Yet Another Sports Blog

Greetings, fellow March Madness fans. March Madness is my favorite annual sporting event, and I’ve enjoyed decades of watching so many great tournaments. I also enjoy a host of statistics surrounding the tournament. As you may be aware, there are ample statistics readily available pertaining to scoring, rebounds, assists, etc., for teams as well as individual players. However, I’m more interested in other things, such as which of the 314 teams and few dozen conferences have performed the best overall, as well as the best 1-, 3-, 5-, and 10-year performances. And which teams/conferences perform well, and not so well, round by round. And which teams/conferences pull off the most upsets – and conversely which are upset the most. And a few other things. But most of this information is not readily available, so in order to satisfy my curiosity, I took the time to compile the data and publish my findings. And I hope some of you find it interesting. Before I get to the data, though, there are several caveats.

First, what follows only takes into account the tournament for the years 1985 through 2025. Why these years only? In 1985, the current 64(+) team tournament format began – and has remained its largely unchanged beautiful self for this entire period (albeit with 10 years of single and 14 years of quadruple play-in games). From 1975 through 1984, the tournament inconsistently ranged from 32 to 54 teams. And if that isn’t different enough, prior to 1975 only one team from each conference (its tournament champion) was allowed to play in “the Big Dance”; hence, there were no meaningful conference comparisons at that time – instead, the “best” conference was the one whose lone qualifying team performed the best. So, by limiting this analysis to 1985 through 2025, we are left with a nice, round 40 years of data. (Yes, that period is actually 41 years, but the 2020 tournament was cancelled due to Covid.)

Second, conference records represent the aggregate records of the schools in the conference at the time of each tournament. So, for example, the ACC includes Maryland’s results from 1985 to 2014, whereas Maryland’s results thereafter are included with the Big Ten. And so on and so on.

Third, prior conferences that merged into another are counted with its current conference. For instance, in 1995 the majority of schools from the Metro Conference and Great Midwest Conference merged to form Conference USA (hereafter just “USA”); thus all Metro and Great Midwest schools that migrated to USA are counted as USA. As another example, all Big 8 schools were wholly incorporated into the Big 12 and are therefore counted as such. Similarly, the PAC-8/10/12 are hereafter referred to as just “PAC”.

Fourth, all data reflects actual results; that is, there are no adjustments for games forfeited at a later date due to rules violations.

Finally, the data only takes into account wins and losses; scores, and more specifically margins of victory, are excluded. I shudder to think how many errors I might have made had I included scores. Nevertheless, while it’s certainly possible I made a mistake somewhere, the data presented herein agrees with what information I am able to verify online. So I am confident the data is virtually correct.

With all of that out of the way, let’s move on to the data. First we’ll cover conference performance, followed by school performance, and end with a section on upsets.

Best and Worst Conference Performance

Here are the conferences that have won at least 50% of their March Madness contests from 1985-2025:

65.6% (394-207) ACC
60.9% (340-218) Big East
60.0% (310-207) SEC
59.3% (351-241) Big Ten
58.4% (302-215) Big 12
56.8% (208-158) PAC
51.3% (102-97) USA
50.0% (27-27) American Conference (formerly the American Athletic, hereafter AAC)

Most people already know the ACC has the best overall record. Just how much better is the ACC than the next best conference? The Big East would need 75 additional wins, or 40 fewer losses, to have the same winning percentage as the ACC. In other words, it’s going to be quite a long time before another conference catches the ACC. (Be prepared for other ACC superlatives throughout.) As one can see, there really are six “power” conferences (ACC, Big East, SEC, Big Ten, Big 12, and PAC). And of these six powers, by most measurements the PAC is the laggard. But the PAC deserves special mention for two reasons: 1) the recent exodus of most of its teams to the other power conferences; and 2) how different some of the data would like if we included tournaments prior to 1985. (If you’re unaware, look up how UCLA absolutely dominated the tournament in the 60’s and 70’s.)

Just making the 50% threshold is the AAC, which has only existed for 11 years. Only time will tell if they can maintain this status.

Another way of looking at the conference performance above is the number of wins per appearance; that is, the total number of conference wins divided by the total number of conference team appearances during these last 40 tournaments. Here are those figures:

1.81 ACC
1.49 Big East
1.45 SEC
1.44 Big Ten
1.38 Big 12
1.30 PAC
1.04 USA
0.96 AAC

In both the winning percentages above as well as these figures, note the significant dropoff after the power six conferences. Notwithstanding that, Conference USA is clearly next in line if anyone suggests there are seven power conferences.

