PAST EDITIONS: 2021, 2022, 2023, 2024, 2025.
If you’d like to know more about how this field shapes up historically, read this.
Will Warren
On with the show.
Well, as mentioned, I hope you're not coming to this being like "I need to find my 11 upsets." (Also, a reminder: an upset is a 4+ seed difference. If you see anyone call a 10 over 7 an upset, delete them from your following list immediately. They are not a serious person.) If we get six upsets, it might be a minor overperformance based on the structure of the field and this college basketball season. This will probably mean a great second weekend and Final Four, but I'm not expecting much from the Round of 64 at all.
This year, there are just 3.0 projected upsets among the 11-16 seeds. Last year, I had a minor freakout because there were only 4.28 projected upsets, the second-fewest ever at that time behind 2017 (4.22). The lowest in the last 30 years is one (2000), followed by two (2007). It just has to beat these two to be less boring...though last year only had three upsets, too. At just 0.8 expected wins from the 13-16 seed lines, that's also the lowest ever, beating last year's previous low of 1.25. Getting nauseous yet?
The good news, as a headline, is that we can strongly bet against an all-1 seed Final Four again. The odds of this are just 4.6%, and even the odds of the Big Three 1 seeds (Duke, Michigan, Arizona) all in the Final Four together is just 14%.
For whenever this gets picked up by people who don’t normally read this website, many of these picks will be wrong. Even the very best brackets miss on an average of 13-15 picks out of 63 total a year. If I missed on 15 total picks, I would be beyond thrilled. Also, who cares, it's a bracket. I look forward to seeing how it all unfolds.
As a reminder, here’s how all of this works: What this is is simply a game-by-game projection of the field of 68 based on a document I’ve put together since 2018, using the odds to try and find hidden value in your bracket based on under-picked teams by the public. I’ve accumulated KenPom odds, based on pre-tournament data, from 2002 to present.
First Four
Unless you’re in a bracket group that I’m not in, you can’t actually make these picks. Alas, I do know that my bread is buttered by gamblers, so these are included in the order of game time. The odds I'm including here are just whatever I see when I search "college basketball odds" as a non-gambler.
(16) Howard (-1.5) over (16) UMBC. I have no strong feeling on this, as they're both about equal at KenPom and UMBC is actually favored there. However, Howard projects to earn somewhere around 3-5 extra possessions in this game compared to the Retrievers, which is enough to help me lean their way in an otherwise dead-even affair.
(11) Texas (-1.5) over (11) NC State. Zero real feel here. Again, this is functionally dead-even, with the two teams separated by three spots at KenPom. Shot volume tracks as a draw. I lean Texas because of their +5.5% delta from 2PT land, and in their defense, they could purely dominate NCSU down low/on the glass. It's just that NCSU could also win this by pure dominance in the backcourt. Neither result here would surprise me.
(16) Prairie View A&M (+3.5) over (16) Lehigh. An upset of sorts! Again, I have no real vibe on these First Four games that are all functional toss-ups...but here, PVAMU projects to earn around six more shot attempts than Lehigh, almost entirely through turnover margin. Both defenses are awful at shot suppression, however, so I would wager sight unseen that this is pretty high scoring if neither team can stop the other from getting second chances.
(11) SMU (-6.5) over (11) Miami (OH). I would like the RedHawks chances more if they were the style of mid-major team that ran up the score in the shot volume department, but SMU projects to get more shots here. Then again, Miami has the Mandate of Heaven. If anything, as a numbers guy, I'd like Miami to cover this +6.5...but I can't quite get to outright. It feels like a game where Miami acquits themselves well but can't pull it out.
Round of 64
East Region
(1) Duke (-28.5) over (16) Siena. 25+ point favorites in the NCAA Tournament are obviously 56-0, but perhaps more interesting is that not a single one of those 56 games saw the 1 seed win by less than 11 points. In fact, we have as many 44+ point wins (three) as we do games within 15.
(8) Ohio State (-2.5) over (9) TCU. This is a simple "pick the better team" wager. 8 or 9s at 60% or better to win, as OSU is here, are 8-4 since 2010. Now, before that, they were 1-7. It wouldn't shock me if TCU runs up a huge shot volume edge, but OSU will have a serious shooting edge of their own, and as one of the less three-point reliant teams in the field, their variance isn't as high as others. It's also the worst shooting offense in the Big 12. I have this as worth all of one point as I'd take neither over Duke, so don't stress it too much.
