The madness of NFL free agency has just about given way to the madness of the college basketball playoff. Just like last year, I’m leveraging advanced team metrics and simulations to find value and potential upset picks for your 2026 March Madness bracket. For a more detailed look at the methodology, check out the introductions to the 2024 and 2025 articles.
Below I’ll present our best estimates for tournament advance probabilities — what is the likelihood that a given team makes it to a given round? Comparing that information with publicly available data on which teams most people are picking can help us find leverage points and possible upsets.
East Region
For the second year in a row, Duke is the best team in the country. Unlike last year, our simulations this time around actually say they are the most likely team to win the tournament. We think they have about a 19% chance to win it all. KenPom and EvanMiya both have as their odds at around 21%. Neil Paine is at 24%. Bart Torvik has them at 20%. JThom agrees with our number — about 19%. The point is, most prediction systems are pretty close — Duke’s chances to cut down the nets in early April are around 1 in 5. However, it’s not an easy path.
First, Duke is dealing with some injuries, including to their starting point guard Caleb Foster, who missed the entire ACC tournament. Duke’s margin of victory in the regular season was over 20. In the ACC tournament, it was less than 6. That’s a small sample, to be sure, but it’s bolstered by advanced stats indicating they turned the ball over more (their turnover rate increased by nearly two percentage points) and shot the ball worse (their effective field goal percentage decreased by about seven percentage points). Throughout the season, Foster’s presence on the floor was worth about 3.8 points per 100 possessions. However, Duke was about 5 points per 100 possessions worse when backup point guard Cayden Boozer was on the floor. Foster hopes to return for the Final Four, but Duke has to get there first.
The Blue Devils were also missing Patrick Ngongba for the ACC tournament, but he is expected back for the national tournament. Duke’s defensive efficiency improved by more than 10 points per 100 possessions when Ngongba was on the floor. His return is crucial for Duke’s tournament outlook.
In any event, the path to the title game goes through some difficult competition. For Duke to make a run to the Elite 8, they would likely have to face St. John’s. They beat Duke head-to-head in about 30% of our simulations. And of course, the Blue Devils have to get to the Sweet 16 first. That could also prove difficult, as they are knocked out by Ohio State in 32% of our simulations.
Value Picks
No. 5 St. John’s: The Red Storm get to the Elite 8 in 14% of our simulations. Less than 10% of Yahoo brackets picked them to advance that far.
No. 8 Ohio State: The Buckeyes have a 20% chance to make the Sweet 16, according to our numbers. Less than 4% of Yahoo brackets have them there.
No. 6 Louisville: Yahoo pickers have the Cardinals going to the Sweet 16 in just over 15% of outcomes. Our simulations see them in the Sweet 16 about 37% of the time. They are No. 9 in offensive shot quality and No. 18 in defensive shot quality. Only eight teams are inside the top-20 in both metrics. Louisville will be playing without Mikel Brown Jr., which dampens some of the enthusiasm, and could even see them exit in Round 1 against South Florida, one of the best offensive rebounding teams in the tournament.
Upset Watch
No. 14 North Dakota State (vs. No. 3 Michigan State): The Spartans are always one of the best rebounding teams in the nation, and this year is no different. What’s interesting about this matchup is that North Dakota State is also one of the best rebounding teams in the nation, particularly on the defensive end. They also force a lot of turnovers — something Michigan State has struggled with. Because they turn teams over and don’t give many second chances, the Bison are No. 8 in the country in defensive shot volume index (dSVI). Only six teams are better in net shot volume (offensive SVI minus dSVI), and Michigan State isn’t one of them (they rank 28th). Less than 4% of Yahoo brackets have picked this upset, but our calculations give it about a 17% chance of happening. That said, betting against Tom Izzo in March may not be a wise decision.
South Region
The South Region is unique in that the 1-seed isn’t obviously the best team in the region, or even the second best, according to our ratings. Both Houston (2-seed) and Illinois (3-seed) are expected to win at least half the time against Florida (1-seed) on a neutral court.
Illinois is No. 1 in the nation in oSVI. They are No. 2 in the nation in offensive shot quality. They are No. 3 in kill shots per game, an EvanMiya metric that measures how often a team goes on a 10-0 run. They are the No. 6 team according to Bart Torvik, No. 7 according to KenPom, and No. 8 according to EvanMiya. Our ratings have them ranked No. 4 in the tournament.
Houston is the No. 5 team in our numbers as well as those of Torvik, KenPom, and EvanMiya. Only four teams have a higher net SVI rating. Only two tournament teams are better in defensive shot quality. Only three teams turn the ball over less often, and only 11 force more opponent turnovers.
The good news for Florida is that the Gators won’t have to face both Houston and Illinois — those two teams will most likely square off in the Sweet 16. But if either team emerges from that half of the South bracket to face the reigning national champions, our simulations suggest it’s a pick ‘em at best.
Value Picks
No. 3 Illinois: Illinois is arguably the best value pick in the entire tournament. Our simulations give them about a 33% chance to reach the Final Four. Only about 10% of Yahoo brackets have them going that far. They are the eventual champions in 10% of our simulations, but only 1% of Yahoo users picked them to win it all.
No. 5 Vanderbilt: Only 10% of Yahoo’s bracketologists picked Vanderbilt to advance to the Elite 8. Our simulations put them there about 24% of the time.
