The National, 3/25/13: Florida Gulf Coast and The History of the Big Upset

Insomnia struck again. As a result I was sitting here pondering the absolutely mind-boggling reality of what Florida Gulf Coast accomplished this weekend, and it got me to thinking. The Eagles have pulled off two huge upsets, in the process removing a division-by-zero error from the matrix of seed-vs-seed results. We're used to huge upsets in the NCAA tournament, of course, but while one big upset is nice, two is really something special.

That led me to go through the brackets for every tournament since the tournament expanded to 64 teams. I had to give myself a criteria to work with, and I decided that a "big" upset is beating a team five or more seed lines above you. That immediately eliminated anyone lower than a six-seed from consideration, and I think that's fair; no team seeded 1-4 can really be said to score a "big upset" in any respect anyway, and a five-seed can really only do so by knocking off a one-seed. Even then, I don't know that it's all that big a deal.

What's really interesting here is the vagaries of time. In the first two years of the 64-team field, eight teams -- four each year -- recorded two or more "big upsets" in one tournament. In the entire decade of the 1990s, only seven teams did so. In all, 36 teams have pulled off the feat (while three more teams still have a chance to accomplish it this year, though circumstances make it almost impossible for any of them to pull it off). That's significantly more than one per year, though not enough to start using terms like "one-and-a-half". However, in 10 of the 29 years since the tournament expanded, nobody's done it, and the really weird part is that despite the fact that every upset actually decreases the chance of more teams doing it... over half the time it happens -- in 11 out of 19 tournaments, to be exact -- more than one team pulls it off.

Before getting to the actual list, a quick-dirty explanation of what each seed has to accomplish to get on the list which will put things in perspective in a way just looking at brackets and doing math won't:

A #15 or #16 seed automatically makes this list by reaching the Sweet Sixteen, which of course has happened exactly once.

A #14 seed makes this list by reaching the Sweet Sixteen UNLESS their second-round game was against the #11 seed; in that event, they would have to reach the Elite Eight by beating the #2 or #7 seed. That has never happened; the only #14 seeds to make this list were Cleveland State in 1986 and UT-Chattanooga in 1997, and they both beat #3 and #6 to reach the Sweet Sixteen.

For a #13 seed to make this list, they merely need to reach the Sweet Sixteen UNLESS their second round game is against the #12 seed; in that event, they need to reach the Elite Eight by beating the #1 or #8 seed, which has never happened.

For a #12 seed to make this list, they need to reach the Sweet Sixteen without running into the #13 seed in the second round. If that happens, they must advance to the Elite Eight by beating the #1 seed, which has never happened, or beat the #8 seed then reach the Final Four by beating the #2, #3, #6, or #7 seed -- which has even had a chance of happening only once, when Missouri reached the Elite Eight in 2002.

(This is the most important one as far as people saying, "Hey, wait a minute, what about...?" There have been a total of twenty #12 seeds in the Sweet Sixteen, but forty percent of them got there by beating #13 seeds. Same with the six #13 seeds to score second-round wins; only three got there on the backs of #5, the other three beat #12 to advance. Putting it simply, and admittedly somewhat harshly in the interest of avoiding "But Western Kentucky!!11!": no, I absolutely did not miss or forget anyone seeded 12th or higher. They're all on the list if they qualified.)

For a #11 seed to make this list, they have to advance to the Sweet Sixteen by beating the #3 seed in the second round. If they had to play the #14 seed in the second round, they can also make the list by beating the #2 seed in the Sweet Sixteen, which has never happened without already making the list in the first two rounds (LSU, George Mason), or by beating the #1, #4, or #5 seeds in the Elite Eight, which has also never happened without having already made the list (also GMU and LSU).

For a #10 seed to make this list, they have to advance to the Elite Eight by beating the #2 seed in the second round and the #3 seed in the Sweet Sixteen, OR, if ONE of those things isn't true, by beating the #1, #4, or #5 seed in the Elite Eight. Otherwise, they have to make the Final Four, which hasn't happened.

For a #9 seed to make this list, they have to advance to the Elite Eight by beating the #1 seed in the second round and the #4 seed in the Sweet Sixteen, OR, if ONE of those things isn't true, by beating the #2 or #3 seed in the Elite Eight. Otherwise, they have to make the Final Four, which hasn't happened.