(12) Northern Iowa (+10.5) over (5) St. John's. This is not one that I adore, but it's the only game in the 1-5 seed lines where I think hidden value exists. For one, Northern Iowa's metrics were tanked by a midseason injury to Tristan Smith. With him in the lineup, per Bart Torvik, they are clearly a significantly better team than the season-long analytics grade them out to be.
T-Rank power feature: barttorvik.com/trank.php?ye...
— Totally_t_bomb (@barttorvik.bsky.social) 2026-03-08T19:47:53.489Z
If that's accurate, it would improve Northern Iowa's odds to 25% from 19% and would make them only a 7.5-point underdog. Given the status of the rest of the field, where arguably the best other 12 (McNeese) is matched up with the strongest 5 in Vanderbilt, this is the lone 12-over-5 that I feel I can recommend. There's also some matchup-specific stuff that intrigues me here. For one, this St. John's offense struggles heavily to score efficiently in the half-court, which is bad when you're playing the second-best defense at transition elimination and a top-20 defense in half-court defensive efficiency. This is a pick that could look extremely dumb, but we have 62% odds this year of a 12-seed winner, and aside from Texas Tech/Akron this is the only one I found myself giving serious consideration to.
(4) Kansas (-14.5) over (13) California Baptist. The 4/13s are where I'm just going to eat it this year if I'm wrong. The odds of the 4s going 4-0 are an astonishing 60%, easily the highest in history. If I miss on one, I'll own it, and this is the only one I pondered even for a few seconds. (The Troy/Nebraska thing feels like a sucker bet to me. I'd like to be wrong, because I love the Trojans.) Even so, this seems DOA. This is the 284th-best half-court offense at Synergy against a top-10 defense. While Cal Baptist is going to kill Kansas on the boards and it will be very funny, their path to actually winning the game would be to keep this below 65 at very highest, and I'm not sold on it.
(6) Louisville (-5.5) over (11) South Florida. This appears to be the cause du jour of College Basketball Twitter right now, taking South Florida in this game. I can certainly see the rationale! Louisville has been a moderate disappointment this year and struggles mightily to beat better teams. Here is my problem with said cause du jour: this is a bad transition defense...against Louisville. Re-read that for me.
Now, the same can be said of Louisville's own bad transition defense, but this is not as efficient of a USF offense as you may think. It's the Bulls defense that ranks higher, and against top-100 competition this season, the Bulls averaged 1.087 PPP...which, yes, is worrying when against everyone else you averaged 1.22. Furthermore, I worry about how their huge rebounding edge is usually neutralized by better competition, especially when drawing a Louisville team that's excellent on the defensive boards. The Cards project to have the rebounding edge here. I think this could be a monster variance game in both directions - they both love their threes, after all - and I am well aware of my own idiotic stat, that 13 consecutive Tournaments have had an upset in the first two TV windows. But South Florida isn't a top 25 team, which probably means Louisville will be fine.
(3) Michigan State (-16.5) over (14) North Dakota State. Worst possible draw for NDSU on the 3 line, as I thought they might be more live against other 3 seeds. NDSU is tremendous at running up the score on the boards, has a positive turnover margin, and has the size and skill to compete with the 3 line. The problem: they didn't draw Virginia, the 3 seed that was the least-dominant on the boards while also having a negative turnover margin. I would take MSU and the points here.
(7) UCLA (-5.5) over (10) UCF. I'm offering a grade of Incomplete to this game until I'm more aware of the statuses of Tyler Bilodeau and/or Donovan Dent. I appreciate Mick (and Jeff)'s notes here, but to be honest, coaches say this stuff all the time. It's like how BJ Edwards is going to magically come back from an injury that looked to end his season so SMU can be in the field.
UCLA’s Mick Cronin told @TheFieldOf68 that he expects both Tyler Bilodeau and Donovan Dent to be available for the Bruins first-round game in the NCAA tournament.
— Jeff Goodman (@GoodmanHoops) March 15, 2026
If UCLA is missing both, I'm picking UCF and living with the consequences. If they're missing one or the other, I probably still learn UCLA, as based on season-long numbers they're a hilarious 66% to win this game. If both can go: UCLA, no doubt. Quick back-of-the-hand math provides UCLA with around a 60% chance to win this game with one or the other out, but it obviously dampens my enthusiasm for their capabilities after the Round of 64. I may come back and edit this one, depending on what we hear by Thursday.
(2) UConn (-20.5) over (15) Furman. Seeing some people say Furman is live here. Boy, it would be really funny if they were...but no 2 seed above 94% to win their Round of 64 game has lost yet, and UConn is at about 97%.
