The Ultimate Guide to Finding an Upset Special
Before we get started I want to issue you a challenge, a Bracket Challenge. If you think you have better picks than some of the ones I give away, I’d love for you to join a Bracket Challenge with me. It is just going to be a free thing so there is no harm if you aren’t too familiar with filling out a bracket. But doing one of these is always a good way to get into the game a little more and have some interaction. In this group, I will encourage people to put their serious bracket in there but if you also want to throw a funny one in that is cool. Note: You must have at least created a bracket to join a group, you don’t need to submit it until the games start on Thursday at noon eastern.
There are a few ways to do this. If you want to do it in a browser just sign in to your ESPN account and go to this link. Then scroll to the bottom and search ‘KyleTheCommish’ and there should be only one option.
If you are doing it inside the ESPN Tournament Challenge app (it’s a very good app, highly recommend for filling out the bracket vs browser) there should be a Create/Join Group button on the bracket you want to enter into the group. Click that and there is a search button at the top right. Same thing search ‘KyleTheCommish’.
If all of this is too confusing there is a way to invite you through email and also a way to invite you through a link. I’ll send those out to anyone that requests one.
Good luck! If you have any questions, don’t be afraid to ask.
I made one giant 10k word guide of awesomeness for this March Madness Bracket season. This is one of the 5 sections of that mega-article. If you would like to read the beast of an article with all 5 sections, you can click right here. If you would like to view one of the other 4 sections, click on their names: Introducing the 2022 Bracket, Tournament Facts/History, Tournament Trends, Finding My Champion. Either way, thank you for reading this article and I hope you learn something.
Finding an Upset Special
This is where a lot of my own big brain research started. And by big brain I mean…
Last year, I did an insane deep dive in my March Madness Bracket article. I was trying to find what these teams that made a run into the Sweet 16 had in common. What I found was none of them were perfect. Some of them were offensive juggernauts and others were impenetrable forces on defense. Not surprising since there are many ways to win in March. With this being my 2nd year doing this deep dive, I had the opportunity to look back on what my results yielded. All of the metrics below (besides Points Per Game and Assist/Turnover margin) were used last year and it gave me a couple of teams that won a game in the tourney, but I wasn’t able to find UCLA, Oregon St, Loyola-Chicago, or Syracuse (I didn’t include 15 seeds so I don’t know about Oral Roberts). It did correctly note that Oregon was in a good position for 2 wins but I need to close that gap and identify what those other teams did and why I liked the higher seeds more than them.
There is only one metric from last year that I used in which all five of those teams appeared together… turnovers. There is one metric that I used for the final four metrics that I didn’t include in the upset specials because it wasn’t as strongly correlated (or so I thought)… Assist/Turnover margin. I am plugging that one in this year and taking out turnovers. Here are the results.
Assist/Turnover Margin
I mentioned that the five teams above were part of my control group for finding a more precise metric that all of these teams were very strong in and could correlate to success in the tournament. Unironically, taking care of the ball and sharing the ball with your teammates is a good idea in a team sport. Each of those five teams ranked in the top 60 in the nation in this stat and averaged over a ratio of 1.2. I will use 1.2 as the low bar for these upset teams and I am going to give 2 points to people that qualify with that metric.
This is also one of the best measuring sticks for Final Four teams as well… maybe we will see another 7 seed (or higher) make a run. Perhaps Miami, who checks in with a 1.461, which is 18th in the nation ahead of teams like Purdue, Kentucky, and most importantly Auburn (their potential 2nd round matchup). Maybe you’re looking for that First Four team that always wins a game? Notre Dame is 26th in the nation at 1.376. Better than both of those teams are Colgate and Vermont. The Catamounts (Vermont) only trail Duke and Iowa among teams in the tournament.
Love our content? Check out the GoingFor2 Live Podcast Network!
Points Per Game
I looked at every upset in last year’s tournament. In those games, the winning team scored over 70 points in eleven of the fourteen games. This is something that is necessary for me. If you can’t score you likely aren’t going to beat a better team on paper but because it is a one-and-done type of format anything can happen. Of the 9 teams that made up those 14 upsets, they averaged 73.6 PPG last year. I am going to round down and use 73.5 PPG as the benchmark for finding upsets.
