9, 8-5, 6-3) than Michigan, their NET ranking before today was 94 due to their three terrible losses in November to NET 103 DePaul (73-70), NET 320 Lafayette (53-51) and NET 176 Massachusetts (85-83) and three losses in January to non-tournament, Big Ten teams (66-49 at NET 89 Penn State, 68-65 at NET 101 Minnesota and 68-60 at home against NET 104 Maryland.)
Rutgers seems to be safely into the NIT. Michigan might have to win at least four of its final seven games before the Big Ten Tournament to finish no worse than 17-13 to remain close to the NCAA Tournament bubble and guarantee an NIT bid. They can’t do worse than 2-5, or they’d need to win at least three games in the Big Ten Tournament to qualify for the NIT with a record of at least .500.
Why no WKU? They are tops of the CUSA right now. No reason why North Texas should be considered the ‘automatic qualifier’
Our model uses KenPom to simulate games, and in our simulations, North Texas wins the Conference USA Tournament more often than Western Kentucky does. (UNT wins about 30% of the time, WKU wins about 20% of the time.) That’s how we choose to do it, but going with the current leader in the standings makes sense too. Just a preference thing between different sites.
Joe,
Another good bracket, your NCAA projections are way better than your NIT projections. I don’t disagree with a single team you have in your NCAA bracket. But if you want to put lunch on your NIT bracket with Minnesota making it I’m all for it, we can do online Panera or Taco Bell, under 10 bucks.
Hey Peter, thanks for your continued engagement with the site. I feel like I haven’t adequately explained to you why Minnesota continues to be in our model’s NIT Bracketology, so here goes: Every projection we put on these pages—the probabilities and both the bracketologies—is made by our model. I’m not making subjective decisions on which teams are projected to land in each tournament. I did make decisions regarding how the model is constructed, some of which were subjective, but I’m not placing these teams into brackets every day. The model is. Obviously, the model can and probably does have its shortcomings. We made the decision not to code in what happens in unusual circumstances. I wrote about these in our “How Our College Basketball Model Works” post, which is linked at the top of this page, our NIT Bracketology page, and our College Basketball Probabilities page. If you haven’t read that post already, it has longer explanations of everything having to do with this model than anything I’m fitting into a comment. Anyway, a résumé like Minnesota’s is one of these unusual circumstances. There’s limited evidence suggesting there are certain thresholds below which win-loss record is weighted disproportionately heavily. We don’t have enough data to know with certainty, though, which is why this facet of the selection process is not built into our model this year. With the NIT and the .500 threshold in particular, there have only been three tournaments since, to our understanding, the no-teams-below-.500 rule was removed. In those three seasons, no team had a résumé on par with Minnesota’s in the areas our model weights most heavily despite having a losing record. Both are in uncharted territory. I don’t know what the committee will do with Minnesota if they finish under .500 (which is more likely than not at this point), and I don’t have much of an opinion on what they should do (though I hope we can learn more from this year’s tournament and make our model stronger next year). For whatever it’s worth to you, our model—which was built on data from recent tournaments and wasn’t coded to consider unusual circumstances—has their most likely placement a three-seed in the NIT at this time. I’m not going to make any kind of bet with you. I’m sure our model has shortcomings (I don’t know if this Minnesota thing is one of them, but it could be), and the goal is always to make it better. For the NIT specifically, our expectation this year is that its final NIT Bracketology, made following the NCAAT Selection Show, will have roughly 30 or 31 teams correctly in the field, 19 or 20 on the correct seed line, and nine or ten within one seed line of the correct seed line. Make of that what you will.
Best bracket you have produced, pretty spot on as we stand today. Lots could change, I keep hearing Rutgers has to win another road game or two, I see a lot of brackets WITHOUT Rhode Island, which to me is a complete head scratcher. I think the Arkansas/Miss St./Alabama SEC trifecta of bubble teams could get real interesting. NC State, Richmond, Providence and yes Purdue have legit shots at sneaking in with a good run these last two weeks. Keep up the good work!!
Glad to hear you’re feeling confident in our model! Trimming the error margins did iron out a lot of the oddities—we should have done that sooner. Thanks for continuing to visit the site.
Replace Purdue with USC, Minnesota with Arizona St. and Miss St. with Richmond and you have a good bracket.