West Region
(1) Arizona (-31.5) over (16) LIU. Fins Up! Here's hoping.
(9) Utah State (-2.5) over (8) Villanova. My really boring strategy for 8/9 games is just to pick whoever's favored, barring a really obvious shot volume mismatch or an injury. This and an additional game we'll cover later offer a strange combination of all three. Utah State is favored, has the edge on both the boards and in the turnover department, and Villanova lost starter Matt Hodge, a 6'8" stretch 4 that shot 40% from deep in Big East play, to a torn ACL. That injury increases USU's odds from 52% to about 54%, using Torvik data. The value pick here actually is the Aggies, as only 46% of ESPN users are taking them.
(5) Wisconsin (-9.5) over (12) High Point. This is the 12/5 game I've had the highest amount of pondering over. This Wisconsin team isn't built to run away from pretty much anybody, and even in the Big Ten Tournament they had to escape Washington 85-82 first to get to the Illinois/Michigan affairs. They seem to me to be the classic case of a "plays up or down to their competition" team.
Now, on one hand, Wisconsin has an 83% shot to win this game at KenPom, which makes them and Vanderbilt two of just 10 5 seeds ever to be at 80% or above. The previous eight went 6-2, though, including a relatively recent example in 2021-22 Iowa losing to Richmond. High Point offers a rare combination of intrigue here, with a Shot Volume Index north of +10 (just the eighth 12 seed to ever have this) against a Wisconsin team that doesn't rebound well and, again, plays up or down to whoever's on the schedule.
Ultimately, the game it appears closest to in my database is actually the aforementioned 2005 Michigan State/Old Dominion game, where ODU had a huge shot volume edge but simply wasn't good enough to beat their 5-seeded opponent. This WOULD fit my idiotic First Two TV Windows Upset Theory, though. I guess ultimately I'd lean conservative here because 1) the win rate for these high-end shot volume 12s still isn't great, and 2) Wisconsin is actually the least-picked 5 to win, per Yahoo's Pick Distribution.
(4) Arkansas over (13) Hawaii. On the other hand, this is a simple 'pencil it in and move on.' Of the 12 4 seeds at 85% or better to win in the Round of 32, 11 of them won and nine made the Sweet Sixteen. Arkansas, by a mile, has the best odds of anyone in these games, and unfortunately this isn't one of those games where the 13-seed is likely to produce the shot volume advantage. (I project Arkansas to generate around ~7 more shot attempts, meaning Hawaii would have to shoot 45% from 3 or Arkansas 20% for this to be fun.)
(11) Texas or NC State (+1.5 with either) over (6) BYU. Autoplay for me. Since Richie Saunders' injury, BYU has gone 5-5 and played at the level of the 44th-best team in America with the 105th-best defense. This isn't a crime, but it scares me that prior to said injury, the Cougars were playing like a top-25 team with a top-10 offense and top-20 results. If the former is what BYU truly is now, there is a real argument that either First Four winner should be favored in this game straight-up. If it's somewhere in the middle, this is still a game where BYU is ~55% to win, well below the average for 6 seeds.
The upside for BYU: both Texas and NCSU have sucked lately, too. They've played like teams in the mid-50s nationally, though I'm going to note here that Texas has much better two-way indicators and has been quite unlucky with opponent 3PT%. Really, this is about an insane thing I saw while preparing this: 79% of people have BYU in the Round of 32. 79%! 28% in the Sweet Sixteen. Is the Dybantsa narrative that powerful? Guys, it's been purely his team for 10 games now and they're playing like an 11 seed. Willing to be wrong here but a massive value play to my eyes.
(3) Gonzaga (-20.5) over (14) Kennesaw State. This is arguably the worst 14 seed against arguably the best 3. Kennesaw does a really good job of attacking the offensive glass and turning every game into a free throw contest, but I'm not sure how or where the 203rd-best post defense in America plans on stopping Graham Ike consistently. This also looks like a Gonzaga bloodbath defensively.
(7) Miami FL (-1.5) over (10) Missouri. This is being played in St. Louis, which has made Mizzou a bit of a popular pick, but I may push back here. Miami will have a major edge on the glass and in the turnover department, and Missouri is one of the worst teams in the field in terms of defensive shot suppression. When your opponent is +10.7 on the glass per 100 possessions...not ideal. No real pick value in either direction, though, and it's just a 1.5-point spread. This is one where if you live in Florida I'd tell you to take Missouri and if you live in Missouri you should take Miami.
