The round of 64 in the NCAA tournament is arguably one of the most exciting weekends in all of sports. Players and teams that almost nobody has heard of became national icons, nominal fans are cheering insistently to keep their bracket from being busted, and you’re either going split screen or constantly switching back and forth to catch all of the action. And then there are the upsets.
While the tournament could still be fun if the bracket went chalk, the upsets and the unpredictably of the games are what make the big dance so great. Nothing ignites the atmosphere of an NCAA tournament like a #15 knocking off a team with national title aspirations or a mid-major player becoming an instant star with a huge performance to topple a power-five giant.
As the amount of skilled college basketball players increases, so has the amount of big-time upsets. From 2012-2018, #14, #15 or #16 seeds have won a total of ten games, an average of 1.43 huge upsets per tournament (they are listed below). In the previous seven years, only three such upsets happened, all by #14 seeds.
#16 University of Maryland – Baltimore County (UMBC) 74, #1 Virgina 54
2017 – None
#15 Middle Tennessee State (MTSU) 90, Michigan State 81
#14 Stephen F. Austin 70, West Virginia 56
#14 Georgia State 57, Baylor 56
#14 University of Alabama at Birmingham (UAB) 60, Iowa State 59
#14 Mercer 78, #3 Duke 71
#15 Florida Gulf Coast (FGCU) 78, #2 Georgetown 68
#14 Harvard 68, #3 New Mexico 62
#15 Norfolk State 86, #2 Missouri 84
#15 Lehigh 75, #2 Duke 70
So how do these huge upsets happen, and is there any way to predict the next big-time giant (or two) falling in the tournament? Looking back to the past ten upsets discussed above, the data tells us that there are specific trends that stand out and are key to some of the NCAA Tournament’s most notable games.
Guards make all the difference
Out of the ten teams that pulled off a big time upset in the last seven years, nine of them were led in scoring by a guard. Kyle O’Quinn of Norfolk State was the only non-guard, a center that now plays for the Indiana Pacers. Moreover, seven of the ten teams were led veterans guards that were either juniors or seniors. Of the three teams not led by a veteran guard, two had a senior guard as their second-leading scorer.
Here’s a look at the 10 teams’ leading scorers:
|Year – Team||Player||Year – Position||Points per game|
|2018 – #16 UMBC||Jairus Lyles||Senior – Guard||20.2|
|2016 – #15 Middle Ten.||Giddy Potts||Sophomore – Guard||14.4|
|2016 – #14 SFA||Thomas Walkup||Senior – Guard/Forward||18.1|
|2015 – #14 Georgia State||RJ Hunter||Junior – Guard||19.7|
|2015 – #14 UAB||Robert Brown||Junior – Guard||13.7|
|2014 – #14 Mercer||Langston Hall||Senior – Guard||14.6|
|2013 – #15 FGCU||Sherwood Brown||Senior – Guard||15.5|
|2013- #14 Harvard||Wesley Saunders||Sophomore – Guard||16.2|
|2012 – #15 Norfolk St.||Kyle O’ Quinn||Senior – Center||15.9|
|2012 – #15 Lehigh||CJ McCollum||Junior – Guard||21.9|
Having a seasoned guard in the mix is crucial to pulling the big upset, as some of the best vets on this list dropped big-time numbers in the wins. Lyles scored 28 in the win over Virginia, while Walkup posted 33 in the win over West Virginia. McCollum dropped 30 in the big win over Duke and Brown poured in 24 points for FGCU Dunk City in their upset win over Georgetown. To top it all off, Hunter hit the game-winning three pointer to propel Georgia State to their win over Baylor. Each of those five guards who had big games were the leading scorer for their teams in those contests as they toppled power-five giants.
Pace and Space
Having a fast-paced style of play is another trend of the teams listed, as each of the squads that took down top-three seeds were ranked within the top 83 teams in the country (out of 351 schools) in field goals attempted. The average ranking of the these teams was 42.7, showing that the ability to run with the big dogs (and outpace them in many of the matchups) is a key to writing a Cinderella story. While some could say that this trend could also be attributed to offensive rebounding, that was a not key to success for many of the teams. None of the ten teams ranked above 54th nationally in offensive rebounds, while the average of upset specialists was 98th in the country in offensive boards.