The lowest conference winning percentages:

0.0% East Coast Conference (ECC, which disbanded in the early 90’s) (0-7)
6.7% Big South (2-28)
10.0% Big Sky (5-45)
10.3% Patriot League (4-35)
13.2% Mid-Eastern Athletic Conference (MEAC) (5-33)
14.3% America East Conference (AEC) (8-48)

Best 1-year Conference Performance

When determining conference performance, a minor dilemma arises. How should one treat games in which conference foes face one another? Almost every compilation of conference performance counts these games as both a win and a loss for the conference, and for good reason. There have been only 33 such games in these last 40 tournaments. (The ACC and Big East lead the way with 8 intra-conference games apiece.) As a result, these match-ups don’t radically alter conference performance data, so all figures herein do not make any special adjustments for such games, except …

It actually can matter how intra-conference games are treated when determining the best 1-year conference performance. Assuming the purpose of conference records is to compare conference performance, then it not only makes sense but is also imperative to not count such games as both a win and loss – especially given all but one of the 33 games occurred within conferences with a 50%+ win percentage. Instead, a seemingly reasonable adjustment would be to “ignore” these games, i.e., treat them as neither a win nor a loss. Having said all that, here are the best 1-year conference winning percentages (so noted if adjusted):

85.7% Big 12 in 1988 (14-4 overall, but 12-2 after adjusting for two intra-conference games)
84.2% Big East in 1985 (18-5 overall, but 16-3 after adjusting for two intra-conference games)
83.3% USA in 2008 (Memphis went 5-1)
83.3% Horizon League in both 2010 and 2011 (Butler went 5-1 both years)
83.3% Big East in 2024 (10-2)

The best 3-year and 5-year conference performances are held by the SEC in overlapping years:

77.1% 1994-1996 (37-11)
72.9% 1992-1996 (51-19)

The best 10-year conference performance:

68.9% ACC 1986-1995 (113-51)

It surprised me that there is only a 3.3% difference between the best 10-year win% and the ACC’s overall win% (65.6%). When digging deeper, I discovered a couple interesting stats that reflect the ACC’s tremendous consistency, relative to the other conferences, throughout these last 40 tournaments. First, the ACC holds the best 10-year win% for every 10-year period (i.e., 1985-1994, 1986-1995, …, and 2015-2025) except for 2005-2014 and 2006-2015, when the SEC has a better win%. Second, and perhaps more striking, the ACC has won less than five games in a single tournament only once (4 wins in 2021), whereas every other conference has won fewer than five games at least six times!

Round-by-Round Conference Performance

I’m not sure how meaningful some of this is, especially in the earlier rounds; but with each subsequent round, I find it more and more interesting.

Play-in-Round

To date, there have been 66 play-in games. The following seven conferences have never lost a play-in game, with the total number of wins indicated:

3 Atlantic 10
2 Big West, Colonial Athletic Association (CAA), Missouri Valley (MVC), and Summit League
1 Big 12 and Horizon

Among conferences that have lost a play-in game, the best win% is 75% held by the ACC (6-2) and America East Conference (3-1, hereafter AEC).

Only two conferences have lost all of their play-in games: the AAC (0-3) and USA (0-1). Among conferences that have won a play-in game, the lowest win%is 14.3% held by the Mountain West Conference (1-6, hereafter MWC).

The Northeast Conference (NEC) has the most play-in game wins with a record of 7-7, and the Southwestern Athletic Conference (SWAC) has the most play-in game appearances and losses, with a record of 6-10.

Round 1 Games

The best conference records in opening round games are owned by the “power six”:

74.1% ACC (160-56)
73.0% Big Ten (176-65)
70.2% Big East (158-67)
66.2% Big 12 (145-74)
64.9% SEC (137-74)
63.5% PAC (99-57)

And the conferences that win fewer than 10% of their matchups in the opening round:

0.0% ECC (0-7)
3.0% NEC (1-32)
3.4% SWAC (1-28)
4.0% Big South (1-24)
6.8% Big Sky (3-41)
7.4% MEAC (2-25)
9.1% Patriot (3-30)

Round 2 Games

Conferences that win at least half their games in the round of 32:

66.3% ACC (106-54)
62.0% SEC (85-52)
61.6% PAC (61-38)
54.5% Big 12 (79-66)
54.4% Big East (86-72)
50.0% Big Ten (88-88)
50.0% Horizon (7-7)

The following eight conferences have lost all of their 2nd round games, with total appearances in parentheses: Ohio Valley Conference, (OVC, 8), AEC (5), Big Sky (3), Patriot (3), MEAC (2), Big South (1), NEC (1), and SWAC (1). Among conferences that have won in the 2nd round, the lowest win% is 12.5% by the Sun Belt Conference (2-14).

The power six conferences really strut their stuff in the first two rounds as seen above. But that is about to change.

Round 3 Games (Sweet Sixteen)

The top ten conference records in the Sweet Sixteen:

100% Metro Atlantic Athletic Conference (MAA) (1-0)
75.0% CAA (3-1)
75.0% SWAC (3-1)
66.7% Big West (4-2)
57.1% A10 (12-9)
56.5% USA (13-10)
55.7% Big 12 (44-35)
55.7% ACC (59-47)
54.7% Big East (47-39)
52.9% SEC (45-40)

Five conferences have lost all of their indicated appearances in the Sweet Sixteen: Ivy League (2), Sun Belt (2), Summit (2), Atlantic Sun (1), and Southland (1). Among conferences with a Sweet Sixteen victory, the lowest win% is 11.1% by the MWC (1-8).