For a #8 seed to make this list, they have to advance to the Elite Eight by beating the #1 seed in the second round, then advance to the Elite Eight and defeat the #2 or #3 seed to advance to the Final Four. This has happened twice: Villanova in 1985 and Butler in 2011. If ONE of those things is not true, they have to defeat a #1, #2, or #3 seed in the Final Four AND/OR in the National Championship Game. This has happened once, but it was also Villanova in 1985, and they'd already put themselves in the history books by then.

For a #7 seed to make this list, they have to defeat the #2 seed in the second round, then the #1 seed in the Elite Eight. If ONE of those things is not true, they have to defeat a #1 or #2 seed in the Final Four AND/OR in the National Championship Game. This has never happened.

For a #6 seed to make this list, they have to defeat two of the following: the #1 seed in their region, a #1 seed in the Final Four, or a #1 seed in the National Championship Game. This has never happened.

We've already stipulated that it's not possible for any higher seed to make this list.

Having dispensed with that, here's the entire list. Year, seed, team -- followed by the teams they beat, and finally the team they lost to (or, in the case of one of our entries, the fact that they won the tournament). Teams listed in [brackets] are teams that don't count toward "big upset" status, included just for continuity.

1985 (12) Kentucky -- (5) Washington, (4) UNLV, (lost to Saint John's)
1985 (11) Auburn -- (6) Purdue, (3) Kansas, (lost to North Carolina)
1985 (11) Boston College -- (6) Texas Tech, (3) Duke, (lost to Memphis State)
1985 (8) Villanova -- [(9) Dayton], (1) Michigan, [(5) Maryland], (2) North Carolina, (2) Memphis State, (1) Georgetown (NC)
1986 (14) Cleveland State -- (3) Indiana, (6) Saint Joseph's, (lost to Navy)
1986 (12) DePaul -- (5) Virginia, (4) Oklahoma, (lost to Duke)
1986 (11) Louisiana State -- (6) Purdue, (3) Memphis State, (2) Georgia Tech, (1) Kentucky, (lost to Louisville)
1986 (8) Auburn -- [(9) Arizona], (1) Saint John's, (4) UNLV, (lost to Louisville)
1987 (12) Wyoming -- (5) Virginia, (4) UCLA, (lost to UNLV)
1987 (10) Louisiana State -- [(7) Georgia Tech], (2) Temple, (3) DePaul, (lost to Indiana)
1988 (13) Richmond -- (4) Indiana, (5) Georgia Tech, (lost to Temple)
1988 (6) Rhode Island -- (6) Missouri, (3) Syracuse, (lost to Duke)
1989 - NONE
1990 (12) Ball State -- (5) Oregon State, (4) Louisville, (lost to UNLV)
1990 (11) Loyola Marymount -- (6) New Mexico State, (3) Michigan, [(7) Alabama], (lost to UNLV)
1991 - NONE
1992 - NONE
1993 - NONE
1994 (12) Tulsa -- (5) UCLA, (4) Oklahoma State, (lost to Arkansas)
1995 - NONE
1996 (12) Arkansas -- (5) Penn State, (4) Marquette, (lost to Massachusetts)
1997 (14) Tennessee-Chattanooga -- (3) Georgia, (6) Illinois, (lost to Providence)
1998 - NONE
1999 (13) Oklahoma -- (4) Arizona, (5) North Carolina-Charlotte, (lost to Michigan State)
1999 (12) Southwest Missouri State -- (5) Wisconsin, (4) Tennessee, (lost to Duke)
2000 - NONE
2001 (11) Temple - (6) Texas, (3) Florida, (7) Penn State, (lost to Michigan State)
2002 (12) Missouri -- (5) Miami (FL), (4) Ohio State, [(8) UCLA], (lost to Oklahoma)
2002 (11) Southern Illinois -- (6) Texas Tech, (3) Georgia, (lost to Connecticut)
2002 (10) Kent State -- [(7) Oklahoma State], (2) Alabama, (3) Pittsburgh, (lost to Indiana)
2003 (12) Butler -- (5) Mississippi State, (4) Louisville, (lost to Oklahoma)
2004 - NONE
2005 (12) Wisconsin-Milwaukee -- (5) Alabama, (4) Boston College, (lost to Illinois)
2006 (13) Bradley -- (4) Kansas, (5) Pittsburgh, (lost to Memphis)
2006 (11) George Mason -- (6) Michigan State, (3) North Carolina, [(7) Wichita State], (1) Connecticut, (lost to Florida)
2007 - NONE
2008 (10) Davidson -- [(7) Gonzaga], (2) Georgetown, (3) Wisconsin, (lost to Kansas)
2009 - NONE
2010 (12) Cornell -- (5) Temple, (4) Wisconsin, (lost to Kentucky)
2010 (11) Washington -- (6) Marquette, (3) New Mexico, (lost to West Virginia)
2011 (11a) VCU -- [(11b) Southern California], (6) Georgetown, (3) Purdue, [(10) Florida State], (1) Kansas, (lost to Butler)
2011 (11) Marquette - (6) Xavier, (3) Syracuse, (lost to North Carolina)
2011 (8) Butler -- [(9) Old Dominion], (1) Pittsburgh, [(4) Wisconsin], (2) Florida, [(11a) VCU], (lost to Connecticut)
2012 (11) North Carolina State -- (6) San Diego State, (3) Georgetown, (lost to Kansas)
2013 (15) Florida Gulf Coast -- (2) Georgetown, (7) San Diego State, (pending)
2013 (12) Oregon -- (5) Oklahoma State, (4) Saint Louis, (pending)