(2) Purdue (-25.5) over (15) Queens (NC). Not expecting much from the 15 line this year, but Queens does have a top-80 offense and should be able to score on Purdue, particularly in P&R, where Purdue's coverage all year long has been leaky. Just don't expect the 345th-best half-court defense in the sport to stop...well, Purdue. If you told me the final score of this game is Purdue 100-76, that would sound right.
Midwest Region
(1) Michigan over (16) First Four. Nothing appealing here, as Michigan will be favored by 30+ against either opponent. UMBC would be a slightly more interesting matchup for obvious reasons, but it would feel a lot like when Michigan bizarrely scheduled Appalachian State in 2014 just to prove a point and beat them by 30+.
(8) Georgia over (9) Saint Louis. I guess this comes down to how real or fake you think SLU's late-season slide was. Since February 1, the Billikens are the nation's 108th-best team, and this is despite being +8% in eFG%. They turn it over too often, and against better competition, they rate out a negative in combined turnover/rebound margin. Georgia does too, but to be honest, Georgia's results and efficiency have been better than SLU's with Quad 4 games removed. I don't like this either, and I don't want it. Maybe this is cope to where if I pick UGA and SLU wins I'll be happy no matter what, because if I pick SLU and they lose I'm two-way miserable.
(5) Texas Tech (-7.5) over (12) Akron. This is the 12-over-5 of choice, it seems. Texas Tech is only being picked by 73% of bracketeers, the lowest by a full 7% of the 5 seeds and hilariously lower than a BYU team they are demonstrably better than. Now, Texas Tech also suffered a devastating season-ending injury, this one to JT Toppin, arguably a top-five player in the sport. They were a legitimate Final Four contender before he went down. After he went down...I mean, it's horrible. They've played like the 14th-best team in the nation. Before he went down, they were playing like the 9th-best team in the nation. Unreal.
I guess I'd ask for people to take a step back, look beyond the wins and losses, and look at who Texas Tech has actually played. I mean, sure, I would've liked for them to not get smoked by Iowa State, but that was two weeks after beating Iowa State on the road. Even from a shot volume perspective, the schedule-adjusted stuff I have calls it a wash, which means you are betting on Akron to out-shoot Texas Tech. I guess I can't get there, especially with Akron's results against the four top-75 offenses they played (1.35 PPP allowed!), but I'll admire the guts of those who can.
(4) Alabama (-12.5) over (13) Hofstra. The more I look at this matchup, the more I think we get a 13-over-4 or, God willing, a 14-over-3 somewhere. I just don't have confidence in actually picking one. That being said, it's weird how instantly Hofstra was written off on Selection Sunday once this was announced. Alabama cannot rebound, doesn't force turnovers, has easily the worst defense of anyone on the top six seed lines, and went 1-3 against the four top-25 2PT% defenses they played. You want me to call a 13 seed that ranks top-50 in OREB%, rarely allows steals, is the only 13 seed with a top-100 offense, and is #3 in America in 2PT% defense dead on arrival?
As an actual bracket pick, the people trying to win money should obviously pick Alabama. (I'll get to where I think you could have a minor edge in a later round against them.) My thought is this: 1-4 seeds with a negative SVI win nearly 0.3 games fewer than expected. We have two candidates for 2026: Kansas and Alabama. If I had to guess what 13-16 pops up and pulls a shocker, I would place my best wager on one of those two being involved. I just wouldn't pick it myself.
(6) Tennessee (-12.5 or -5.5) over (11) Miami (OH) or SMU. Tennessee projects as an extremely safe type of 5 or 6 seed: the Bully Ball team. Tennessee's Simple Shot Volume offensively of 127.9 is a record-setter for this seed line, and the Vols join an extremely rare collection of 5/6 seeds that got back 40%+ or more of their misses. The 10 who went before them: 8-2 in the Round of 64, with only 2006 Oklahoma and 2016 Baylor eating it. Tennessee's average win expectancy of 78% in this round is nearly 10% more than any of those who went before them. Even the SMU game profiles as Tennessee being 70% to win, which puts them closer to teams like 2005 Villanova (69% to win, made Sweet Sixtten) or 2015 West Virginia (66% to win, Sweet Sixteen) than a 2006 Oklahoma.
(3) Virginia (-18.5) over (14) Wright State. For upset purposes - nothing against UVA, never had an issue with it - I really wanted Virginia to draw North Dakota State. Didn't happen. The path towards Wright State being live here feels hard to envision. 14 seeds ranked outside of the KenPom top 125 are 2-17 all-time, and this Wright State team actually looks similar to the 2018 Wright State team that got smoked by Tennessee.