Round 4 Games (Elite Eight)

Conferences that have won at least half of their Elite Eight games:

100% Horizon (2-0)
100% MVC (2-0)
100% MWC (1-0)
75.0% Big West (3-1)
66.7% AAC (2-1)
66.7% CAA (2-1)
60.5% Big Ten (26-17)
57.6% ACC (34-25)
53.8% USA (7-6)
51.1% SEC (23-22)

As is also the case in the 3rd round, not a single power six conferences has one of the top six records by win%. In this 4th round, though, the top six have very few appearances. It’ll take a while, but let’s see if those conferences can maintain the same success rate over, say, a dozen or two Elite Eight appearances.

Three conferences have lost all (“all” being one) of their Elite Eight appearances: the MAA, Mid-American Conference (MAC), and Southern Conference. Among conferences with at least one Elite Eight win, the lowest win% of 8.3% belongs to the Atlantic 10 (1-11).

Round 5 Games (Final Four)

Four conferences have won all of their Final Four games: Horizon (2-0), West Coast Conference (WCC, 2-0), MWC (1-0), and Western Athletic Conference (WAC, 1-0). Among conferences with a Final Four loss, the best win% is 60.9% by the Big East (14-9).

Four conferences have lost all of their Final Four appearances: the CAA (0-2), MVC (0-2), Atlantic 10 (0-1), and Southwest Conference (0-1), which disbanded in 1996. Among conferences with a Final Four win, the lowest win% is 28.6% by USA (2-5).

Round 6 Games

The AAC (UConn in 2014) and Big West (UNLV in 1990) won their one and only appearance in the championship game. These conferences with a championship game loss have won at least 50% of such games:

71.4% Big East (10-4)
63.6% SEC (7-4)
61.1% ACC (11-7)
50.0% PAC (2-2)
50.0% USA (1-1)

Four conferences have lost all of their championship game appearances: the Horizon League (0-2), WCC (0-2), MWC (0-1), and WAC (0-1). Among conferences with a championship, the lowest win% is 23.1% by the Big Ten (3-10).

Conferences with the most Final Four appearances:

34 ACC
26 Big Ten
23 Big East
23 SEC
19 Big 12
11 PAC
7 USA

Conferences with the most Championship Game appearances:

18 ACC
14 Big East
13 Big Ten
11 SEC
10 Big 12
4 PAC

All 40 Titles by Conference:

11 ACC
10 Big East
7 SEC
4 Big 12
3 Big Ten
2 PAC
1 AAC
1 Big West
1 USA

Have you ever wondered about “conference depth”? If so, how should it be measured? While I don’t have a perfect answer, I do have at least one proposal thanks to a stray comment from sometime past. Someone once said to me (although I can’t recall who it was – friend? acquaintance? ESPN commentator? Oprah Winfrey?) “the only reason the ACC has the best record is because of Duke and UNC”. Which sounds pretty reasonable, given Duke and UNC have two of the better March Madness records (see below). That led me to wonder just how well the ACC performs apart from those two, which seems like one fairly decent way to measure the ACC’s depth. Of course, the same could be done for any conference. This is very simple to calculate in the case of Duke and UNC, who have not changed conference affiliation since 1985 (or ever, for that matter). It’s more difficult with, say, the Big East, where UConn (spoiler alert: the best record!) moved out of and back into this conference. But it’s certainly not impossible. So here is that measure of depth for each of the power six conferences (I have chosen not to include other conferences due to the difficulty of the calculation, as well as knowing intuitively none would come anywhere close to the power six figures): each conference’s win% after removing their top two performing teams:

56.7% ACC (186-142 excluding Duke/UNC)
56.6% Big East (232-178 excluding UConn/Villanova)
52.8% Big 12 (187-167 excluding Kansas/Baylor)
52.5% SEC (170-154 excluding Kentucky/Florida)
49.74% PAC (95-96 excluding UCLA/Arizona)
49.73% Big Ten (183-185 excluding Michigan/Michigan State)

The ACC has a very slim margin, which could easily change from year to year.

One final comment on conferences: how amazing that the SEC placed fourteen teams in the 2025 tournament! The only other time a conference placed double digit teams in the Big Dance was the Big East’s 11 teams in 2011. And those 14 SEC teams achieved 23 wins, the most ever by a conference in a single tournament; the previous record was 19 by the ACC in the 2006 tournament.

And now, on to school performance.

Best and Worst School Performance

Here are the top ten schools by win%:

78.2% UConn (68-19)
76.8% Duke (109-33)
75.6% UNC (99-32)
73.5% Kentucky (86-31)
72.3% Kansas (94-36)
71.1% Florida (54-22)
69.7% Michigan (53-23)
67.1% Syracuse (57-28)
66.7% Michigan St (66-33)
66.7% Loyola of Chicago (8-4)

There are three other schools, besides Loyola of Chicago, who have won more than 50% of their games yet have fewer than ten wins: Florida Atlantic (57.1%, 4-3), Loyola Marymount (57.1%, 4-3), and Rhode Island (53.3%, 8-7).

