If you want a sure thing in your men’s NCAA tournament pool, you’ll need to fill out the 9,223,372,036,854,775,808 brackets necessary to guarantee a winner. Just leave yourself plenty of time to finish them all: if you filled out one bracket every second it would take you 292 billion years to cover all the possibilities. Instead, just study up on this perfect NCAA tournament bracket that is guaranteed* to win your pool.
(* As we note every year, this might be more like a Patrick Ewing guarantee than a Joe Namath guarantee.)
The Perfect Bracket is more than just a bunch of game-by-game predictions - it also seeks to differentiate your picks from others in the pool by making selections that have higher value relative to conventional wisdom. The Perfect Bracket carefully selects upsets using DAViD, the Data-Assisted Victory Detector for the NCAA tournament (click here for the full explanation of the method).
Upsets are terrific at getting early separation from the rest of the pack, but don't go crazy in the early rounds. Since 2011, the first year the field expanded to 68 teams, the higher seed has won 71 percent of first- and second-round games, so go with the chalk early and often. And that includes avoiding any temptation to pick a No. 16 seed to upset a No. 1. Yes, I know all about UMBC's historic win over Virginia last year (I picked the Cavaliers to win it all. Ugh.), but No. 1 seeds are still 135-1 against the No. 16 seed in tournament play.
In fact, avoid selecting a No. 1 seed to be an upset victim in the first two rounds. Over the past eight tournaments, top seeds have a 56-7 overall record through the Round of 32. That's an 89 percent success rate. Don't advance the lower-seeded teams too far, either: the No. 14 and No. 15 seeded teams have gone 10-63 (14 percent win rate) against opponents in the first two rounds over the past eight years. And take the low win rate by favorites in the Elite Eight with a grain of salt; those matchups include No. 2 and No. 3 seeds beating No. 1 seeds, hardly a big distinction at that point in the bracket.
So, what does perfection look like? Here's a region-by-region breakdown of this year's perfect bracket.
This region is brutal. Any team looking to exit this part of the bracket has to contend with not only Duke, the overall No. 1 seed and Michigan State, the Big Ten winner, but also No. 4 Virginia Tech (11th best team per Pomeroy ratings), No. 8 VCU (seventh-best defense in the nation), No. 9 UCF (second-highest rate of field goals from free throw line, 46 percent) and No. 12 Liberty (12th best shooting team in the country).
Duke is the obvious favorite here. The ACC champion Blue Devils are the third-best team in the country per Pomeroy's rankings with quality wins over Kentucky and Virginia. Their only two losses, when at full strength this season, are to Gonzaga and Syracuse.
The offense, led by freshmen Zion Williamson and RJ Barrett, scored 122.6 points per 100 possessions (third-best) including a robust 63 percent field goal rate around the basket. In addition, no team was better at converting offensive rebounds into points (1.3 points per putback) than Duke, making it tough for underdogs to get extra possessions that typically fuel an upset.
Virginia Tech, a No. 4 seed, is also worth watching. The Hokies can stretch the floor (they score 1.2 points per possession in transition, only four teams are more efficient on the break) and hurt you from deep (40 percent from three-point range, eighth-best in the nation), opening up opportunities for fourth-year junior forward Kerry Blackshear Jr.
Blackshear is averaging 14.7 points and 7.4 rebounds per game resulting in a box-score plus minus of 10.2 this season, meaning he is worth 10.2 points more than a league-average player on an average team per 100 possessions. Blackshear has also helped pick up the pieces after point guard Justin Robinson was lost to injury, scoring 18.1 points per game in his absence.
But the team you'll want coming out of this region is Michigan State, since the Spartans provide the most value. (More on that below.)
This region should go according to plan.
The Tar Heels are the sixth-best team in the country per Ken Pomeroy's ratings and have lost twice, to Virginia in the regular season and Duke in the ACC tournament, since Jan. 13. North Carolina does have a tendency to surrender a large amount of shot attempts from behind the three-point line (43 percent of field goals against, 298th), but aside from No. 5 Auburn (43 percent from beyond the arc, seventh-most) and No. 7 Wofford (40 percent) there aren't many sharpshooting squads in the Midwest.
North Carolina will face stiff competition from No. 2 Kentucky. PJ Washington and Tyler Herro have fueled Kentucky’s rise to the top. Washington is averaging a team-high 14.8 points and 7.6 rebounds per game while shooting 42 percent from beyond the arc and Herro is averaging 14.2 points per game with a three-point success rate of 37 percent. The three-point percentages for both players are a key factor in Kentucky’s wins this season: according to Taylor Combs of kentuckysportsradio.com, Washington and Herro are 51 out of 109 from long range (47 percent) in Kentucky’s 14 wins against teams ranked in the top 100 of the NCAA’s NET rating. In five losses against those teams they are a combined 7 for 40 (18 percent).