Watch out for South Dakota St. They average a ridiculous 85.5 points per game which is 2nd in the nation behind only Gonzaga. UAB also averages over 79 points a game. Yes, their schedules don’t often see a bunch of stellar defensive teams like UAB will face in Houston, but if you have the ability to score that much you can win any game. UNC is the highest Power 5 upset special team on this list at 77.5 PPG, which checks in at 22nd in the nation.
Effective FG%
My control group from 2017-19 had every team except one with an effective FG% rate of 52% or greater. However, Oregon St and Syracuse both did not meet that threshold this past year. Cuse was close at 51.2% but Oregon St was a lot like the other lone team that made the Sweet 16 which was South Carolina back in 2017. They were both below 50% which is in the bottom half of the country. However, both teams made a run to the Elite 8 because they were able to score over 70 points in every game they won in the tourney (except one). They are definite outliers in this bunch so I am going to keep the rate at 52% which will include about a quarter of teams in the country.
This stat is another very effective measuring statistic for Final Four teams. Guess who the #1 team in the nation is? THE JACKRABBITS THAT’S WHO! South Dakota St scores and scores and they are pretty damn efficient about it too. A 59.7% effective FG% just beats out Gonzaga for the top spot in the nation, with Vermont in 3rd. Loyola-Chicago joins the top 10 party with 56.3%. For major programs, you don’t have to scroll too far. Ohio St, Virginia Tech, Miami, and Notre Dame are all in the top 30 at above 54%.
KenPom Efficiency Rankings
I love using KenPom. If I am between two teams I almost always will default to KenPom. Therefore it is only right that I designate which teams reach certain benchmarks set by our control group. The averages from that group are as follows: 113.4 offense, 94.7 defense. You will look at those numbers and they will mean nothing, so let me explain. The offensive number is how many points the team would score per 100 possessions. The defensive number is how many points the team would allow per 100 possessions.
For comparison’s sake, 113.4 on offense includes 27 teams in the nation whereas 94.7 on defense is a bit more lenient and includes 35 teams. Not just tournament teams but every team in the nation.
Below are the teams that hit one of the metrics. Offense is in the first group. Defense is in the second. Zero teams hit both metrics.
- Virginia Tech 114.3/96.4
- Miami 114.8/102.1
- South Dakota St 116.8/105.3
- Ohio St 116.6/100.8
- Michigan 114.2/98.6
- Davidson 116.8/102
- 3 teams just miss the cut offensively: UNC 113.1/97, UAB 112.5/98.5, Notre Dame 112/98.4
- San Diego St 104.1/85.9
- Iowa St 104.4/90.4
- Boise St 108.5/92.1
- Creighton 105.6/92.1
- San Francisco 110.8/92.4
- Indiana 107.6/92.6
- Loyola-Chicago 110.9/93.2
- TCU 108.2/93.3
- Seton Hall 108.5/93.5
- Memphis 110.4/94.1
- 1 team just misses the cut defensively: Murray St 111.5/95.1
The fact that no teams seeded 7th or higher hit both marks (top 35 on defense and top 26 on offense) is quite disappointing. Last year, we had two teams (UCLA and Creighton) achieve both and achieve success in the tourney. We also had Oregon and Wisconsin who were right near that mark that won their first games each. The closest teams in 2022 based on my estimation are UNC, Memphis, Murray St, and Virginia Tech. UNC and Memphis are on the 8/9 line. Murray St is a 7, but Virginia Tech may be in a good spot as an 11 to go through a 6 and then a 3/14 matchup. A lot of people are probably keying on the Hokies because of their ACC run but KenPom agrees.
3 Pointers
One of the best ways to win in March Madness is just to shoot your opponent out of the gym. If your team just can’t stop hitting 3s, they will be tough to beat. So I need to take a look at 3 point FG% AND 3 point FG Attempt Rate. A team can shoot 3s great, but if they don’t utilize it very much, what’s the point of them being included.
36% is the average for our control group and is a pretty elite group. Only 15% of schools in the nation average that or above. In the tournament, we only have 19 teams.
Our control group didn’t rely on the 3 ball as their only source of offensive power. They were about average with only a 37% 3 point attempt rate. Any 7 seeds or worse that fall too far under or too far above that rate I will throw out. The reason I want too far above thrown out as well is that if these teams have an off-night shooting, they may not have other means of scoring consistently. I will mention the teams with very high 3 point att rates and high shooting percentages though. 32% and lower and 42% and higher catches 4 teams, however, the highest any of them were at was 43.1% so I will keep them in the sample. So we are down to 13 teams that are great at shooting 3 pointers but have other means of offense.
