Analysis | The perfect bracket to win your March Madness men’s pool

Before we continue, I want everyone to know I don’t have any hard feelings for all the heat I took last year after my projections placed San Diego State in the title game (correctly, as it turned out). The Perfect Bracket also hit on five more of the Elite Eight teams — Kansas State, Miami, Texas, Connecticut and Gonzaga — none of which were No. 1 seeds. I understand why you were skeptical, and I want you to know we can move on, mostly because I am taking the high road after being correct.

Instead of basking in the glory of my predictions, let’s turn our attention to how I came about rating San Diego State so high. The fifth-seeded Aztecs were ranked No. 16 in analyst Ken Pomeroy’s adjusted efficiency ratings before Selection Sunday, a similar figure to No. 3 seeds last year. The most similar efficiency profiles to San Diego State’s, according to Bart Torvik’s comparison tool, were Arkansas in 2022 and Louisville in 2015 — teams that both reached the Elite Eight, indicating the Aztecs had a lot of potential in pools as an undervalued team.

We will use similar techniques to make this year’s Perfect Bracket as perfect as it can be. We start by projecting each game and comparing win rates to what would be expected by seed matchups alone — with the aim of identifying teams that provide stronger upset potential and/or are likely to be overlooked by your competitors.

We also look at how similar teams performed in the tournament, using a weighted blend of the four factors — shooting, turnovers, rebounding and free throws — and comparing that to what’s expected of an average seed. (For more, read our methodology at the bottom of this story.)

For example, Purdue, the No. 1 seed in the Midwest region, is similar in performance to the 2023 Purdue squad that, as a No. 1 seed, famously lost in the round of 64. The Boilermakers also appear similar to five other teams that failed to survive the tournament’s opening weekend. In other words, it’s tough to trust Purdue in this year’s tournament, and you are probably better off selecting a different national champion.

Remember, the goal of the Perfect Bracket is not to be competitive in your pool; the goal is to be the highest–scoring bracket at the end of the tournament, thanks to a selection of teams that are more valuable than they appear on paper. In other words, it’s designed to be first or somewhere approaching last, with no in-between. So who is this year’s San Diego State? Let’s find out.

We have already talked about why No. 1 Purdue is probably not the best choice to represent this region in the Final Four, but in case you weren’t convinced, consider that Coach Matt Painter and the Boilermakers have been a single-digit seed in the last eight NCAA tournaments and a top-four seed in each of the last six. They’ve advanced past the Sweet 16 just once over that span, been eliminated by a lower-seeded team six times, and, in the last three years, seen their seasons end against a 13 seed, a 15 seed and, most famously, a 16 seed.

However, I don’t think there are enough pitfalls to keep the Boilermakers from advancing to the Elite Eight this time around. I think our energy is better spent looking at the rest of the top half of the region for value. And so how about this: No. 12 McNeese State and No. 13 Samford playing for the chance to lose to Purdue in the Sweet 16.

McNeese State has a good chance, relative to its seed, to get past No. 5 Gonzaga in the first round. No. 13 Samford could take advantage of an injury-riddled No. 4 Kansas squad. Senior center Hunter Dickinson and senior guard Kevin McCullar Jr. were unavailable in the Big 12 conference tournament, with Kansas getting routed by Cincinnati. And while Dickinson should be available against Samford, McCullar’s status is still up in the air. That’s significant, as McCullar leads the team in scoring (18.3 points per game). If Samford can capitalize on the Jayhawks’ injuries, then our bracket picks up a big win against the field. If Kansas prevails, we don’t lose much in these early rounds when chalk prevails, especially if Purdue advances to the Elite Eight.

The more important question is what to do about No. 7 Texas? If we think Purdue is weak, Kansas is impacted by injuries and the public will overwhelmingly support Purdue, No. 2 Tennessee and No. 3 Creighton, then selecting a lower seed to represent this region would make sense. The Longhorns were bounced out of the Big 12 tournament after one game, but remain efficient on offense with a good enough defense to earn respect. Coach Rodney Terry’s team moves the ball around (57 percent assist rate on field goals, 57th best in the nation, per Pomeroy) and doesn’t rely too heavily on the three-point attempts. When the Longhorns do fire from behind the arc, they are accurate, converting 36 percent this season. That formula could get you far in the tournament.

Let’s say Texas falls to the winner of the play-in game between No. 10 Virginia and No. 10 Colorado State or to Tennessee in the round of 32. What have we lost? The opportunity to have a No. 2 meet a No. 3 in the Sweet 16 with the region’s No. 1 waiting in the wings? If that happens, you’d have to be right on a lot of early upsets to pull away from the pack. But identifying a strong No. 7 seed as a potential winner of a region with a weak No. 1 seed will make our bracket stand out.