Being able to spread the floor and dish assists was also a big factor, as six of the ten teams ranked either first or second in their conference in dimes per game. Nationally, the ten teams averaged to rank 55th in the nation in assists, with six of ten ranking in the top 35 (or top 10% of the country).
As you might expect, each of the Cinderella teams won their conference tournament (except for Harvard, who plays in the Ivy league with no conference tournament). Each of the teams had also won at least two games in a row coming into the big dance, and more importantly, seven of the 10 teams had at least a five-game winning streak. The three teams that did not were all 14 seeds, and each had played fairly well down the stretch, with UAB winning five of their last seven games, Mercer winning eight of their last ten, and Harvard capturing six wins in their last eight games. Their winning streaks going into the tournament are listed below:
|Year – Team||Streak|
|2018 – #16 UMBC||5|
|2016 – #15 Mid. Ten.||6|
|2016 – #14 SFA||20|
|2015 – #14 Georgia State||5|
|2015 – #14 UAB||3|
|2014 – #14 Mercer||4|
|2013 – #15 FGCU||5|
|2013- #14 Harvard||2|
|2012 – #15 Norfolk St.||7|
|2012 – #15 Lehigh||8|
On the flip side, some the supposed favorites who were shocked in their opening game of the tournament were not riding the wave of momentum heading into March Madness. Just five of the ten won their conference tournaments, and that group also included New Mexico (in a non power-five tournament), Missouri (who was 2-2 in their last four regular season games) and Virgina (who lost one of their best players, DeAndre Hunter, in the ACC conference title game). Two seeds Georgetown (2013) and Duke (2012) lost two of their last three games of the season, while Baylor (2015) and Duke (2014) had won just two and three games in a row (respectively) before falling in their conference tournaments.
Swipe and Score
Turnovers and steals were two other key statistics for the 10 teams that managed to knock off top three seeds in the round of 64. Each of the teams was ranked in the top five in their respective conferences in steals, with six of the ten either ranking first or second. Four of the teams were also ranked in the top five nationally, with only two ranking outside of the top 100.
On the flip side, turnovers were a different story for the mid-major upset specialists. Only two of the teams were in the bottom half of the country in turnovers, with the ten teams averaging to rank 227th in the nation in the statistic. Four of the teams ranked 333rd or lower, meaning that they were within the best 6% of the country offensive miscues.
This was reflected in the actual upsets, as seven of the the ten David’s knocking down Goliath’s won the turnover battle in their respective games. They also averaged just 9.6 turnovers in their upsets, while all but one of the Cinderella stories forced over ten opponent turnovers.
Experience against power-five schools
For the most part, these teams did not have cupcake schedules leading up to their conference seasons. Six of the ten teams played at least three power-five schools, while four teams claimed a victory over a power five school. Six teams also played at least one top-25 team in the non conference season. While many of these games resulted in losses, these contests gave those squads experience against a bigger schools. Each of the 10 teams either won or lost by 10 points or less to a power-five school, except for UMBC.
- Rebounding wasn’t a huge factor in the wins for the upset specialists. Just four of the ten teams outrebounded their opponents in their big wins.
- Three point shooting was a big key for some of the wins, as seven of the 10 lower-ranked teams outshot their opposition (percentage-wise) from downtown.
- Each of the upset specialists had at least 19 wins heading into the tournament, with eight out of the ten teams have 24 or more wins. Only three had double-digit losses heading into the big dance.
The unpredictably of March Madness is part of what makes it one of the greatest events in sports, but even the big dance has accumulated data to create trends for Cinderella stories. So when the tournament bracket is revealed on Selection Sunday, remember these trends. If you do, maybe you you won’t be shocked as shocked as everyone else when one of the supposed national title contenders takes a tumble in the first round of the tournament to an upstart mid-major squad.
Once the bracket for 2019 is released, BFTB will identify potential upsets based on these trends. Come back to view that article and to help fill out your bracket! On a separate note, Breakdown from the Bench also talked to players from UMBC, Florida Gulf Coast, Middle Tennessee State, and Loyola Chicago on the psychology of taking down an NCAA giant and making a March Madness run. Look for that article by Tanner Peterson coming soon.