Beware of the Perennial March Madness Sleepers and Underachievers
Some teams just seem to thrive (or collapse) when it's tourney
They are a bracket-maker’s worst nightmare. They are the teams who surprise us all by going much further — or exiting much earlier — than anyone expected. While there have been a fair share of one-off’s, like #15-seed Florida Gulf Coast slam-dunking their way past #2-seed Georgetown and #7-seed San Diego State en route to the Sweet Sixteen in 2013, some teams are more likely to surprise us than others.1
Two statistics, PAKE and PASE, are great measurements of how a team did in the NCAA Tournament against expectations. PAKE (Performance Against Komputer/KenPom Expectations) is based on KenPom adjusted efficiency numbers and accounts for the true strength of each opponent in each tourney game, independent of a team’s seed. It measures a team’s tourney wins expected based on their adjusted efficiency rating. PASE (Performance Against Seed Expectations), meanwhile, is quite simply that: tourney wins versus the number of wins expected for a given seed.
I was able to gather tournament data dating back to and including the 2008 NCAA Tournament. Below is a table of all teams that appeared in the NCAA Tournament over that time frame with their tourney statistics, plus an aggregate PAKE+PASE statistic that I created, which combines the two metrics.
It is surprisingly one of the traditional blue-bloods, North Carolina, that ranks as the top performer over expectations in the NCAA Tournament. The next few were not super surprising. Tom Izzo has been known to shine in March, leading the MSU Spartans to Final Fours as a 7-seed and 5-seed in that timespan. UConn won National Championships as a 4-seed (2023), 7-seed (2014), and 3-seed (2011), while Butler ran all the way to the National Championship game in back-to-back years as a 5-seed and 8-seed. Interestingly 7 of the top 11 teams in performance above expectations are largely considered among the sports traditional powers — UNC, UConn, Michigan State, Kentucky, Syracuse, UCLA, and Villanova. Looking towards the bottom of the list, it’s Georgetown an Virginia firmly planted in the basement. We already mentioned Georgetown’s infamous loss to FGCU, and second round losses as a 2-seed and 4-seed in 2008 and 2015, respectively, don’t help their cause. I think Virginia’s ranking speaks for itself, as who could forget their historic loss to 16-seed UMBC!
Here is another chart plotting program-level PASE vs. PAKE for 2008-2024 tournaments.
Conference-Level Tournament Performance
Looking at the conference level, what conferences should you back or fade in tourney-time?
Should we maybe re-think the perception that the ACC has been down in recent years? The conference blows the others away in NCAA Tournament performance against expectations. Meanwhile the Mountain West has simply been terrible. Despite putting 37 teams into the tournament since 2008, only 12 have won a game and just five have advanced to the Sweet Sixteen. While the ACC is tracking to have it’s second-worst year per computer metrics (with it’s second-worst being 2023, and well within reach of being surpassed this year), ACC teams that do make it into the big dance might deserve a second-look. Here is that PASE vs. PAKE plot for conferences, similar to the one above.
Hopefully this info helps you all as you fill out your brackets! While March Madness will inevitably surprise us all in some way, maybe we can at least use this to spot an upset or two.
FGCU has gone 0-2 in just two NCAA tournament appearances over the ensuing 11 years.