There are 11 conference tournament winners here, making this the toughest region in the bracket. It is home to the top overall seed, No. 1 Connecticut, which could be the first repeat national champion since Billy Donovan and Florida went back-to-back in 2006 and 2007. An argument can be made that this iteration of the Huskies is better than the one that won it all last season. The team’s offensive efficiency after adjusting for tempo and strength of schedule is higher, per Pomeroy’s metrics; its rim protection is better, and it blocks more shots. Plus, freshman guard Stephon Castle has established himself as a likely first-round pick in June’s NBA draft.

Other worthy contenders in this region include No. 2 Iowa State (the fifth-best team in the nation, per Pomeroy’s ratings), No. 3 Illinois (10th) and No. 4 Auburn (fourth). Of those, focusing on Auburn to advance to the Elite Eight presents the most value.

Connecticut is the chalky play, one that many brackets will lean on, which means Auburn could be a differentiator. The Tigers are good enough to be title contenders but obscure enough to not be very popular, especially in a region with so many other options. They don’t allow opponents to get many quality shots either at the rim (48 percent shooting against, 99th percentile per Synergy Sports) or behind the arc (30 percent shooting, 95th percentile) yet they don’t sacrifice performance on offense, either, scoring 120.6 points per 100 possessions after adjusting for strength of schedule (10th per Pomeroy’s ratings).

Believing Auburn will face Connecticut in the Sweet 16 opens us up to take a few more risks, albeit little ones, like No. 12 UAB defeating No. 5 San Diego State. If San Diego State wins, it shouldn’t impact our long-term prognosis much, since we’re depending on Auburn to advance to the next round. The sacrifice of a point for the possibility of correctly identifying a 12 seed to win a game is worth the risk, especially in pools that give bonuses for correct upset picks.

The real surprise could be in the bottom half of the region, where we’ll have No. 6 BYU advancing to the Elite Eight at the expense of fellow Big 12 team Iowa State. (If it makes you feel better, the Cougars beat Iowa State by 15 in January.) The Cougars’ efficient offense looks like it lives and dies by the three-point shot — they take more than half of their field goal attempts from behind the arc and convert on 35 percent — yet they are also very adept at scoring near the rim, which should help decrease their scoring variance on any given night. It also appears BYU was supposed to be the best No. 5 seed, but was moved down a line by the committee to accommodate the Cougars’ requirement to not play on Sundays.

One upset I highlighted earlier this week was No. 10 Drake over No. 7 Washington State. Drake should be able to bully the Cougars in the post and prevent them from getting many second-chance opportunities from offensive rebounds.

I think this is the region that gets weird. No. 1 Houston was embarrassed by Iowa State, 69-41, in the Big 12 conference tournament and doesn’t look as strong as it did earlier in the season, especially after several injuries. Still, Houston is one of the toughest defensive teams in the country, and the offense is led by guard Jamal Shead, who averages 13.1 points and 6.2 assists.

The Cougars will open with a No. 16 Longwood squad that is one of the best offensive rebounding teams in the country. Longwood is also adept at getting to the free throw line. Those two ingredients can be found in many tournament upsets over the years. Am I saying Houston is going to be the third No. 1 seed to be defeated by a No. 16 seed in the men’s tournament? No. At least I don’t think so. But the fact I am even considering it means we need to look elsewhere for our picks from this region — and perhaps at No. 7 Florida.

Florida Coach Todd Golden is very happy when his team gets into track meets. The Gators use almost 72 possessions per 40 minutes (18th in the country, per data from Pomeroy) and are quite efficient on offense, scoring 120 points per 100 possessions after adjusting for strength of schedule. I’d give them serious consideration as an Elite Eight pick.

It’s probably more valuable to pencil in No. 4 Duke to the Elite Eight. Kyle Filipowski is a matchup nightmare for virtually everyone in the country. There aren’t many 7-foot centers who can beat you down low, off the dribble and beyond the three-point line. He’s making more than half of his shots (51 percent, including 36 for 103 on three-point attempts) and is one of the ACC’s leading rebounders (8.2 per game) and shot blockers (1.6 per game).

Plus, based on the Blue Devils’ most similar efficiency profiles, they can be expected to win two games, on average, in the tournament. Houston can be expected to win 2.7 games — a notable difference, but close enough to justify taking a less popular No. 4 seed over the No. 1 seed in pursuit of maximizing the value of your picks.

No. 2 Marquette, by comparison, should be expected to win 1.9 games based on its most similar efficiency profiles, and No. 3 Kentucky should be expected to win 1.7 games. Again, this is about identifying value to make our bracket unique, and give us a realistic chance to finish on top.

If you are in a huge pool, one with thousands of entrants, consider advancing Duke and Florida to the Elite Eight. Both have their merits and that selection would go a long way to differentiating your bracket.

