Upsetting The Two-Seed: Cole’s March Madness Cheat Sheet
Ryan Coleman breaks down why 10 of the last 20 two-seeds haven’t made it out of the first weekend of the NCAA Tournament and how you can pick the next big upset
With March Madness right around the corner, I’m sure millions are thinking, “this is the year I have a perfect bracket.” Of course, that won’t ever happen, but there’s no better feeling in the world than picking the right upsets and champion, ultimately winning your bracket pool.
In the days leading up to the greatest tournament in American sports, I will look at both the past and the future. With work, you will be able to predict the correct first round upsets, Cinderella teams, and decide the most important choice of all: who will hoist the championship trophy on April 8th in Minneapolis. To start, I took a look at the two seed, and what it has taken to upset these elite teams in recent years.
Upsetting the Two Seed
In each of the past five years, only half of the two seeds have made it past the first weekend. That’s ten of, supposedly, some of the nations best teams. So what does it take to pull off an upset? I’ll analyze each two seed upset before the Sweet 16 in the past five years, and what it took to bring these giants down.
All statistics per Kenpom.com
2018: UNC (2) lost to Texas A&M (7) and Cincinnati (2) lost to Nevada (7)
UNC (26–11): In 2018, North Carolina depended on their biggest strength: offensive rebounding. UNC dropped five of their six worst offensive rebounding (OR%) games, including to Texas A&M when they had lowest percentage at 17.3%. The Tar Heels shot 6-for-31 (19.4%) from three that game, their second lowest percentage on the season. U
Unsurprisingly, A&M ranked 40th in defending the three and 18th in defensive effective field goal percentage (eFG%). UNC let a lot of shots fall as A&M shot 21-for-36 (58%) from inside the arc. Looking holistically, seven of UNC’s losses came to teams in the top 50 of defensive eFG%. A&M ranked 18th for that stat, speaking to why North Carolina struggled from both outside and inside the arc, as they only shot 20-for-47 (42.6%) that game from two, their 6th lowest of the year. An interesting statistic to note is that UNC rarely forced turnovers that season, as they were 301st in defensive turnover percentage (TO%) at 16.3%.
Cincinnati (31–5): So this team was really good at…well, everything. They were 2nd in defensive efficiency, offensive rebounding percentage, and defensive 2P% while ranked 20th in defensive turnover percentage. Unfortunately, Nevada was ranked first in the nation at ball control with only a 13.5 TO% on the season.
In this upset, Nevada showed their elite ball control as they held their TO% to 3.1%, Cincy’s next lowest opponent TO% all year was 12.3%. The Bearcats lowest 14 3P% games held all their losses with only two wins over a top 80 team in that batch, showing a dependency on the three. Cincy had a bad draw as the Wolf Pack ranked 14th in the nation at defensive 3P%. Nevada was also money from two, shooting 24-for-43 (55.8%) which was the second highest FG% from two Cincy allowed all season (Nevada ranked 83rd in 2P%).
Side Note: You may be thinking “well, of course teams are dependent on the 3-point shot to win, you only beat the other team by scoring more points than them,” and yes, that makes perfect sense. But, if you look at team like Duke this year, they lost against Gonzaga while marking a 38.5 3P%. While this is not great, it‘s this Duke unit’s 8th best percentage from three in the games they have played so far. Looking at the other end of the spectrum, of their lowest three 3P% games, the Blue Devils had two wins over top five teams per Kenpom. So yes, of course, making three-point shots helps any team win, but some teams depend more on the shot from beyond the arc, or any other stat (such as offensive rebounding for the 2018 UNC team) than what is generally expected.
2017: Duke (2) lost to South Carolina (7) and Louisville (2) lost to Michigan (7)
Duke (28–9): Immediately looking at Duke’s record, it seems turnovers were their biggest X-factor. The four games with their highest TO%, they lost. This includes the game against South Carolina which put the Blue Devils at a TO% of 24.2. While hindsight is 20/20, if we look at how South Carolina was ranked 5th in the nation at defensive TO%, it is obvious why Duke struggled.
