No 16-seed had ever beaten a 1-seed in the men’s NCAA tournament until 2018. Now it’s happened twice.
In the first 28 tournaments that featured 15-seeds (1985-2012), four 15s had beaten 2-seeds. Seven have done so in the past 11. No 15-seed had reached the Sweet 16 until Florida Gulf Coast did it in 2013; now it has happened for three straight tournaments, and Princeton is one win away from putting a No. 15 in the Elite Eight for the second straight year.
The Sweet 16 will not feature any of Kansas, Duke, North Carolina or Kentucky for just the fourth time in 70 years … and the second time in three. We’re openly pondering a Final Four with no 1-seeds, and the Elite Eight could feature up to four programs from what we typically call mid-major conferences.
Parity is as strong as ever in college basketball. Players, coaches and fans from Kennesaw State to Furman to, of course, Fairleigh Dickinson, all got their moments in the spotlight on Thursday and Friday, and Princeton and Florida Atlantic will get another go-round this coming weekend. It made for one of the NCAA tournament’s more delightful opening weekends in recent memory (even without all that many close second-round games).
We can debate how we’ve gotten here — it probably has more to do with the diminishing quality of the top teams than the bottom teams — and we can debate whether this is actually good for the sport. TV ratings tend to benefit from the presence of at least a few blue bloods, after all, and lord knows sports like European soccer haven’t suffered all that much from a lack of input from upstarts. But the NFL also hasn’t suffered because of the big-brand Dallas Cowboys’ annual failures (they haven’t reached a Super Bowl in 27 years).
At its best, parity convinces a huge number of fan bases that their team can win it all — or at least make a huge run — and hopeful fans are engaged fans. ESPN’s TV ratings this season were reportedly as high as they’ve been since before the pandemic, and Thursday’s first-round TV viewership was the highest in eight years. At worst, the success of mid-majors in the NCAA tournament hasn’t hurt anything. At best, it has helped significantly.
So why aren’t more mid-majors getting tourney bids?
Last year I ranked all of the NCAA tournaments going back to 1979, concocting a formula that took upsets, close games and generally memorable moments into account.
In the top eight tournaments from that list — 1985, 1990, 1983, 2006, 2010, 2014, 1987 and 1997 — 75% of possible at-large bids went to teams from what we could call power conferences: the ACC, Big East, Big Ten, Big 12 (or, for the 1980s tournaments, the Big 12 and SWC), Pac-10/Pac-12 and SEC. Outside of that group, an average of 4.9 other conferences got multiple teams into the field, 5.3 if we include 1980s independents as their own entity. Even the more recent tournaments on that list were diverse and outstanding. The 2014 tournament, for instance, featured 10 total multibid conferences. The Atlantic 10 got six teams in that year; the A10’s Dayton reached the Elite Eight, and the AAC’s UConn — technically a mid-major at that specific moment in time — won it all.
Over the past seven tournaments, the percentage of at-large bids going to power conferences has increased to 85%. On average, only 3.7 other conferences, besides the power six, have gotten multiple bids. This year, only three did: the Mountain West, AAC and West Coast Conference, a.k.a. the most renowned of the mid-major conferences. Of the last four at-larges in the 2023 field and the first four out, seven were from power conferences.
You could make the case that the last couple of NCAA tournaments have been excellent despite themselves. The selection committee gave 13 tourney bids to Big Ten and ACC teams this year, and only two made the Sweet 16; the committee gave a total of six to the WCC, AAC, C-USA and Ivy, and four made it to the second weekend.
Most #MarchMadness bids by conference:
8 – Big Ten
8 – SEC
7 – Big 12
5 – ACC
5 – Big East
4 – Pac-12
4 – Mountain West
2 – WCC
2 – AAC pic.twitter.com/755EcxaeIN
— NCAA March Madness (@MarchMadnessMBB) March 12, 2023
Now, it must be acknowledged that part of the reason for an increase in power conference tourney presence is the fact that power conferences themselves are bigger. Membership in these conferences has increased by nearly 50% since the tournament expanded to 64 teams in 1985.