The team most likely to upend this bracket is Wofford. The Southern Conference champion Terriers are in the AP Top 25 for the first time in program history, went 18-0 in league play, create 6.4 extra possessions per game via offensive rebounds and turnovers plus make 42 percent of their three-point attempts, fourth-best among Division I schools this season. Those are all attributes common to underdogs that are capable of pulling off an upset. Still, our projections show Kentucky with a 74-percent chance of advancing to the Sweet 16, so it doesn’t quite warrant the risk of taking the Terriers.
No. 8 Utah State could have gone far if not for its draw. They create almost three extra possessions per game off offensive rebounds and turnovers plus are adept at getting open looks at the rim. But because they will almost certainly face North Carolina if they advance to the Round of 32, their chances of being one of this year's surprise teams is greatly diminished.
No. 3 Houston had the fourth-highest NET ranking per the NCAA but was the 10th best team per the consensus ranking of 67 different metrics. Their first-round opponent, No. 14 Georgia State is one of the best three-point shooting teams (39 percent, 15th) in the country, with three players - Devin Mitchell, Malik Benlevi and Jeff Thomas - all shooting over 40 percent from behind the arc on five or more three-point attempts per game. Plus, junior guard D'Marcus Simonds, the 2018 Sun Belt Player of the Year, was named to this year's All-Sun Belt first team after is averaging 18.9 points, 4.9 rebounds, 3.7 assists and 1.3 steals per game. His ability to score down low in the post (1.1 points per possession, including passes) gives the Panthers a versatile offense that is tough to stop.
Plus, Georgia State's defense creates extra possessions via steals (11 percent, 27th), giving them even more opportunities to punish teams from deep. As a result, Houston may have a problem getting out of the first round.
This part of the bracket looks to be ripe for upsets. The top seed in the region, the Virginia Cavaliers, likes to grind down opponents and play at a snail's pace. But that's a problem. Since 2011, there have been eight teams seeded No. 1, No. 2 or No. 3 that have also ranked in the bottom 20 nationally for adjusted tempo. One was last year's Virginia squad. Two others, No. 3 Syracuse in 2014 and No. 2 Virginia in 2015, lost in the Round of 32. Just two of the eight, No. 1 Wisconsin (2015) and No. 3 Michigan (2018), made it to the national title game, though neither won.
Research by Eli Boettger also showed that teams in the top 50 for tempo accumulated slightly more wins than their seed would suggest in the NCAA tournament, while slow-tempo teams like Virginia (300th or lower in tempo) won fewer than expected.
That could open the door for No. 2 Tennessee to emerge as the region's representative in the Final Four. Grant Williams, a national player of the year candidate, is averaging 19.2 points, 7.5 rebounds this season and has improved his true shooting percentage from 54 to 65 percent in just one year. Few players can neutralize him in the post (1.2 points per possession, 98th percentile) and he has enough range to step outside the arc and hit three-point shots (38 percent). He's also adept at getting himself to the free throw line (7.1 fouls drawn per 40 minutes, 10th best rate in the nation). Defensively he held opponents to 41 percent shooting around the rim in 2018-19, making him the type of balanced player that can take a team far in the tournament.
No. 5 Wisconsin could also surprise. The Badgers had a rough start to 2019 - Wisconsin lost three of the first four games of the new year - but have righted the ship and enter the tournament with the 12th best adjusted net efficiency (plus-23.7). Coach Greg Gard sees his squad take care of the ball (14 percent turnover rate on offense, eighth best), play stout defense around the rim (less than a point per possession allowed, 98th percentile) and against ballhandlers on the pick and roll (0.7 points allowed per possessions, 85th percentile), the most frequent play against them this season.
Other upset possibilities in this part of the bracket include No. 13 UC Irvine over No. 4 Kansas State (23 percent chance, more than twice what we would expect from an average 13 seed), No. 10 Iowa over No. 7 Cincinnati (52 percent) and No. 11 Saint Mary's over No. 6 Villanova (49 percent).
Gonzaga had a disappointing ending in the West Coast Conference men's tournament, losing to St. Mary's in the final, but they still own the No. 1 offense in America after adjusting for tempo and strength of schedule (125.1 points per 100 possessions) and can boast about having one of two wins in the country over No. 1 Duke at full strength for an entire game. Led by All-America candidates Brandon Clarke and Rui Hachimura, two players ranked in the 98th percentile for scoring efficiency (1.3 and 1.2 points per possession respectively), Gonzaga can beat teams in transition, around the basket, down low in the post and off the dribble.
However, the Bulldogs allowed No. 2 Tennessee and No. 1 North Carolina to shoot a combined 61 percent from behind the arc this season, both resulting in losses, signaling the Bulldogs could have trouble with some of the better-shooting teams in the tournament.
Other formidable opponents in this region include No. 2 Michigan and No. 4 Florida State.
Michigan is short on experience (1.5 years on roster, 239th lowest in NCAA) and relies heavily on its starters (18 percent of minutes played by reserves, 350th) but they do have one of the nation's best defenses per Pomeroy (2nd) and don't put opposing teams on the foul line too often (25 percent, 11th).
The Seminoles' depth is formidable. The team's reserves play more than a third of the game (38 percent of minutes, 30th most this season) with 6-foot-10 sophomore center Mfiondu Kabengele leading the team in points scored per 100 possessions (35.2) despite not starting a game this season.