Three teams still have a chance to make the 2013 list:

(13a) La Salle -- [(13b) Boise State], (4) Kansas State, [(12) Mississippi]... [(9) Wichita State] doesn't help either. They'll have to reach the Final Four.

(9) Wichita State -- [(8) Pittsburgh], (1) Gonzaga... [(13) La Salle] won't help, and they need Ohio State to win Thursday to have any chance of making the list without winning a Final Four game.

(6) Arizona -- What they've done so far is irrelevant; they can only make the list by beating Louisville in the Final Four and Kansas for the National Championship.

Which all brings us to the question: is Florida Gulf Coast the greatest Cinderella of All Time? As far as the first weekend is concerned, obviously so. If we give each team points for each upset based on the difference between their own seed and the seed of their victim, we can rank this. FGCU has so far accumulated 21 "seed points" in upsets -- 13 for Georgetown, 8 for San Diego State, and it goes without saying that nobody's done so well in the first two rounds. Only Cleveland State and Tennessee-Chattanooga come close, racking up 19 points apiece.

They've still got a ways to go, though. VCU was good for 24 seed points in 2011, George Mason picked up 27 in 2006, Villanova in 1985 accumulated 29, and Louisiana State is the reigning champion, having racked up a whopping 32 points as an 11 seed in 1986 by blowing through #6, #3, #2, and #1 on the way to an unlikely Final Four appearance. However, a win over Florida next weekend would be worth another 12 points for the Eagles, and that would indeed move them to the top of the list with 33.

More silliness: I realized while doing this that the #12 seed "boost" applies to more than just the first round. We all know that #12 seeds advance to the second round slightly more often than #11 seeds to, but they reach the Sweet Sixteen at an even higher clip compared to #11 seeds. The #11s are only 15-24 in second-round games (including a 3-0 mark against #14 seeds), but the #12 seeds are an amazing 20-21 in second-round contests, winning exactly 40% of those games even when they're facing the #4 seed, never mind the #13s.

Having figured that out, I decided to track the percentage chance for each seed to reach the Sweet Sixteen. It'll blow your mind.

16: 00.00%
15: 00.86%
14: 01.95%
13: 05.17%
12: 17.24%
11: 12.93%
10: 18.10%
09: 04.31%
08: 08.62%
07: 16.38%
06: 33.62%
05: 33.62%
04: 43.97%
03: 51.72%
02: 64.66%
01: 87.07%

Obviously, the big drop for the 8-9 seeds is because they play the #1 seed in the second round, but you'll note two things. #10 seeds are more likely to reach the Sweet Sixteen than #7 seeds even though #7 seeds are more likely to beat #10 seeds in the first round. That's because #10 seeds are 17-25 against #2 seeds in the second round, while #7 seeds are a paltry 17-50.

That preternatural performance against #2 seeds is why #10 seeds reach the Sweet Sixteen more often than #12 seeds, but those two seeds advance more often than any other seed outside the top six, and why it really is vitally important to get into the top four seeds if you've got serious intentions of winning a national championship. Otherwise, your chances of even making it to the second weekend are just not that good at all.