Overall, Duke depended on the three, as their lowest ten shooting games held six of their losses. The Blue Devils did not force turnovers either, ranking 253rd. South Carolina was pretty efficient in this game as well. Duke on the season had an okay defensive efficiency, ranking 47th in the nation, while South Carolina’s defensive efficiency was ranked 3rd.
Louisville (25–9): This team was talented top to bottom but ranked 232nd in FT% at 68.6 and a free-throw-attempt-to-field-goal-attempt-ratio (FTA/FGA or FTR) ranking 220th. This meant they didn’t drive and draw fouls much. Louisville was incredibly dependent on the three. Of their lowest 11 games in 3P%, they had all nine of their losses and the other two games were against Southern Illinois and Evansville, ranked 149th and 153rd respectively.
Also, they lost their lowest three FTR games, including their game against Michigan…which was at 19.7%, meaning that Michigan was careful not to foul and the Cardinals weren’t hitting shots. Louisville did not force many turnovers, which impacted them greatly, as they lost six of the nine games where their opponent had the lowest TO%. Their game against Michigan was last (or first, I guess) with the lowest opponent TO% at nine. Michigan had elite ball control, with the 4th best TO% in the nation. To be fair, Michigan shot well as they went 22-for-40 (55%) from inside the arc.
2016: Xavier (2) lost to Wisconsin (7) and Michigan State (2) lost to Middle Tennessee State (15)
Xavier (28–6): The 2016 Musketeers were pretty dependent on the 2-point shot. Their lowest 12-rated games in 2P% included all six of their losses; the other six games were against sub-40 ranked teams. Wisconsin ranked 42nd in the nation at defensive 2P% and limited Xavier to a 44.7% clip from inside the arc.
On the other side of the ball, Xavier ranked 160th in defensive 2P%, and the 11 games where their opponents had the highest 2P%, Xavier had all of their losses including to Wisco where the Badgers shot 17-for-32 (53%). Wisconsin ranked 255th at 2P%, but in this game they turned it on.
Michigan State (29–6): The Spartans lost to Middle Tennessee State, one of the biggest, baddest, toughest, best…oh, wait, nope, they were a 15 seed. If I could tell you that I can look at this game and accurately predict future 15 vs. 2 games, I’d be a millionaire. I cannot do that, but I can look at that statistics and try and give an explanation as to why Michigan State went down. MSU was horrible at forcing turnovers, ranking 343rd in the nation with a 14.2% turnover rate (mhhhh, I wonder where we saw this before). The Spartans dribble-drive was rarely used as they rated 328th in the nation at FTR.
Speaking to their weaknesses, they were solid at defensive 3P%, but when other teams had high 3P%, they struggled. Of the six games where this was the case, MSU lost five of them, including when MTSU shot 11-for-19 (57.9%) from beyond the arc, the best any team shot against MSU all year. Now maybe, just maybe, this could have been foreseen. MTSU did rank 16th in the nation at shooting the three so hey, there’s something here to exploiting MSU’s weaknesses. MSU offense that game was solid — they shot 63% from two and 45% from three. It’s just that they let up 90 points, the second-most they allowed all year.
2015: Virginia (2) lost to Michigan State (7) and Kansas lost to Wichita State (7)
Virginia (30–4): Overall, the 2015 Cavaliers team were bad at forcing turnovers and had a low FTA/FGA, ranking 248th and 232rd at each, respectively. Obviously, UVA was — and always will be — an elite defensive team, but when they faced a team with good offense (like MSU, who was ranked 14th in offensive efficiency; North Carolina, who ranked 11th; or Duke at 3rd) UVA lost. MSU’s 29th ranked 3P% showed as they shot 6-for-12 (50%) from beyond the arc. The Cavaliers also lost to teams that didn’t allow a high 3P%. This is important, as UVA shot 2-for-17 (11.8%) from three against MSU, a team that ranked 44th in the nation at defensive 3P%. UVA lost three of their final five games, so maybe momentum was not great, either.