(10) Santa Clara (+4.5) over (7) Kentucky. It's incomprehensible to me that anyone still believes in 2026 Kentucky, but here we are. An astonishing 66% of people are taking the Wildcats to the Round of 32. Against top-100 competition, do you know what the power rankings for these two teams are? Kentucky, the 'favorite', is 43rd. Santa Clara: 24th. They project to have a significant shot volume edge and they're playing a team ranked 355th in the Away From Home metric at Haslametrics. It's the easiest value play in any 8/9 or 7/10 game available.
(2) Iowa State (-24.5) over (15) Tennessee State. Boring, but this is a weak 15 seed whose top 10 comparisons at Torvik went 0-10. None of them stayed within 15 points of their 2 seed.
South Region
(1) Florida over (16) First Four. Not really worth investigating.
(9) Iowa (-2.5) over (8) Clemson. Back to what I mentioned earlier, for 8/9 games I take whoever's favored, barring a really obvious shot volume mismatch or an injury. Here you actually have all three: a favorite who projects to have a significant edge in the shot volume department playing against a Clemson team down their highest-usage starter in Carter Welling. There's no edge in picking Iowa, though, as 57% of ESPN users have taken them.
(5) Vanderbilt over (12) McNeese. The Vanderbilt love is almost getting too severe. They're the most popular 5-seed pick to make the Sweet Sixteen, after all, and McNeese is a fairly stout 12-seed. But I'd feel a lot better if two items had occurred here. One: McNeese had drawn a 5-seed with worse guards. The #1 defensive TO% in America went 3-3 against teams in the top half of offensive TO% and 23-2 against every other Division I team they played. Two: their results against the few top-100 teams they played had gone better. McNeese grades out 106th against top-100 competition, and even if you bump it back to top-150 to give them more room, they still come in at 72nd. That's the best of the 12 seeds, but this has more Gonzaga/McNeese than Clemson/McNeese vibes to me. I'll note here that seeing the Bracket Breakers guys have this as their top upset makes me very queasy and I could be really, really wrong.
(4) Nebraska (-13.5) over (13) Troy. Not falling for it with Troy. I don't buy this Nebraska team for a deep run, as over their last 12 games they're not only 6-6, they're the 29th-best team in America. I'd argue this is Bad when your likely Round of 32 opponent rates higher. But it's also not great when your Round of 64 opponent, Troy, rates 213th in that same span despite holding opponents to 27% from 3.
(11) VCU over (6) North Carolina. Another auto-play here. Since the Caleb Wilson injury, UNC has played at the level of the 39th-best team in the nation, with neither their offense nor defense ranking in the top 40 nationally. VCU during that same span of time: 29th. There's already a lot to not like about UNC: they don't force turnovers, they do a very bad job of suppressing shot volume defensively, and against Quad 1 + 2 opponents, they were one of just five top-25 NET teams (Alabama, Texas Tech, Kansas, Wisconsin) to finish with a negative combined rebound and turnover margin. VCU is very strong, and against top-100 teams only, this is #37 VCU playing #34 UNC...and that latter number was with Caleb Wilson playing. I'll go down with the VCU ship once more.
(3) Illinois (-24.5) over (14) Penn. Penn is an exceptionally exciting underdog as a team with excellent shooting, a quality turnover margin, and an ability to play slow. I wish they could've drawn a more vulnerable 3-seed. Illinois is probably their worst possible draw: a team that never turns it over, will wreck them on the boards, and never, ever fouls. The Illini can shoot themselves out of any game, but they could also shoot themselves into a 33-point win or something.
(7) Saint Mary's (-3.5) over (10) Texas A&M. I wondered if this would be the rare popular 10-over-7 pick, which never happens, but SMC is only being picked by 51% of bracket users as I type, which makes them unusually valuable for a 7. Now, I do fear that this will be a functional draw in shot volume and that TAMU is simply too long and athletic for SMC, but they have a giant projected edge on the boards. Worth noting that neither team was great against top-100 competition (SMC 43rd, TAMU 48th). I wouldn't pick either past Houston anyway, so this is worth all of one point in your average pool.
(2) Houston (-23.5) over (15) Idaho. Idaho is a fun team that, again, needed a better draw.
NOTE: From here on, there's nothing in the way of real matchup breakdowns. They're all hypothetical anyway. Everything from now to the end is about Bracket Value, i.e. shots here and there that can help you win crowded pools...or at least find a few extra points where others will step in it. Hopefully. No spreads included from here, either, sorry.