Just in the past 10-12 years, the Pac-12 roped in 1998 finalist Utah, the Big 12 incorporated TCU and the Big East dissolved and then reformed to include 2010-11 finalist Butler, plus Creighton, DePaul and Marquette. And we’re about to go even further down this road: The Big 12 is adding Houston (a No. 1 seed in this year’s tournament), Cincinnati, BYU and UCF, and it’s possible that the Pac-12 could soon expand to include San Diego State and perhaps SMU or UNLV. “Gonzaga to [Big 12/Pac-12]?” rumors have picked up steam again recently, too. And as second-tier conferences have lost members to larger conferences, they’ve in turn plucked members from further down the food chain.
This consolidation has had an obvious effect; it’s not as if the selection committee itself has had evil, elitist intentions in leaning toward the power conferences. Still, those conferences continue to get the benefit of the doubt even if or when they don’t earn it in the tournament.
Are we using the wrong selection criteria?
Per the NCAA, Purdue conqueror Fairleigh Dickinson came into the tournament with the lowest-rated strength of schedule in the country. The only BPI top-100 team the Knights had played was Pitt, and the Panthers had beaten them by 22. A total lack of schedule strength in no way prevented them from beating the East’s top-seeded Boilermakers, then acquitting themselves well before falling to 9-seed FAU in the second round.
Who says you need to be battle-tested to win in March?
— NCAA March Madness (@MarchMadnessMBB) March 18, 2023
Is there something to that and the fact that, in the past four tournaments, 14 of the 17 top-four seeds that lost to double-digit seeds had finished the season ranked higher in Strength of Record (a résumé rating) than in BPI (a quality rating)? Might we be overvaluing schedule strength and at least slightly undervaluing actual quality when it comes to selecting or seeding tourney teams? Granted, no criteria would have made FDU look particularly good — after all, the Knights didn’t even win their conference tournament — but it might be fair to wonder if some of the high seeds that have lost recently shouldn’t have been seeded as high.
Then again, what’s the alternative? Using tools that aren’t opponent-adjusted? Rewarding teams for good records, context-free, and indirectly discouraging strong nonconference scheduling? That partially solves one problem and creates quite a few more.
Really, the selection criteria are probably about as good as they’re going to be. If we want to create a fairer field, we should perhaps follow a different path: using tourney success to determine future tourney bids.
Bring on the coefficients!
There’s no law saying we have to give out exactly one automatic bid per conference. In European soccer and plenty of international sports, the number of bids into a competition is derived at least in part from how you’ve performed in the past. The UEFA coefficient, a score derived from competitive success over a five-year period, determines the number of clubs different countries get into soccer’s Champions League, and if it works for the Premier League and LaLiga, why couldn’t it work for the ACC and Big 12?
We could use the idea of a coefficient in different ways.
Based on the scoring system of choice, we could use a conference’s five-year performance to hand out multiple automatic bids to deserving conferences.
For an example, I created a points system that awards one point for a first-round win, two for a second-round win, etc., up to six for a national title game win. (We won’t worry about First Four wins.) So if you win the title, you earn a total of 21 points for your conference (1+2+3+4+5+6), and if you win your first-round game and get eliminated, you earn one.
Looking at the past five NCAA tournaments (2017-19, 2021-22), you would end up with these per-tournament averages using current conference membership*.
* What I mean by “current conference membership” is, when Loyola Chicago moved from the Missouri Valley to the Atlantic 10 last year, it would have taken its coefficient points with it.
Who’s to say we couldn’t award a second automatic bid for any conference with an average over 1.0? That would mean that the brilliant run of Saint Peter’s last season would have earned the MAAC an extra slot in this year’s field. (Since Iona won both the regular-season title and conference tournament, that would mean the second bid could either go to second-place Rider or, more likely, tournament runner-up Marist.) The Atlantic 10’s recent success would have cemented a spot in the field for conference and tournament runner-up Dayton.