Kansas (27–9): The Jayhawks were pretty bad at forcing turnovers, ranking 268th in the nation, but they were solid at drawing fouls. They depended on the two a lot as their eighth-lowest games in 2P% consisted of six losses with the other two being three-point and five-point games. Against Wichita St, KU shot 14-for-36 (38.9%) from two and 6-for-21 (28.6%) from three. With KU’s dependence on 2P%, it makes sense that the Shockers ranked 17th in the nation at defending the two. Wichita was also really good limiting their turnovers and steals, ranking 6th and 2nd in the nation for, which is important because KU ranked 185th in the nation for forcing turnovers.
KU got bopped by Wichita 78–65, which speaks to the power turnovers have on the game. And to be fair, Wichita shot 10-for-20 from downtown that game, which was the second best 3-point percentage a team shot against KU all season. But even if you move that number to Wichita’s average 3P% on the season at 35% (which would be 7–20 for the game), the Shockers still would have won by four.
2014: Kansas (2) lost to Stanford (10) and Villanova (2) lost to UConn (7)
Kansas (25–10): Guess what the 2014 KU team was bad at? Forcing opponents to turn the ball over, ranking 292nd. KU was consistently terrible at defending the three, ranking 240th in the nation in opponents 3P%. Guess what Stanford was good at? Shooting the three! Not elite, but top 80 in the nation.
In the Jayhawks’ seven lowest 3P% games, they lost five of them, with the other two coming to sub-100 ranked teams. They shot 5-for-16 (31%) from three against Stanford so not terrible but nothing great either. Fun fact: Stanford did not make a single three this game, they went 0-for-9 but still went 21–38 (55%) from inside the arc. Kansas’ lowest five games in offensive efficiency they lost, including Stanford, and Stanford was ranked 40th in defensive efficiency.
Villanova (29–5): Ahh, this team was better at forcing turnovers, ranking 60th. But, they were quite bad at defending the three, ranking 231st. Nova depended on on the deep shot ,as they were 7th in the nation for 3PA/FGA with 44.8% (almost half of their shots were 3’s).
The problem was, they ranked 117th in the nation at 3P%. UConn was conveniently ranked 15th in the nation at defensive eFG%, and this showed, as Nova had their lowest eFG% all year against UConn. Nova turned the ball over a fair amount at 22.6% which was their lowest of any game they lost. UConn shot 45% from three but Nova still lost by 12. If that percentage came down to 6-for-20 (30%) UConn would have edged them by three.
Whoof, that was a lot. So what does it all mean?
Basically it comes down to this: the lower-ranked teams that win are the ones that match up well with higher-ranked teams’ respective strenghts and weaknesses. Texas A&M was solid at offensive rebounding, something that UNC depended heavily on to win. South Carolina’s ability to force turnovers matched up well with how Duke struggled at ball control.
Of course, there is always going to be luck with predicting how any given team is going to shoot on a given day, whether good or bad. I mean, how do you predict MTSU shooting 57.9% from beyond the arc? That’s ridiculous! But with the hindsight of ten different two seeds getting upset before the Sweet 16, a trend emerges with just how that happened. Look at how well each team controls the ball or forces turnovers, and if they depend on any one stat to win games.
My pick this year? Louisville over Michigan State. The Spartans don’t force many turnovers, ranking 344th in the nation. This means that Louisville will have an opportunity to score nearly every time down the court. Furthermore, Louisville already beat Michigan State this year in a 82–78 overtime thriller.
Obviously, Louisville and MSU have to win both their respective first-round matchups. If they make it to the second round, though, watch out for an early exit from Michigan State. Bring on the madness!
Ryan Coleman is a college basketball fanatic who loves March more than the other 11 months combined. You can follow him on Twitter here.