No. 3 Texas Tech, however, has a lot of value here. The Red Raiders enter the tournament with the best defense in the nation per Pomeroy with high marks for turnover rate (23 percent, 11th), rim protection (48 percent against, 92nd percentile per Synergy Sports) and block percentage (15 percent, seventh). Texas Tech can light up the scoreboard, too, (36th best offense per Pomeroy), so don't think this is just a one-dimensional squad primed for an early-round disappointment.
Sophomore guard Jarrett Culver leads the team in points (18.5) with 6.3 rebounds and 3.6 assists per game and is one of the most dominating players in isolation, scoring 1.1 points per possession this season. Davide Moretti is the only player in Big 12 history to finish the season shooting 50 percent from the field, 45 percent from three-point range and 90 percent from the free throw line.
The Red Raiders have won nine of their last 10, the only loss to West Virginia in the Big 12 tournament, and appear ready for a sustained run in this year's tournament.
Final Four and beyond
A perfect Final Four is nice to have for your pool, but not essential. Over the past eight years only a fraction of people (0.25 percent) have gone 4-for-4 and that was even lower last year (0.1 percent). You're much more likely (23 percent over the past three years) to get two right, preferably the same two teams that will face one another in the championship game.
On paper, No. 1 Virginia has the best chance to make the Final Four, appearing 52 percent of the time as the representative of the South region. No. 1 Duke (42 percent) and No. 1 Gonzaga (41 percent) are next, followed by No. 1 North Carolina (37 percent) and No. 2 Michigan State (33 percent). However, we are looking for value, not necessarily chalk, which is why Michigan State is an intriguing pick: No. 2 seeds are picked, on average 11 percent of the time to make the Final Four, yet the Spartans have a likelihood that is triple that, making them a value play.
No. 2 Tennessee (24 percent) and No. 3 Texas Tech (21 percent, 6 percent average for 3 seeds) are also offering probabilities that are higher than you would expect based on seed alone.
The most important pick in any bracket is the national champion. In most scoring systems, choosing the correct national champion is worth the same number of points as going 32-for-32 in the first round, and, according to data from the past eight years of the official bracket game of the NCAA tournament, every one of the past eight winners had the participants and winner of the national championship game right. Luckily, winnowing down the 68 teams in the tournament to a select few who should be the national champion is relatively easy.
Over the past eight years, every national champion except one, Connecticut, a No. 7 seed in 2014, was a No. 1, 2 or 3 seed. Since 1985, the first year the field was expanded to 64 teams, all but four of the 34 winners were one of the top three seeds in the tournament; 21 of the 34 (62 percent) were No. 1 seeds.
Winners have also played in one of the top five strongest conferences per the Simple Rating System, a schedule-adjusted margin of victory rating that is expressed in points per game, with an SRS of zero indicating an average team. And all but three of the past 17 winners have had their own, individual SRS rank in the top four nationally.
The qualifying conferences this year include the Big Ten, ACC, Big 12, Southeastern and Big East, indicating Duke (ACC), Virginia (ACC), and Michigan State (Big Ten) are the best bets to become this year's national champion. You could add North Carolina (ACC) to the list since they are only a few decimal points shy of Michigan State for the fourth-highest SRS of 2018-19.
Everyone will be on Duke - the Westgate SuperBook in Las Vegas has them as a 2-to-1 favorite with Gonzaga the next choice at 5-to-1 - and Virginia's slow tempo makes them vulnerable again in the early rounds. That leaves Michigan State as a solid title contender that will also be overlooked on most brackets: less than five percent of ESPN brackets rely on a No. 2 seed to win it all.
In addition, many pundits considered Michigan State worthy of a No. 1 seed. The Spartans are the fourth best team in the nation per Pomeroy's ratings with both their offense (4th) and defense (8th) ranking in the top 10. In addition, there is almost nowhere on the court the Spartans don't excel. The team ranks in the 93rd percentile around the basket (1.3 points per possession) and in the 89th percentile on shots beyond the three-point line (1.1 points per possession, 37 percent shooting). Michigan State also finds a way to score on guarded catch-and-shoot attempts (1.1 points per possession, 83rd percentile).
Their star, Cassius Winston, has some health concerns but despite nursing two lower body injuries he tallied 14 points and 11 assists in the Big Ten title game; he's also one of 10 semifinalists for the Naismith Trophy, awarded each season to the top player in the nation after averaging 18.8 points and 7.6 assists per game.
The loss of forward Kyle Ahrens hurts, he was taken from the court on a stretcher in the first half of the Big Ten title game, returning on crutches in the second half, but a bench player using 13 percent of the team's possessions when he is on the court shouldn't alter their title hopes to a significant degree.
Instead, bank on one of the best teams in the tournament to fly under the radar all the way to the national title.
Neil Greenberg is a staff writer with The Washington Post whose beat is sports analytics. His analysis and insight can be found on the Fancy Stats blog, where he covers all pro sports, as well as college football and basketball.