The price we're paying for no upsets: the third-best Sweet Sixteen in history
Okay, "history" as in 2002 to now
I think we got spoiled. I’m using “we” collectively here to mean the college basketball viewership. Think about it! Since COVID hit, here’s a summary of every Tournament’s opening weekend since. I’m using the correct definition of ‘upset’, i.e., a seeding difference of 5 or more.
2021: A bonkers first two days with seven upsets (tied for the record), which was merely followed by a crazier Round of 32 that featured five additional upsets. The 12 opening weekend upsets remains the record for any NCAA Tournament from 1985 to present since the expansion to 64+ teams. At a total seed count of 94, this remains the lowest-quality (in terms of seeding) Sweet Sixteen in history.
2022: Six upsets in the Round of 64 highlighted by the biggest of them all: 15-seed Saint Peter’s beating 2-seed Kentucky. Obviously, Saint Peter’s would go on to make the Elite Eight, pulling off a shocker of shockers. The Sweet Sixteen had four teams seeded 10 or lower for the second straight year, which is nuts because that hadn’t happened in any tournament since 2011. Now it had happened twice in a row.
2023: A 16-seed beat a 1, for starters, but for the first time ever, a 15-seed beat a 2 in the same tournament. In fact, teams on all but one seed line (14) won at least one game. Half of the 1 & 2 seeds were gone by the time the Sweet Sixteen started, tied with 2018 for the fewest-ever 1 & 2s in the second weekend.
Heading into 2024, the expectation was set. You can make a reasonable argument that we, as humans, just watched three of the five craziest NCAA Tournaments in modern history in back-to-back-to-back years. Naturally, the expectation was that we would have a fourth. Analysts spent most of the season discussing top 10 teams’ win-loss records away from home. Every upset loss was an indicator of a crazy March. Every conference tournament windfall for a future 11/12/13 seeds was simply a reason to prepare yourself for a mad three weeks of basketball.
Indeed, for the Round of 64, we got a taste of madness. There were no 16-over-1s or 15-over-2s, but we got 14-seed Oakland beating 3-seed Kentucky, along with 13-seed Yale beating 4-seed Auburn. In total, the Round of 64 produced six upsets, which is about one more than the historical average for the first round. With the Round of 32 incoming, many anticipated we’d see more madness.
And then there was none. Not a single upset occurred in the Round of 32. Vegas favorites went 15-1, with the only ‘upset’ being 6-seed Clemson overcoming being a two-point underdog to 3-seed Baylor. Every attempted upset bid ended in a loss. The only double-digit seed in the Sweet Sixteen - actually, the only seed lower than 6 - is 11-seed NC State, who faced 14-seed Oakland in the Round of 32.
How’d we get here? The bad news: it was actually very easy to see coming. The good news: this is one of the best Sweet Sixteens ever produced.
This post is free, but everything else in the month of March(+early April) isn’t. Subscribe here for $18/year.
The seeding should’ve told you
I include “me” in the you section here. I had one double-digit seed in my Sweet Sixteen (New Mexico) but also had two of the 1 & 2 seeds exiting in the Round of 32: North Carolina and Marquette. Neither happened, and no huge upset happened at all. Again, we should’ve seen this coming. From a post I wrote the night after the field was revealed:
Now, let’s get into gap territory. And boy, it’s a bummer if you like upsets.
The gap between 1 and 16 seeds is +34.96 AdjEM on average, the third-highest since 2002 and the highest in any Tournament since 2019.
The gap between 2 and 15 seeds is +24.59 AdjEM on average, the fourth-highest since 2002 and the highest since 2021. (Which did have a 15-over-2!)
The gap between 3 and 14 seeds is +18.54 AdjEM on average, the third-highest since 2002 and only surpassed by 2022 (no upsets) and 2017 (also no upsets).
Lastly, the gap between 4 and 13 seeds is +15.2 AdjEM on average, the second-highest of all-time and only surpassed by 2017 (no upsets).
What this means is as follows: of your 16 protected seeds, you have an unusually high shot of seeing a lot of them make the Round of 32. In fact, if you were to average the efficiency margins of the 2024 13-16 seeds, they would be the worst in our 22-year database by over half a point. The only years in the same stratosphere were 2002, 2004, 2015, and 2017…which saw the 1-4 seeds go a combined 61-3 in the Round of 64. For those of us who adore the chaos this Tournament so often gives us, this doesn’t look like a good year for it.
Collectively, the 1-4 seeds went 14-2, and the only close games were the upsets. The area where I was wrong was thinking the Round of 32 might help us out. All of the gaps between high seeds and their counterparts (1 vs. 8/9, for example) had wider gaps than the historical average. In fact, the gap from 4 seeds to 5/12s was the third-largest on record. The gap from 1 seeds to 8/9 was the largest since 2019. The gap between 2s and 7/10s was right at the historical average and right in line with 2021 and 2022, but perhaps we got a bit unlucky there.
In all, the collective gaps from high seeds to lower-seed counterparts was the third-highest ever, only surpassed by 2019 and 2015. Not coincidentally, those years also had very calm paths to the Final Four. Half of the eight teams who made it were 1 seeds, and only 2015 Michigan State was seeded lower than a 5. (2019 Auburn technically was surprising, but it’s worth noting they entered that Tournament 13th in KenPom, a borderline top-10 team.
As such, we probably should’ve seen this coming. This was never going to be a 2021, 2022, or 2023; I hated seeing it, but the comparisons to 2007, 2008, 2015, 2017, and 2019 kept ringing true in the data. All of those were relatively calm Tournaments with very boring Round of 32s.
Now, for the good news: this is the third-strongest Sweet Sixteen since 2002, at least by pre-Tournament KenPom data. At an average AdjEM of +24.1, the teams who’ve made this Sweet Sixteen are only topped by 2015 (+24.75) and 2019 (+26.66). This is evidenced by data, for one, but the seeding is meaningful here too. From our pal David Worlock, the Director of Media Coordination/Stats for the NCAA:
Considering that 40 would be the lowest possible, we’re all of 13 total seed numbers off of a ‘perfect’ bracket here for chalk. At least since I’ve been basketball-conscious (coincidentally 2002-present), this is the fourth-chalkiest Sweet Sixteen I can remember. What this means, though, is a truly awesome Sweet Sixteen. You see that average AdjEM above, but check this out:
In terms of top-to-bottom gap, there’s just a 19.2-point gap between the top (UConn) and the bottom (NC State) teams left in the field. That’s the smallest gap since 2018, which saw four of the eight Sweet Sixteen games decided by four points or fewer.
Nine of the pre-Tournament KenPom top 10 (excluding Auburn) are left standing, the first time this has happened since 2019 (when all top-10 teams made it) and only the second time it’s ever happened.
There are two Sweet Sixteen matchups between top 10 pre-Tournament KenPom teams (Houston/Duke in the South, Iowa State/Illinois in the East), also the first time this has happened since 2019 and just the second since 2009 (North Carolina/Gonzaga, Memphis/Missouri).
The lowest FanMatch rating at KenPom for a game in this year’s Sweet Sixteen is 67 (NC State/Marquette). While an imperfect measurement, that’s the highest low in the Tournament since 2017.
Arguably, the third-best Sweet Sixteen in history may even be selling it short. This is one of the most power-packed Sweet Sixteens anyone may ever see, and it feels like a last gasp of a bygone era. The transfer portal helping talent become more spread out than ever in college basketball means that Sweet Sixteens like this may be few and far between going forward. You may not like chalk in the moment, but you’re more likely to appreciate this one down the line than you were the Elite Eight game between North Carolina and Saint Peter’s.