As for the worst performing schools, over 30% of the 314 schools that have appeared in the Big Dance have yet to secure a win. Of these schools, here are the half dozen with the most losses: Iona (13), Boise St (9), Nebraska (8), Akron (7), Colgate (7), and S Dakota St (7). Among schools with a win, here are those with the lowest win%:

7.1% Penn (1-13)
7.7% Montana (1-12)
8.3% Winthrop (1-11)
10.0% East Tennessee St (1-9)
11.1% Belmont (1-8)
11.1% Southern (1-8)

Among schools with multiple wins, the lowest win% belong to Utah St (11.8%, 2-15) and Chattanooga (18.2%, 2-9).

Among schools with double digit wins, the lowest win% belongs to VCU (38.5%, 10-16), Providence (41.7%, 10-14), Missouri (42.5%, 17-23), Mississippi St (43.5%, 10-13), and Clemson (44.0%, 10-14).

Best 1-year School Performance

There is a 40-way tie for the best 1-year March Madness performance by a school, as every champion had a 6-0 record. Would a 7-0 record count as surpassing a 6-0 record? I suppose so, and it’s certainly possible should a play-in team ever win a championship. Two play-in teams have reached the Final Four – 11 seeds VCU in 2011 and UCLA in 2021. Yet the lowest seed to win a championship thus far is 8th seeded Villanova in 1985. So we’ll just have to be patient until someone can break this tie.

Best 3-year Performance

What constitutes the best 3-year March Madness performance? There have been three repeat champions from 1985 to 2025: Duke in 1991-92, Florida in 2006-07, and UConn in 2023-24. Yet no school has won three consecutive championships during this period. Nevertheless, three schools have perfect March Madness records in a 3-year period. How can this be? From 1985-87 and 2013-15, Louisville and UConn, respectively, won the championship sandwiched between two years of missing the tournament; thus both were 6-0 during those periods. Even better, Florida followed up their back-to-back championships with missing the tournament, thus finishing 12-0 from 2006-08. If one does not hold it against those schools that they did not qualify for the tournament, then these are indeed the best 3-year school performances.

If, on the other hand, one feels that missing the tournament should disqualify a school from consideration, then the best 3-year performance obviously necessitates three consecutive years of qualifying for the tournament. But how often does this happen? Honestly, I don’t know, although I would guess this number to easily be in the hundreds. Regardless, despite the lack of a three-peat champion, two teams have come as close as possible without quite achieving it. In 1990, Duke lost to UNLV in the finals, then followed that with their repeat championships, thus finishing with a 17-1 record from 1990-92. This record was matched by Kentucky from 1996-1998, whose loss to Arizona in the 1997 final was bookended with two championships.

Best 5-year School Performance

Revisiting whether missing the tournament should disqualify a team when determining performance, UConn had a record of 13-1 (92.9%) from 2010-14 as well as from 2011-15 – by virtue of winning two championships, losing in the second round, and missing the tournament the other years. No team with five consecutive years of tournament participation has as high a win% as this.

Among those with five consecutive years of tournament games, Duke’s 1990-92 run was preceded by two Final Four appearances; hence their 1988-92 record was 25-3 (89.3%). Not far behind, Kentucky’s 1996-98 run was sandwiched between two Elite Eight appearances; so their 1995-99 record was 23-3 (88.5%). Two impressive runs that are hard to imagine happening in this day and age, although UConn has a chance to surpass them with a championship this year or next.

Best 10-year Team Performance

From 2010-2019, UConn’s 10-year win% was 87.5%! Their record was 14-2 as a result of two championships, two second round exits, and six years of not being chosen on Selection Sunday.

If we limit ourselves to teams with ten consecutive years of tournament appearances, we are now talking about a very limited sample size. Nevertheless, Duke’s 1985-1994 record was 39-8 (83.0%), and included those back-to-back championships, three losses in the finals, and two Final Four losses. That’s seven Final Four appearances in one decade – once again, very hard to imagine a school accomplishing that in this era of one-and-done and the transfer portal.

Round-by-Round Team Performance

Play-in Round

With only 24 years of play-in games thus far, this round would be more interesting if some of the same teams participated in these games each year. That is certainly not the case. But if one team does stand out, it would have to be Texas Southern; this school has participated in the most play-in games (6), and is tied for the most wins (3) and losses (3). Mt St Mary’s (3-2) is the only other school with three wins, and Boise St (0-3) and NC Central (0-3) are the only others with three losses.

Round 1

Only four schools have never lost in the first round: Northwestern (3-0), Tulane (3-0), Duquesne (1-0), and Furman (1-0). Among all other schools, the top ten in win% are:

92.3% Kansas (36-3)
92.3% Auburn (12-1)
91.7% UNC (33-3)
89.5% Duke (34-4)
88.2% Kentucky (30-4)
87.5% Maryland (21-3)
85.2% Gonzaga (23-4)
84.0% UConn (21-4)
83.3% Michigan (20-4)
82.8% Syracuse (24-5)

As for the other extreme, please refer to the section above detailing the worst performers among the 30%+ of schools that have yet to win. Obviously, the overwhelming majority of those teams lost in the first round. However, two schools have never made it to the first round due to losing their one and only tournament appearance in a play-in game: Alabama A&M and North Florida.