Round of 32
East Region
(1) Duke over (8) Ohio State. I'm going to note now before you get your hopes up: I'm not taking any 8/9s over a 1. It's not worth the potential lost value in this bracket, and the 1s sit at 51% to all make the Sweet Sixteen. Even last year, when we thought we had the greatest 1 seeds ever, those odds were just 43%. My guess is that one of the 1 seeds has a real issue and trails with 10 minutes to go, but that's it.
(4) Kansas over (12) Northern Iowa. I don't enjoy this either, but I would simply note this item to you: St. John's (51% to S16) is actually being picked more by your average poolmate than Kansas (43%). That's despite Kansas having superior odds (46% to 44%) to get there. Those in giant pools (like, 900+ people) should consider UNI here, as only 2% of people have them to the second weekend despite having an 8% chance. I'm not sure this Kansas team is even good, mind you, but someone is legally required to win this four-team pod.
(6) Louisville over (3) Michigan State. Guys, coaches we think are 'bad' beat coaches we think are unbeatable all the time. Bill Self lost to Bobby Hurley THIS YEAR. Tom Izzo has lost in March to, in no order, Jerome Tang, Late Stage Jim Boeheim, Kermit Davis, and Trent Johnson. Losing to Pat Kelsey wouldn't be nearly that stupid.
In actuality, this is mostly a bet on a pair of items: that Mikel Brown Jr. will be mostly healthy, and that Michigan State's vote share for the Sweet Sixteen (72%!) is way too far beyond their actual odds of 58%. Also, Louisville's been the better team over the last 10 games: 20th versus 24th. Also also: with Mikel Brown available, per Torvik, this is a 17th-place 6 seed playing 12th. Not that outlandish.
(7) UCLA over (2) UConn. Some stats for you. The 2 seeds this year, compared to the 1 seeds, are at least vulnerable. They've got a 28% shot of getting to the Sweet Sixteen intact. Three of them - Iowa State, Houston, Purdue - are among the top nine 2 seeds ever in terms of odds to escape the Round of 32. Of the six who've gone before them, five at least made the Elite Eight. (The sixth was 2016 Michigan State. Oops.)
Weirdly, two of these (Iowa State and Purdue, possibly because of previous bracket trauma) actually represent plus value in the average bracket. The third, Houston, does not, but they're the highest-rated 2 seed on the board. They do draw the strongest 7 and second-strongest 10, but Houston's astonishing ability to generate extra possessions (+11.8 SVI) combined with being a top-10 team puts them in rarefied air for March purposes. There are 19 1 or 2 seeds since 2002 who've combined a +10 SVI with a top-10 ranking, and they are an unbelievable 18-for-19 in making the Sweet Sixteen. Your test cases this year: the Coogs and the Cyclones.
That leaves UConn as this year's most likely candidate to lose on the top two seed lines. Do I love this pick? Not really. But despite being easily the least-great of the 2 seeds, UConn has the second-most picks to the Sweet Sixteen and Elite Eight of the 2 line. You can generate major value by picking against them starting here, or if you feel less excitable, the Sweet Sixteen when the average human has them in their Elite Eight.
West Region
(1) Arizona over (9) Utah State. Unfortunately, every 1 vs. 8/9 game this year falls in the category of putting the 1 seed straight to the second weekend without much thought. Neither possible opponent here interests me much.
(4) Arkansas over (5) Wisconsin. Neither pleasant nor likable a pick, but again, you are somehow currently getting plus value because only 53% of people have Arkansas in their Sweet Sixteen versus 58% odds. Take it and run with it, particularly if my Wisconsin Theory comes true and they actually lose to High Point.
(3) Gonzaga over (11) NC State/Texas. There will be a double-digit seed that makes this year's Sweet Sixteen, and if forced to pick, I believe it will be NC State or Texas, whoever wins that game. They've got the best odds in my book to get it done, both analytically and when facing top-100 competition. (In fact, the only team that outpaces the First Four squads here against top-100 teams is Santa Clara, but they play an Iowa State team I'll be covering shortly as a strong bet for the second weekend. The lone team outpacing these two over the last 10: South Florida.) My advice: if you ARE in a bigger pool, take NCSU/Texas. If you're not, Gonzaga it. I think NC State would have a better shot here, but Texas may match up better with BYU in the Round of 64.
(2) Purdue over (7) Miami FL. Covered earlier, but Purdue and Iowa State are two very rare things: 2 seeds with plus value in the Round of 32. You almost never get plus value on a 1 or 2 seed until the second weekend at earliest, but it's a combination of how large the gap from 2 to 7 is/how many people see the name Purdue and believe they'll freak it. That being said...I do actually think Miami would match up quite well here if it happens.