There is one added wrinkle in this region, with No. 10 Boise State competing with No. 10 Colorado in a play-in game, with the winner set to face Florida. In the right circumstance, Colorado would be a slam-dunk pick for me to advance at least to the round of 32. The Buffaloes are the 26th-best team in the country, per Pomeroy’s ratings, and are balanced on both sides of the court.

It would be a great story for our Washington-based readers if No. 12 James Madison were to defeat No. 5 Wisconsin, but that’s a tall order. Badgers guard AJ Storr, a sophomore transfer from St. John’s, leads the team in scoring, while freshman John Blackwell gives the Badgers a lethal scoring threat off the bench (47 percent shooting behind the arc) they lacked in previous seasons.

The top three seeds in this region — No. 1 North Carolina, No. 2 Arizona and No. 3 Baylor — are all very solid. A tenacious defense propels the Tar Heels, who are efficient at grabbing defensive rebounds (24 percent of opponents’ misses, ranking 14th in the country) and defending in transition (allowing just 87 points per 100 possessions on fast breaks, per Synergy Sports — one of the best rates in the country). Both Arizona and Baylor rank in the top eight of Pomeroy’s offensive efficiency ratings.

However, I still think we take some chances in this part of the bracket, specifically with No. 5 St Mary’s.

Fueled by a defense that ranks 16th in the country in Pomeroy’s defensive efficiency ratings, the Gaels produced a 16-game winning streak. They are great at chasing opponents away from the three-point line (29 percent of field goal attempts allowed, eighth in the country), preventing second-chance opportunities and protecting the rim (less than a point per possession puts them in the 98th percentile, per Synergy Sports). They also like to slow down the pace on offense, keeping even the most efficient offensive opponents from getting many possessions.

This could be a bracket-defining pick like San Diego State was last year, with this St. Mary’s squad having similar underlying factors to Houston in 2022 (a team that made the Elite Eight) and reigning national champion Connecticut. Ultimately, I think Arizona will be this region’s representative in the Final Four, but the opportunity to get value from a team like St. Mary’s is hard to ignore.

I was bullish on No. 4 Alabama early this season, but my optimism has soured heading into the tournament. The Crimson Tide rely heavily on the three-point shot (47 percent of field goal attempts, 16th highest). It hasn’t punished them yet, but all it takes is one off night to send such a team home early. The Tide also sends opponents to the free throw line often, something that can assist underdogs. As we have mentioned before, extra possessions and chances to score for the less talented team are a key ingredient when looking for potential upsets.

There are other upsets in the West that are intriguing, too, such as No. 11 New Mexico over No. 6 Clemson and No. 10 Nevada over No. 7 Dayton. New Mexico is among the 25 best teams in the country, per Pomeroy’s ratings, and the matchup against Clemson is a good one for the Lobos — particularly because of their ability to grab offensive rebounds and create turnovers. Nevada, meanwhile, is one of the more experienced teams in the tournament, and shouldn’t let Dayton run amok from behind the arc.

The Final Four for the perfect bracket is unconventional, with No. 4 Auburn and No. 2 Arizona facing off in one semifinal, and No. 4 Duke meeting No. 7 Texas in the other. Yet when you consider that the consensus picks would include all four No. 1 seeds advancing, with Connecticut winning it all on one out of every four brackets as of Monday afternoon, our bracket makes perfect sense. If you followed the “People’s Bracket” last year — the most popular picks for every round in ESPN’s contest — you would have finished with 45 points in standard scoring systems. My bracket was far different from the consensus, and finished with 96 points.

Auburn’s résumé stacks up well against any of the No. 1 seeds, and the Tigers are the seventh-best team in the country according to the consensus of 58 rating systems. Arizona is fifth, Duke is eighth and Texas is 24th. Remember, the Texas play is situational based on popularity and injury concerns. Plus, we have the Longhorns losing to Duke in the Final Four, so while an earlier loss by them wouldn’t be ideal, it wouldn’t be crippling, either.

The key pick, of course, is Auburn beating Duke for the national title. If that happens, it will be another good year for the perfect bracket supporters.

The Perfect Bracket intelligently selects upsets by projecting each individual matchup from the ground up, starting with an estimated number of possessions for each team and taking into account any additional possessions to be had via offensive rebounds and turnovers. Then, estimated scores are derived by adjusting each team’s points scored and allowed per 100 possessions for strength of schedule.

Once we know the projected scoring margin, we can infer an implied win percentage. For instance, teams that are favored by two points would have an expected win probability of 57 percent. That rises to 77 percent if the predicted scoring margin is seven points.

Finally, these win rates are compared to what we would expect by looking at seed matchups in a vacuum, with teams providing stronger upset potential given more weight. Strong teams that are likely to be overlooked by your competitors are also given extra weight.