One noteworthy item pertaining to teams without a March Madness win: Nebraska is the only school from a power six conference without a win (ever!, not just from 1985 to 2025). However, this year the odds are heavily in favor of them finally securing a victory (or possibly several).

Round 2

Loyola of Chicago (3-0) has the most wins without a loss in the 2nd round; the only other undefeated Round of 32 schools won their one and only appearance: Chattanooga, Cornell, Davidson, Florida Atlantic, Florida Gulf Coast, Oral Roberts, Oregon St, South Carolina, St Peter’s, and Valparaiso.

Among teams with a 2nd round loss, the best records are:

87.0% Arizona (20-3)
85.7% Houston (6-1)
82.4% Duke (28-6)
80.0% Clemson (4-1)
80.0% Vanderbilt (4-1)
76.7% Kentucky (23-7)
76.2% UConn (16-5)
75.8% UNC (25-8)
75.0% Texas Tech (6-2)
73.9% UCLA (17-6)

Dozens of schools have lost all of their appearances in the 2nd round. Among these, the teams with the most losses are: New Mexico (7), Murray St (5), Saint Louis (5), Charlotte (4), and Colorado (4). Among the schools without a Round of 32 win, the worst records are:

16.7% VCU (1-5)
20.0% Dayton (1-4)
20.0% St Mary’s (1-4)
20.0% UAB (1-4)
21.4% Iowa (3-11)

Round 3 (Sweet Sixteen)

Temple (5-0), Kansas St (4-0), and St John’s (3-0) are the only schools with multiple wins without a loss in the Sweet Sixteen, while these schools won their lone appearance: Davidson, Dayton, Florida Atlantic, George Mason, Kent St, Loyola Marymount, Navy, Oregon St, South Carolina, St Peter’s, and VCU.

These schools with a loss in the 3rd round have the best win%:

83.3% Florida (10-2)
75.0% UConn (12-4)
75.0% Missouri (3-1)
73.9% Kentucky (17-6)
72.0% UNC (18-7)
71.4% Memphis (5-2)
70.0% Villanova (7-3)
69.6% Kansas (16-7)
67.9% Duke (19-9)
66.7% UNLV and Virginia (both 4-2), and Minnesota, Providence, and UMass (all 2-1)

The following schools have yet to win in the indicated number of Sweet Sixteen appearances:

4 Vanderbilt and Washington
3 Texas A&M
2 BYU, Cal, DePaul, Nevada, Ole Miss, Richmond, Southern Illinois, and Western Kentucky
1 Arizona State, Ball St, Bradley, Chattanooga, Cleveland St, Cornell, Eastern Michigan, Florida Gulf Coast, Georgia, George Washington, LaSalle, Louisiana Tech, Miami of Ohio, Milwaukee, Missouri St, New Mexico St, Northern Iowa, Ohio, Oral Roberts, Penn St, Princeton, UAB, UTEP, Valparaiso, Virginia Tech, Washington St, and Wyoming

Among schools with a Sweet Sixteen victory, these are the lowest win%:

14.3% Iowa St (1-6)
20.0% Maryland (2-8)
20.0% Pitt (1-4)
25.0% West Virginia (2-6)
25.0% Clemson (1-3)
25.0% San Diego St (1-3)

Round 4 (Elite Eight)

Butler and Maryland are the only schools with multiple Elite Eight wins without a loss (both are 2-0). Other Elite schools that won their lone Elite Eight appearance are Florida Atlantic, George Mason, Loyola of Chicago, Mississippi St, San Diego St, South Carolina, VCU, and Wichita St.
The best records among schools with a loss in this round belong to:

75.0% Indiana (3-1)
75.0% UNLV (3-1)
75.0% Wisconsin (3-1)
73.7% Duke (14-5)
72.7% Michigan St (8-3)
71.4% Syracuse (5-2)
71.4% UCLA (5-2)
71.4% Villanova (5-2)

Interestingly, there are four schools that, upon reaching their only Sweet Sixteen appearance, made it to the final Four: George Mason, Loyola of Chicago, South Carolina, and VCU. Of these schools, only 7 seed South Carolina was not an 11 seed.

And these ten schools have the most total wins without ever reaching the Final Four:

31 Xavier
26 Tennessee
24 Temple
23 Iowa St
21 Pitt
19 Notre Dame
18 Florida St
18 Iowa
17 Kansas St
17 Missouri

Here are the winless schools in the Elite Eight, listed by total appearances:

5 Temple
4 Kansas St
3 Missouri, Tennessee, and Xavier
2 Florida St, Notre Dame, and USC
1 Boston College, Clemson, Creighton, Davidson, Dayton, Iowa, Iowa St, Kent St, Loyola Marymount, Navy, Oregon St, Pitt, Rhode Island, St Josephs, St Peter’s, Tulsa, and Wake Forest

Among schools with an Elite Eight win, these are the lowest records:

20.0% Texas (1-4)
25.0% Oregon (1-3)
25.0% Purdue (1-3)
25.0% Virginia (1-3)

Just in case you missed it, Temple (5-0) and Kansas St (4-0) are perfect in the Sweet Sixteen – yet both are perfectly imperfect in the Elite Eight, losing all nine of those opportunities to advance to the…