(Each of the 4 teams that were caught by my 10% filter was above the 42% rate. Colgate was at 42.1% and the other three were all between 43.0% and 43.2%. I designated which teams were above with a ^ sign in the data below.)
As you may have guessed by now… South Dakota St is who I am picking to win the entire tournament. They once again lead the nation in another category. They shoot and sink the 3-ball at an astounding 44.2% clip. The next closest team in the nation is a whole 4% points behind them (spoiler alert: that team is Colgate) and the first big boy school checks in at 3rd which is Virginia Tech at 39.3%. Almost 5% better than the closest Power 5. Another team that has been pretty great in some of these other statistics that I haven’t mentioned by name yet is Davidson, who checks into the top 10 at 38.4% with UAB following up behind them in 9th. Loyola-Chicago and Notre Dame are 2 of the 4 teams I flagged for shooting too many 3’s but they are both above 37.7% at hitting them, which is 20th in the nation.
Opponent 3 Point %
One last thing that I haven’t mentioned yet is defense. If shooting your opponent out of the gym is the best way to beat them, the 2nd best way might be to prevent the other team from shooting you out. When looking at defensive statistics over the control group, a few of them were great on defense, but there weren’t many commonalities between statistics. Steals, rebounding, opposing FG%, blocks, nothing was consistent. The best I could find was opponent 3 pointers. This year it grabbed only 23% of the nation which is pretty good.
Our control group had a pretty stingy 32.4% on their opponent 3 point FG percentage.
San Francisco checks into the top 10 on a list and leads the way for our upset teams at a measly 28.9% for opponents. South Dakota St would be screwed if they made only that amount of 3s! Iowa St also makes an appearance as they are also allowing under 30% from 3. The only other upset teams that are in the top 30 in the nation are New Mexico St and Murray St.
The Data/Teams
Last year, this data found a few teams that pulled off some March magic. Rutgers and Abilene Christian both had a nice first upset win but weren’t able to advance from there. UCLA made it to the Final 4 as an 11 seed but only made this list in both 3 pointers categories. By no means is this list an exact science but every team with 0 or 1 points lost in the first round except for 2 (out of 14). Food for thought.
I included all teams seeded 7th or worse in my above measurements. However, I don’t really expect a 16 or a 15 to win multiple games. If we get rid of those, here are all of the teams that were included in each yield. Team seed in parenthesis.
Assist/Turnover Margin (2 points): Loyola-Chicago, Murray St, Michigan St, Michigan, Indiana, UNC, Marquette, Rutgers, South Dakota St, Virginia Tech, Notre Dame, Miami, Davidson, Colgate, Vermont
Points Per Game: Longwood, Ohio St, Marquette, Davidson, Vermont, Miami, Montana St, Memphis, Colgate, San Francisco, Murray St, UNC, UAB, South Dakota St
Effective FG%: Michigan St, UNC, New Mexico St, Michigan, Wyoming, Marquette, Memphis, Chatanooga, Akron, UAB, Murray St, San Francisco, Montana St, Notre Dame, Miami, Virginia Tech, Davidson, Colgate, Loyola-Chicago, Vermont, South Dakota St
KenPom Efficiency (will be designated with a k): UNC, Memphis, Murray St, and Virginia Tech
3 Pointers: UNC, Montana St, Vermont^, Ohio St, Notre Dame^, Michigan St, Loyola-Chicago^, UAB, Davidson, Longwood, Virginia Tech, Colgate^, South Dakota St
Opponent 3 Point%: Boise St, UAB, Montana St, Memphis, Notre Dame, Marquette, Virginia Tech, Creighton, Loyola-Chicago, Michigan St, Longwood, Seton Hall, Chatanooga, Yale, San Diego St, Wyoming, Murray St, New Mexico St, Iowa St, San Francisco
Based on this data… who looks like they could be this year’s Cinderella?