For any conference with an average over 5.0, we could assure a third bid as well, which wouldn’t make much of a difference for the Big 12s and ACCs of the world (they typically get way more than three in) but would in this case mean that the West Coast would get another team in addition to Gonzaga and St. Mary’s — third-place Santa Clara, perhaps? Or the highest-seeded WCC tournament semifinalist (BYU)?
What if we used the same idea, but applied it to clusters of conferences instead? We could divide the conferences into four groups of eight using a quality measure of choice. For convenience, I’ll use five-year BPI averages, but the NCAA could obviously use NET rankings or whatever else. We look at the average success of each quadrant and determine bids from there.
Quadrant 4 (Big Sky, Big South, MEAC, NEC, OVC, Patriot, Southland, SWAC): 0.0 tournament points on average.
These conferences receive zero extra bids.
Quadrant 3 (America East, ASUN, Big West, C-USA, CAA, Horizon, Ivy, Sun Belt): 1.0 average tournament points.
These eight conferences get one additional bid, given to the highest-rated non-automatic qualifier (likely either North Texas, Liberty or UAB).
Quadrant 2 (A10, MAAC, MAC, MVC, MWC, Southern, Summit, WAC): 7.0 average tournament points.
These conferences are granted two or three extra bids, which they actually received in this year’s tournament — the MWC got four teams into the field.
Quadrant 1 (AAC, ACC, Big 12, Big East, Big Ten, Pac-12, SEC, WCC): 87.8 average tournament points.
Give them as many extra bids as you want — they already receive at least that many.
Both of these coefficient approaches reward conferences in some way for their actual tournament performance, and merit is never a bad thing. But they also don’t actually change the tournament field much. The conference approach exchanges, say, Pitt, Nevada (a mid-major!) and Arizona State for Marist, Dayton and BYU. The quadrant approach trades Pitt or Nevada for North Texas. That certainly impacts fans of those teams but creates only a small ripple otherwise. That’s not necessarily a bad thing. But we could get much bigger and weirder.
BASE THE ENTIRE FIELD ON LAST YEAR’S TOURNAMENT
What if we gave a second automatic bid to any conference that won at least one round-of-64 game last year? Because of wins by Saint Peter’s, Murray State, New Mexico State and Richmond, respectively, that would mean giving extra 2023 bids to the MAAC (likely Marist), Ohio Valley (regular-season champion Morehead State ), WAC (regular-season champion Utah Valley) and Atlantic 10 (Dayton). That’s a more significant change, but again, it’s based totally on merit!
Hell, what if we came up with a completely different way of assigning bids? Let’s keep all 32 automatic bids for conference champions, then divvy out 32 more bids based on the makeup of last year’s round of 32. (To add up to 68, we can send out four bonus at-large bids or get NIT success involved.) So the Big 12 and Big Ten head into 2022-23 knowing they’re probably getting seven bids (the automatic bid + six from last year’s round of 32), and a conference like the Mountain West, which went a dismal 0-4 last season, has to build momentum from scratch with potentially only one bid.
OK, that last one is probably too far, but any of these are technically doable and merit-based. But there’s one other option as well. It’s probably more likely.
We could finally just expand the tournament again
I’ll pause while you scream, “NO. NEVER. STOP. BAD.”
All right, are we good now?
The perfection of the 64-team bracket — and, if we’re being honest, its symmetrically brilliant fit on a single sheet of paper — has kept us at or near 64 teams for nearly four decades now. In 2010, when the rumor emerged that the tournament was expanding to 96 teams, the negative reaction was nuclear-grade. Last summer, SEC commissioner Greg Sankey brought up the idea again following Texas A&M‘s narrow miss, and ACC commissioner Jim Phillips followed suit. The dismissal from fans and commentators was again swift. We love our tournament, and our brackets, exactly the way they are. We love the lost work productivity of the first Thursday and Friday of the tournament. It is perfect and has been especially perfect the past couple of years. The tournament did expand in 2010, but only to an awkward 68.