Final Four Round

Teams that have not lost in the Final Four by number of appearances:

5 Michigan
2 Butler and Gonzaga
1 Baylor, Purdue, San Diego St, Seton Hall, Texas Tech, Utah, and Virginia

The best records among schools with a Final Four loss :

85.7% UConn (6-1)
66.7% Florida (4-2)
66.7% Arkansas (2-1)
66.7% Indiana (2-1)
64.3% Duke (9-5)

Here are the Final Four teams without a win, listed by games:

2 Auburn, LSU, and Oklahoma St
1 Alabama, Cincinnati, Florida Atlantic, George Mason, Loyola of Chicago, Marquette, Miami, Minnesota, Mississippi St, North Carolina St, Oregon, Providence, South Carolina, St John’s, Stanford, Texas, UMass, VCU, Wichita State, and West Virginia.

These schools with at least one Final Four victory have a win% below 50%:

25.0% Michigan St (2-6)
33.3% Ohio St (1-2)
33.3% Oklahoma (1-2)
33.3% UNLV (1-2)
33.3% Wisconsin (1-2)
40.0% UCLA (2-3)

Championship Games

Here is the win% for all 34 teams that have played in the championship game:

100% UConn (6-0!!!)
100% Villanova (3-0)
100% Louisville (2-0)
100% Baylor (1-0)
100% Maryland (1-0)
100% UNLV (1-0)
100% Virginia (1-0)
75.0% Florida (3-1)
66.7% UNC (4-2)
60.0% Kentucky (3-2)
55.6% Duke (5-4)
50.0% Kansas (3-3)
50.0% Arizona (1-1)
50.0% Arkansas (1-1)
50.0% Indiana (1-1)
50.0% Michigan St (1-1)
50.0% UCLA (1-1)
33.3% Syracuse (1-2)
20.0% Michigan (1-4)
0.0% Georgia Tech (0-1)
0.0% Georgetown (0-1)
0.0% Houston (0-1)
0.0% Illinois (0-1)
0.0% Memphis (0-1)
0.0% Ohio St (0-1)
0.0% Oklahoma (0-1)
0.0% Purdue (0-1)
0.0% San Diego St (0-1)
0.0% Seton Hall (0-1)
0.0% Texas Tech (0-1)
0.0% Utah (0-1)
0.0% Wisconsin (0-1)
0.0% Butler (0-2)
0.0% Gonzaga (0-2)

A few comments on the final two rounds of the tournament.

UConn is the king of clutch! Winning six championships in only 7 Final Four appearances is pretty amazing. The only team to beat UConn once they’ve reached the Final Four is Michigan St in 2009. UConn is also interesting because they’ve only played in 25 of the last 40 tournaments, meaning they win just about once in every four tournaments. If Kansas and Duke had the same success rate, they would each have about nine championships. (Kansas and Duke have the most tournament appearances since 1985 – 39 and 38, respectively.)

Only two teams are perfect upon reaching the final weekend: Baylor and Virginia each won the championship in their lone Final Four appearance. At the opposite extreme, Wisconsin has appeared in the most Final Fours (3) without winning a title.

Upsets! Upsets! Upsets!

What constitutes an upset can be a very subjective thing. Some folks would judge an upset strictly by the point spread. Fair enough. But since every team in the tournament is seeded, when it comes to March Madness almost everyone considers an upset to have occurred when a lower seeded team defeats a higher seeded team. Very simple. (Yes, everyone can find an example where the selection committee totally botched a school’s seeding – but overall the committee has done an admirable job.)

So which conferences and schools are the most proficient (and deficient) at pulling off upsets? Great question – but that just leads to having to define proficiency. For instance, should one just tally up the number of lower seeded teams beating higher seeded teams? That would certainly be simple to do. Or instead, should each upset be weighted in some way so that Fairleigh Dickinson’s upset of Purdue (one of only two 16 seeds defeating a 1 seed, the other being UMBC over Virginia)) counts much more than a 9 seed beating an 8 seed? And if so, how should each upset be weighted?

A weighted system obviously makes much more sense; but what is the ideal weighting system? In pondering this, and to save myself much mental consternation, I opted to utilize the simplest weighting system imaginable – one based upon the difference in value of the opponents’ seeds. Thus, for instance, 16 seed Fairleigh Dickinson would be awarded 15 “upset points” for defeating 1 seed Purdue. And of course a 9 seed would receive only 1 upset point for beating an 8 seed. (Perhaps it should be the other way around since 9 seeds have a better record against 8 seeds than vice versa.) By this method, one can easily determine which schools/conferences are the upset masters, and which not so much.

Furthermore, upset points at the same time generate both positive and negative points. For instance, Fairleigh Dickinson’s 15 upset points are positive (16-1=15), while Purdue would receive 15 negative upset points (1-16=-15). And these negative upset points can be used to determine those schools and conferences that are the most upset prone. (For purposes of clarification, negative upsets points will hereafter be referred to as “x negative points”, rather than using a negative sign.)