6 Points: None
5 Points: UNC (8k), Loyola-Chicago^ (10), Murray St (7k), Michigan St (7), Marquette (9), South Dakota St (13), Virginia Tech (11k), Notre Dame^ (11FF), Davidson (10), Colgate^ (14), Vermont^ (13)
4 Points: Miami (10), Montana St (14)
3 Points: Michigan (11), Longwood (14), Memphis (9k), San Francisco (10), UAB (12)
2 Points: Indiana (12FF), Rutgers (11FF), Ohio St (7), New Mexico St (12), Wyoming (12FF), Chatanooga (13)
Teams seeded 7-14 with ZERO OR ONE point: Akron (13), Richmond (12), Boise St (8), Creighton (9), San Diego St (8), Seton Hall (8), TCU (9), Yale (14), Iowa St (11), USC (7)
*k = In the Kenpom top 40 for adjEM
^ = Shoots more than 42% 3 point shots
FF = First Four
Summary & My Upset Specials
We had 6 key measurements to try to find an upset based on research over multiple years of teams seeded 7th or higher that have made it to the Sweet 16. Keep in mind that we are looking for these teams to win multiple games. I love hitting on upsets in the first round as much as anyone, but I am getting greedy and want more.
None of our teams hit in every single category. 10 teams didn’t even get up to 2 points in our metrics which knocks out almost 30% of our field. We even had three 8 seeds and a 7 seed in that group. Yikes.
The good news is there are plenty of teams to like. If we start with our 4 teams that were flagged by our KenPom metrics, 3 of them have 5 out of a possible 6 points, UNC, Murray St, and Virginia Tech. Va Tech missed on PPG, UNC missed on defending the 3 ball, and Murray St missed on shooting the 3 ball. Of those 3 missed metrics, I think I could live with defending the 3 ball as the metric I didn’t hit.
Loyola-Chicago shoots the 3 point shot better than most teams in the country and they shoot it at a high clip as well. Confusingly, they didn’t hit the PPG threshold. This is because their pace of play is so slow. Generally, this is a good thing for the tournament so I am keeping my eye on them. It’s impossible to look past Michigan St in March but they have the same issue as Loyola. They aren’t able to score enough PPG for our threshold. Notre Dame is the only First Four team amongst our leading group because they also didn’t hit the PPG threshold. If they can beat Rutgers, they will face a struggling Alabama team in the First Round.
Marquette and Davidson are 2 mid-majors going up against big-time programs in UNC and Michigan St respectively. Marquette didn’t meet our 3 point threshold. Davidson can shoot the dickens out of the ball but struggle to defend the 3. Both of those weak points could be exploited.
I noted in the “NCAA Facts ‘Is Seeding Important? First Round found on the Tournament Facts/History article'” section that in 8 of the last 11 tournaments a 13 seed has beat a 4 seed. If it happens again this year, I’d be willing to put money on it being South Dakota St over Providence. They didn’t reach the mark in defending the 3 pointer, but who cares when they score more points than anyone (besides Gonzaga) in the nation? Providence is going to try to slow the game down but they aren’t that great defensively. Also, Providence is only ranked 49th in the country per KenPom, ahead of only 25 teams in the 68 team field. This one smells like an upset. And don’t forget Vermont who plays a very similar style of insane scoring output but not great defense. Arkansas is a better opponent than them defensively but Vermont is actually the team that will want to slow the game down. That is a very interesting matchup in a game that will be played in Western New York.
I would be remiss if I did not at least mention my beloved Colgate. Last year, I picked them to upset Arkansas, and for 34 minutes of the game they were going back and forth with them on points. Unfortunately, their gas had run out and JD Notae and crew got all the way to the Elite 8. As good as Colgate is on paper, they struggle defensively and I think they will struggle against the size and strength of Wisconsin.
All of that is great, which of them are going to win 2 games? As much as I like Loyola-Chicago and Sister Jean, Villanova will not be surprised by the slow pace that they employ if they meet in R2. Additionally, I will be picking South Dakota St>Providence but Iowa is the 2nd most efficient offense in the country. if they meet in R2, I wouldn’t pick them. However, I keep looking at the East Bracket and saying it is prime for upsets. I think we could see UNC beating Marquette and then an injury-stricken Baylor Bears to make a little run. I am also certainly picking Va Tech>Texas and they will give Purdue all they can handle. Purdue generally has underachieved in the tournament. My last Upset Special is going to be Tom Izzo and Michigan St upsetting and ending Coach K’s magical run at Duke. The ACC is not strong this year and I think Duke has not been good of late.
Who is your Cinderella? Who do you think will cut down the nets? Let me know in the comments or at @KyleTheCommish on Twitter.
ATTN Dynasty Commissioners: Do you want to do something cool for your league? How about a 1-hour live show dedicated to YOUR league? Team-by-team breakdowns, rankings, and more. For details and to book a show, visit: GoingFor2.com/plp.