— NCAA March Madness (@MarchMadnessMBB) March 12, 2023
For all the eye-rolling we give to the concept of bracket creep, however, and for how quickly we dismiss the (correct) idea that expansion would be a massive, tradition-bucking cash grab for the NCAA and those to whom the NCAA writes March Madness checks, it’s actually rather wild to think about how little the bracket has crept. In the nearly 40 years since the first 64-team tournament in 1985 — which came into existence after an extended fight about how much of a tradition-bucking, regular-season-diluting cash grab it would be — we’ve added only four teams to the field, a 6% increase. The number of teams in Division I has increased by 29% in that time, the U.S. population by 41%, but we have demanded our perfect bracket remain in place.
The way the college sports world tends to work, though, the only way we’ll get more mid-majors into the field is by adding more power-conference teams, too.
ESPN’s John Gasaway has already highlighted how and why tournament expansion wouldn’t help only the big guys. To be sure, it would help the big guys: a field of 80 or 96 teams this year, for instance, would have almost certainly included 18-15 Oklahoma State, 17-14 Wisconsin and perhaps 17-16 Colorado or 17-16 Washington State. But it would have also opened the door for 27-7 North Texas, or 26-7 Sam Houston, or 27-8 Liberty, or 26-9 UAB, or 22-12 Dayton, any of whom were far better than FDU in the regular season and were more than capable of winning a couple of tournament games. If said expansion resulted in more automatic bids for conference champions — or at least champions of conferences with a certain level of tournament success, à la the ideas above — that’s even better.
We could expand the tournament all the way to 96 teams without fundamentally altering whatever it is we hold most dear about the current setup. Even at 96, the logistical change wouldn’t have to be significant: Instead of eight venues hosting eight teams playing six games over a Thursday-Saturday or Friday-Sunday, we easily make that 12 teams in each venue, with eight playing games on Tuesday or Wednesday to qualify for the Thursday or Friday games.
We could redraw things from scratch, seeding four regions from 1-24, giving the top eight seeds byes and ending up with first-round pairings like 9 vs. 24, 10 vs. 23, etc. That would potentially create more difficult matchups for top seeds — No. 1 Houston might play, say, 16-seed Liberty or 17-seed Michigan in the round of 64 — and instead of beating Purdue, a miracle run by a team like FDU might first go through a 9-seed in the first round, then an 8-seed in the second. If that’s simply too much change for our delicate sensibilities, then we could instead go wild with play-in games: The top 8 seeds in each region still get byes, while eight teams play for the right to earn the four 9-seeds (FAU vs. Penn State, for instance), eight play for the No. 10s (Providence vs. Rutgers?), etc., all the way down to the 16-seeds.
Out of curiosity, I mocked out a 96-team field that, in essence, folded most of the NIT’s field (minus teams that finished at or under .500) into the NCAAs and added in a pair of NIT-eligible teams that begged out of the field (a disappointed North Carolina and a banged-up Dayton). Among these 96 teams were 11 from the Big Ten, nine from the SEC, eight from the Big 12, seven each from the ACC and Pac-12 and five each from the Big East and MWC. But this field also featured multiple bids for 22 conferences and offered spots to some of the best teams from the mid-major universe: Liberty, Sam Houston, Yale, Dayton, UAB, North Texas, Bradley, Hofstra, Toledo, Utah Valley and others.
That feels like a fair trade to me. Maybe it doesn’t to you. Regardless, there are big and small ways to reward the mid-major universe for its contributions to what remains one of the greatest parties on the sports calendar. Who knows what that group could do with a few more merit-based seats at the table?