But it’s not just about total upset points. Fairleigh Dickinson is a perennial 16 seed (although they were seeded 15th once). They’ve had five other opportunities to pull off an upset without success. How should they compare with the only other successful16 seed? UMBC also has 15 total upset points, but they accomplished it in fewer tries. By further calculating potential upset points, we can divide actual upset points by this value to determine the percentage of successful upset points. Naturally, a high percentage of successful positive upset points is good and a low percentage is bad, whereas the opposite is true of negative upset points (high percentage bad, low percentage good).

Before continuing, I realize this is not exactly the most sound mathematical analysis; nevertheless I at least found it pretty interesting. But I’ll understand if you just skip this whole section. And before we get to the conference and school upsets, here is some interesting general information on upsets by round.

First here are the seed vs. seed records in the opening round:

1 vs. 16: 98.8% (158-2)
2 vs. 15: 93.1% (149-11)
3 vs. 14: 85.6% (137-23)
4 vs. 13: 79.4% (127-33)
5 vs. 12: 64.4% (103-57)
6 vs. 11: 61.3% (98-62)
7 vs. 10: 61.3% (98-62)
8 vs. 9: 48.1% (77-83)

In the first round of the tournament, there have been a total of 10,240 potential upset points from 1985 to 2025. This can be easily calculated by adding 15 (16-1) + 13 (15-2) + … + 1 (9-8), multiplying by four (for each regional), and further multiplying by 40. Potential upset points in subsequent rounds can’t be easily calculated apart from looking at each and every matchup. Nevertheless, here are the actual percentages of successful upset points by round (there are no potential upset points for play-in games since those games match equally seeded teams):

Round 1: 16.6% (1,701 of 10,240 potential upset points)
Round 2: 25.5% (930 of 3,647)
Round 3: 24.0% (330 of 1,373)
Round 4: 37.2% (186 of 500)
Round 5: 26.2% (51 of 195)
Round 6: 26.8% (19 of 71)
All rounds combined: 20.1% (3,217 of 16,026)

I was surprised the first round had the lowest success rate, by a relatively significant amount, of upset points. And since there are so few upsets in the championship games, here is what makes up those 19 upset points, in chronological order:

7: 8 seed Villanova over 1 seed Georgetown in 1985
1: 2 seed Louisville over 1 seed Duke in 1986
5: 6 seed Kansas over 1 seed Oklahoma in 1988
3: 4 seed Arizona over 1 seed Kentucky in 1997
1: 3 seed Syracuse over 2 seed Kansas in 2003
1: 3 seed Florida over 2 seed UCLA in 2006
1: 2 seed Villanova over 1 seed UNC in 2016

Conference Upsets

Conferences with the ten highest percentages of converted positive upset points (the ACC and SEC rounded to the same value, hence the extra decimal):

40.55% SEC (296 actual of 730 potential upset points)
40.48% ACC (255 of 630)
34.3% Big East (243 of 709)
33.5% PAC (210 of 626)
31.6% MVC (165 of 522)
30.2% AAC (26 of 86; low numbers due to just 11 years of existence)
28.6% CAA (143 of 500)
27.5% WCC (115 of 418)
26.9% Big Ten (204 of 759)
26.7% USA (114 of 427)

Please note all of the power six conferences made the top ten with the exception of the Big 12 (which came in at #11 with a 25.1% success rate).

And here are the conferences with the lowest percentage of potential upset points achieved:

0.0% ECC (0 actual upset points of 101 potential; only 7 tries prior to conference dissolution)
1.5% Big South (5 of 327)
2.2% SWAC (9 of 410)
3.3% NEC (15 of 460)
5.3% Big Sky (29 of 546)

Shifting now to conferences that are upset the least, here are the conferences with the lowest percentage of actual negative upset points vs. potential negative upset points:

0.0% Southland (0 actual negative upset points of 7 potential negative upset points)
0.0% MAC (0 of 6)
4.9% Horizon (2 of 41)
8.3% CAA (1 of 12)
9.6% SWC (11 of 115)
12.5% Ivy (1 of 8)
13.0% Big West (24 of 184)
15.5% ACC (432 of 2,793)
15.9% WCC (54 of 340)
18.2% MAA (2 of 11)

Several conferences have relatively few potential negative upset points because they rarely have teams seeded higher than 9th. Note the ACC is the only power six conference to make the top ten.

And here are the conferences that are upset the most:

50.0% OVC (8 actual negative upset points of 16 potential)
38.5% MWC (84 of 218)
37.7% MVC (40 of 106)
28.5% AAC (51 of 179)
25.6% Atlantic 10 (121 of 473)

School Upsets

Three schools have never had an opportunity for an upset; Alabama A&M and North Florida were mentioned above as having lost their only tournament appearance in a play-in game. The other school is Jacksonville, which lost their only tournament appearance as an 8 seed in 1986.