No. 7 Saint Mary’s: Houston is one of the better teams in the field, but in about 20% of our simulations, it is Saint Mary’s, not Houston, that emerges to advance to the Sweet 16. Less than 7% of Yahoo brackets agree.
Upset Watch
No 12. McNeese (vs No. 5 Vanderbilt): Although Vandy is one of the top values in the South Region, they aren’t a lock to get out of the first round. McNeese is No. 1 in the country in net shot volume. They lead the nation in opponent turnover rate. On the other side, they take care of the ball better than most teams and rebound well enough to give themselves plenty of second chances. Vanderbilt is a mediocre rebounding team at best, and they often struggle to defend without fouling. Only one tournament team sends opponents to the free-throw line more often. As a team, McNeese shoots almost 75% from the line — above the NCAA average. On the other hand, McNeese also sends teams to the line a lot, and Vanderbilt shoots almost 80% as a team (No. 4 in the nation).
West Region
Another region where the No. 1 seed isn’t the favorite to make the Final Four, the West Region features a very strong No. 2 seed in Purdue. The Boilermakers have the most efficient offense in the country, thanks in large part to their No. 1 rank in offensive shot quality. What’s particularly interesting is that they’ve earned that spot without relying on threes and shots at the rim. Only 75% of their shots have come at the rim or from beyond the arc, which ranks 328th. The implication is that they are able to create wide open shots from so-called suboptimal locations, and to do so to such a degree that the location of the shot doesn’t have a major impact on the quality score. This might make their offense more resilient against tough tournament defenses. For what it’s worth, there’s a correlation of about 0.2 between offensive shot quality score and rim-and-three rate. Purdue is proving that’s not a necessary connection.
Our numbers have Purdue winning the entire tournament about 9% of the time. Yahoo users only have them in the winner’s circle in about 3% of brackets. Arizona’s path would have to go through Purdue, but they are still the fifth-most likely team to hoist the championship trophy. However, over 20% of Yahoo brackets have picked them to win, so they’re one of the worst values.
Value Picks
No. 4 Arkansas: Arkansas’ path to the Elite 8 goes through Arizona, so it’s not an easy feat by any stretch. Still, we project them there in 25% of cases — only about 10% of Yahoo brackets see them getting past Arizona.
No. 11 Texas: Play-in winners are always underpicked, in part because many brackets are entered before they have become play-in winners. We have Texas in the Sweet 16 about 17% of the time. Yahoo’s collective prognosticators only give that about a 3% chance of happening.
Upset Watch
No. 12 High Point (vs. No 5 Wisconsin): High Point is No. 3 in net shot volume. They are fifth-best in the nation at taking care of the ball and fourth-best at forcing turnovers. They shoot well from the field and get to the line often (and make their free throws). On the other side, Wisconsin does a good job of avoiding turnovers but struggles to force many turnovers. They are not a good offensive rebounding team, so they won’t be picking up many extra possessions. And they shoot a ton of threes, which can be a high-variance strategy. They may also be playing without forward Nolan Winter, one of their most impactful players. The Badgers were 16 points per 100 possessions better when he was on the floor. If he’s unable to return in time for Wisconsin’s Round 1 matchup, there may not be another chance this season.
Midwest Region
Like Duke, Michigan is dealing with significant injuries. However, L.J. Cason’s impact on Michigan’s defensive efficiency is actually a net negative, at least according to the numbers. The Wolverines give up about 4.1 fewer points per possession when he’s not on the floor. That said, a recent loss to Purdue and close games against Wisconsin and Ohio State remind us that this 1-seed is vulnerable.
This region also features a very strong No. 2 seed. Iowa State is No. 2 in net shot volume and No. 2 in dSVI. Only three teams force more turnovers than Iowa State, and none are better in defensive shot quality. The Cyclones lead all teams in kill shots per game — they’ve gone on 42 10-0 runs this season. Moreover, they are the No. 11 team in the nation in three-point percentage. Our simulations have Iowa State appearing in the title game about 11% of the time. Yahoo users have only advanced them that far in about 5% of brackets.
Value Picks
No. 5 Texas Tech: The Red Raiders advance to the Elite 8 in 14% of our sims. That’s not a huge number, but Yahoo brackets are underestimating the likelihood that Texas Tech can beat Michigan. They think it happens in less than 5% of cases.
No. 11 Miami (OH): The RedHawks enter the tournament with just a single loss on the season, and their reward is that they get to play an extra game. Play-in winners get no love from Yahoo bracket pickers. Even after defeating a tough SMU team, few users at Yahoo will go back and change their brackets, meaning there’s still significant value on Miami. Our numbers give them about a 14% chance to make the Sweet 16. Less than 3% of Yahoo brackets have picked them so far.
Upset Watch
No. 10 Santa Clara (vs. No. 7 Kentucky): Santa Clara is only a slight underdog according to our simulations — they beat Kentucky about 47% of the time. The Broncos are among the better teams in the tournament in both offensive rebounding and forcing turnovers, which lets them generate a lot of extra possessions. Kentucky is only average at keeping opponents off the offensive glass. They turn the ball over on 15% of their possessions — solid, but not great. Nearly 70% of Yahoo brackets picked Kentucky to win, so there is significant value in picking the underdog in what we think is close to a toss-up. In fact, Santa Clara is actually the better team according to Bart Torvik’s numbers.