Midwest Region
(1) Michigan over (8) Georgia. All of my complaints about 1 vs. 8/9 games aside, the Michigan/Saint Louis potential matchup I think I would be most interested in. That or a Duke/Ohio State game. This would probably be Michigan eventually motoring away to win by 12-18.
(5) Texas Tech over (4) Alabama. I'm going to run back what I mentioned earlier. Did you know that, since JT Toppin's injury, Texas Tech has still played like a top-15 team? Did you know that, over their last 10 games, Alabama has played at the level of the 23rd-best team in the sport with the 84th-best defense? Yet nearly twice as many people have Alabama in their Sweet Sixteens. I'll live with the consequences if I get it wrong.
(6) Tennessee over (3) Virginia. Not only is Tennessee rated equal or better almost everywhere - a rare situation where the 6 seed might actually be favored - this is a doubly-rare situation where the 6 seed has extremely positive bracket value compared to the 3. I am riding with the mercurial Vols this March.
(2) Iowa State over (10) Santa Clara. God, how I wish Santa Clara was playing UConn or even Purdue instead. I would've liked that matchup a whole lot more than this. I'll run my Houston stat again here: there are 19 1 or 2 seeds since 2002 who've combined a +10 SVI with a top-10 ranking, and they are an unbelievable 18-for-19 in making the Sweet Sixteen. Your test cases this year: the Coogs and the Cyclones.
South Region
(1) Florida over (9) Iowa. Actually, this may be the 1 vs. 8/9 I'd like to see most. If you don't think Iowa would be live here, you haven't seen Iowa play and completely manipulate time and space during a game. Possessions feel like they last for four hours. It would throw this imperfect Florida team for a loop, and Iowa can absolutely exploit Florida's ballhandling struggles. I imagine UF would still win by 11 or something, but this would be interesting for a while.
(5) Vanderbilt over (4) Nebraska. Auto-play. Vanderbilt is the better team, and while Vandy is getting too much love there isn't much I love about Nebraska at this point either.
(3) Illinois over (11) VCU. Guys, I'm sorry: this is going to be a really boring "how stats and history" post. It's just not a setup for many upsets. I think what TeamRankings' Predictor Tool says is more or less right: Illinois would have an 80% or greater chance to win this game in basically any setup or scenario.
(2) Houston over (7) Saint Mary's. Neither setup gets Houston anywhere close to a situation where I'd ponder picking against them. They still represent good value to anyone's bracket at this stage.
Sweet Sixteen
(1) Duke over (4) Kansas. Again, I'm not going to pick against the obvious best team until I am given good reason to do so, whether in the bracket or purely analytically. This isn't it for me, and again, I'm not convinced Duke isn't facing Northern Iowa or someone here.
(6) Louisville over (7) UCLA. You're taking a monster risk here, certainly, but if it works you're going to be sailing smoothly. Considering there's not a team in this region I am taking that isn't named Duke, this is probably where you can gain a few extra points, given UConn's wobbliness and given that Louisville is roughly a top-15 team in the nation at full strength. (Also, this STILL wouldn't be a top-25 opponent!)
(1) Arizona over (4) Arkansas. Statistically, at least three of the four 1 seeds should make the Elite Eight. There's 60% odds of at least three getting there. I think you're aware of which one I'll be betting against.
(2) Purdue over (3) Gonzaga. This isn't a pick I'm proud of or even one that I particularly like. I laid out earlier how and where I think your most likely double-digit seed to make the Sweet Sixteen is whoever wins the Texas/NC State game, and they would be beating Purdue (or Miami, or Missouri) in this game instead. But purely as a bracket play, Gonzaga represents serious value in that they might be straight-up better than Purdue while only being picked by 30% of people to make the Elite Eight. I'd still take Purdue, though. Maybe it's stupid.
(1) Michigan over (5) Texas Tech. This is the game where I think Texas Tech would miss JT Toppin the very most. If he were playing here, they would have a great shot at an awesome upset. Unfortunately, he won't be playing, and I'm riding with Michigan until I have a great reason not to.
(2) Iowa State over (6) Tennessee. The things that make Tennessee match up really well with either 11 seed or with Virginia don't make them match up well with Iowa State, who can force Tennessee into a ton of turnovers and could neutralize their rebounding advantage. More notable for you is that Iowa State somehow still has plus bracket value here: 50% odds to make the Elite Eight, with just 47% of users picking them to actually do it.