The following schools convert more than double the average potential upset points (20.1%):

65.0% Loyola of Chicago (26 actual upset points of 40 potential upset points)
58.6% Penn St (17 of 29)
57.1% Cleveland St (28 of 49)
53.3% Florida St (32 of 60)
52.9% Furman (9 of 17)
51.9% LSU (55 of 106)
50.0% Kentucky (44 of 88)
48.7% Oregon St (19 of 39)
48.6% Villanova (53 of 109)
47.8% Richmond (55 of 115)
47.7% Syracuse (52 of 109)
47.4% UC Irvine (9 of 19)
46.3% UNC (37 of 80)
45.4% Auburn (44 of 97)
45.0% San Diego (9 of 20)
44.4% Loyola Marymount (20 of 45)
43.7% Boston College (31 of 71)
43.5% Wyoming (20 of 46)
43.4% North Carolina St (56 of 129)
43.2% Kansas (19 of 44)
43.0% Wichita St (34 of 79)
42.9% South Carolina (12 of 28)
42.9% UMBC (15 of 35)
42.3% DePaul (22 of 52)
41.4% Florida Atlantic (12 of 29)
41.4% Middle Tennessee St (29 of 70)
41.4% UConn (29 of 70)
41.3% UCLA (50 of 121)
41.0% Michigan (41 of 100)

As for schools that convert a low percentage of potential upset points, there are dozens of schools yet to score an upset. Keep in mind the majority of these schools are lower seeds that have yet to be favored to win. Having said that, here are the ones with the most potential upset points without a win:

130 Iona
110 Texas Southern
99 North Carolina A&T
97 Robert Morris
89 Colgate
82 Belmont
75 Mt St Mary’s
72 Holy Cross
71 Akron
71 S Dakota St
71 UNC Asheville
70 Louisiana Monroe

Among schools that have at some point pulled off an upset, these convert the lowest percentage of upset points:

1.5% UCSB (1 actual upset point of 67 potential upset points)
1.6% UCF (1 of 63)
3.1% Charlotte (1 of 32)
3.7% St Joseph’s (1 of 27)
3.8% Penn (5 of 131)
4.3% Virginia Tech (1 of 23)
4.7% Montana (7 of 148)
4.8% Fresno St (1 of 21)
4.8% Northwestern (1 of 21)
4.9% Winthrop (5 of 103)

For the schools upset the least, here are the ones who have never lost to a lower seed, ordered by potential negative upset points:

56 Butler
21 Arizona St
13 Southern Illinois
10 Rhode Island
7 Louisiana Tech and Northern Iowa
6 UTEP and Western Kentucky
5 Buffalo
4 Northwestern
3 Wofford
2 George Washington
1 Eastern Michigan, George Mason, Old Dominion, and Pacific

Among schools that have been upset at least once, these ten have the lowest percentage of negative upset points:

5.3% Houston (13 negative upset points of 221 potential negative upset points)
5.9% Arkansas (13 of 221)
8.5% Oregon (10 of 118)
9.1% Navy (1 of 11)
9.1% UNC (65 of 717)
10.8% UConn (45 of 417)
10.9% Maryland (27 of 248)
11.5% Duke (101 of 880)
11.9% Kentucky (72 of 604)
11.9% Texas A&M (7 of 59)

At the other extreme, these are all the schools that have never won a game in which they were favored, in order of negative upset points:

18 Nebraska
15 Oregon St
7 Drake
3 Wyoming
1 Boise St, Charleston, Evansville, Green Bay, Jacksonville, South Florida

And these schools have lost more than 50% of their potential negative upset points:

90.6% South Carolina (29 negative upset points of 32 potential negative upset points)
83.3% Colorado St (5 of 6)
83.3% Dayton (15 of 18)
78.6% Penn St (11 of 14)
75.0% Richmond (3 of 4)
71.4% New Mexico St (10 of 14)
70.0% Nevada (14 of 20)
60.7% Charlotte (17 of 28)
58.5% Georgia (24 of 41)
56.1% Clemson (37 of 66)
54.8% Cal (23 of 42)
51.9% Wichita St (27 of 52)

With all this upset data, which schools stand out either for good, bad, or both? On the good side, as an example, Kentucky converts 50% of their potential positive upset points, and is also one of the best at avoiding upsets (losing only 11.9% of potential negative upset points). Two other schools that are at both of these good extremes are UNC (46.3% positive and 9.1% negative) and UConn (41.4% positive and 10.8% negative). Butler also stands out for having by far the most potential negative upset points without losing to a lower seed (while at the same time having a much better than average conversion rate of positive potential upset points at 34.3%).

On the not-so-flattering front, Charlotte stands out for having a very low positive upset percentage of 3.1% while also losing 60.7% of potential negative upset points.

Perhaps the most interesting teams, though, are Penn St and South Carolina. Both teams rank pretty high at pulling off upsets, while also being among the schools that are upset the most. I’m certainly not a proponent of gambling, but one could have made some serious dough had they bet with these teams when they were underdogs, and bet against them when they were favored. All in hindsight, of course.

One other school worth mentioning – Xavier has both the most actual (62) and potential (185) positive upset points, resulting in a respectable upset rate of 33.5%.

Lastly, the following schools accumulated the most upset points in a single tournament:

33 St Peter’s in 2022
32 LSU in 1986
29 Villanova in 1985
27 George Mason in 2006
24 VCU in 2011
24 UCLA in 2021

Well that’s all the information I have to pass along. With the 2026 tournament right around the corner, let the madness begin.