(5) Vanderbilt over (1) Florida. I mean...I just saw it? Vanderbilt is an extremely risky pick, but there are two reasons I'm not interested in taking Florida far. For one, they're getting the Defending Champion Bump. 75%! of users at Yahoo have them in the Elite Eight. Their actual odds to get there: 53%. It is only slightly more likely Florida makes the Elite Eight than doesn't, and with that much bracket value available, this is where I'd take the next-best team from the upper half of their region to take advantage. Also defensible here: Nebraska, but you have to stomach just how bad they've been offensively for five weeks. Vandy is extremely tricky to evaluate, by the way: they certainly could lose in the Round of 64...but if they get past it, they have two very favorable matchups to make the program's first Elite Eight since 1965.
(2) Houston over (3) Illinois. Illinois is likely the superior value play in your standard bracket, but most analytical systems aren't going to factor in that Houston is basically getting to play a home game in the Sweet Sixteen. This may or may not actually matter, but the other aspect of the bet is taking into account the Iowa State note from earlier. Those same 19 teams are 13-for-19 in making the Elite Eight. Also, 1-4 seeds with -1 or worse turnover margins per 100 win 0.3 games less than expected, per my database. This year's test cases: Michigan, Florida, Illinois, Michigan State, and Kansas, who would expect to combine for 12.6 wins but (in theory) should only win 11.
Elite Eight
(1) Duke over (6) Louisville. A lot of smarties out there are jumping off the Duke bandwagon due to the Foster/Ngongba injuries, and perhaps they are right. However, Ngongba seems fairly likely to play no later than the Sweet Sixteen, and Foster isn't actually confirmed to be out for the season. It's boring, but no one in the East Region represents enough in the way of plus value to make passing on Duke anything beyond a negative.
(2) Purdue over (1) Arizona. This will require serious explanation. I wouldn't mind picking Arizona to win it all, and they are one of the four or five truly defensible picks to my eye unless we do live in Hell and Florida goes back-to-back. (Nothing against Florida. I just like unique champions.) (Second parentheses: the fourth and fifth defensible picks to my eye are Houston or Purdue. They were better than Florida this year against top-100 competition.)
The reason why I'd pick Purdue is NOT because I think Purdue is better or is even more likely to actually get this far. It's more that Purdue is your best chance at differentiating from the crowd. As I type, 57% of people have Arizona in their Final Four. Just 16% of people have Purdue. Against top-100 teams, Arizona was third-best nationally...and Purdue fourth. That's a pretty huge percentage gap for two teams that were near-equal against the best of the best. Filtered exclusively for games against top-100 teams, Arizona has 45% odds to get to Indianapolis. Purdue: 30%. It's worth a swing to me.
(1) Michigan over (2) Iowa State. This is not. Michigan, in those same odds, is a true 50/50, while ISU is 22% to get to Indy. There's some positive value in taking the 'Clones, but not enough to make it worth it unless, again, you're in a pool of 500+ people. Or you live in Michigan. In which case, stop reading this and go to a Great Lake for me.
(2) Houston over (5) Vanderbilt. 34% of bracketeers have Houston in their Final Four, which makes Vanderbilt the better value play, but in those top-100 odds, Houston has a 30% chance to get to Indianapolis, much like Purdue, and has the best odds to see Indy amongst their region. Again, that doesn't factor in playing two functional home games, either. Think of this as more of a bet against Florida - 42% of people have them in their Final Four, and even in pure KenPom odds, they're only 29% to make it.
Final Four
(1) Duke over (2) Houston. If you're feeling frisky, Houston is the slightly superior value pick, but I've been on the 1 vs. 1 train in the national final for two months. I don't really see a reason to stop here.
(1) Michigan over (2) Purdue. This, on the other hand, represents a very rare item: plus value for a 1 seed to make the national title game. Michigan's odds to get there are bang-on 30% at Torvik, and only 28% of people in the nation have them playing on Monday night this year.
Title Game
(1) Michigan over (1) Duke. Well, can't undersell it here. Michigan is this year's Bracket Value Team of the Year. The pick to win it of just 14% of brackets, they have 19% odds to get it done at KenPom, 18% at Torvik, and 27% (!!!) at EvanMiya. No team in the entire field of 68 promises more long-term value, even with the LJ Cason injury, than the Wolverines do.
Your disclaimer here: the last two years, this process has managed to pick the eventual runner-up as your champion. But! It's also picked at least two Final Four teams three years running. Life can and has been worse than that. Happy bracketing.
