Breaking it down by BPI
March, 17, 2013
By Alok Pattani | ESPN Stats & Information
ESPNSixty-three of the 68 teams in the tournament would have also made the field if they were selected just based on BPI.Let’s use ESPN’s primary basketball team ranking tool, College Basketball Power Index (BPI), to break down the NCAA men’s basketball championship field of 68.
Snubs according to BPI
BPI was not designed to predict which teams would make the field of 68, but rather which teams deserved to make the field (and where they deserved to be seeded) based on merit.
That being said, 63 of the 68 teams -- or more precisely, 32 of the 37 at-large teams -- that made the field would have also made the field if it was selected entirely according to BPI.
The best at-large team that didn't make it was 41st-ranked Stanford. The worst at-large team that did make it was 63rd-ranked Illinois.
Here are the five teams that were left out but would have made the tournament if BPI alone was used to select the tournament -- Stanford (BPI rank: 41), Iowa (43), Baylor (46), Virginia (48), and Maryland (50).
Here are the five teams that made the tournament field but BPI would not have placed them in the tournament based on their ranking among other potential at-large teams –- California (BPI rank: 53), La Salle (54), Temple (56), Villanova (60) and Illinois (63).
Thoughts on the No. 1 seeds
The four No. 1 seeds were all ranked in the top 6 in BPI, including No. 1 overall seed Louisville (also the top team in BPI).
BPI would have had Florida (second in BPI) and Duke (third) ahead of Gonzaga and Kansas, but the ratings among the top teams were relatively close. Gonzaga was only 0.6 behind fourth-ranked Indiana in the BPI.
Also of note is that three of the four No. 1 seeds actually ranked better in BPI than RPI (Kansas being the lone exception).
The No. 1 and 4 teams in RPI -- Miami and Duke -- ended up being No. 2 seeds, and RPI No. 2 New Mexico ended up being a No. 3 seed.
So it seems like RPI wasn't the lone determining factor in choosing those top seeds, as some would suggest.
Overseeded and underseeded
Among the teams actually in the field, Butler and Kansas State were the only two schools to be placed four seeds higher than their BPI would suggest.
BPI had Butler as worthy of a No. 11 seed, but the school was given a No. 6 seed. Kansas State was deemed a No. 8 seed by BPI, but a No. 4 seed by the selection committee.
Pittsburgh is the most underseeded team, according to BPI. The Panthers were tabbed a No. 8 seed by the selection committee, but BPI calculated them worthy of a No. 4 seed.
Six other schools were seeded three spots lower than their BPI rank. They are noted in the chart on the right.
Based on BPI, the West looks like the region with the best teams on average from top to bottom, but the Midwest and South have a better collection of Top 4 seeds.
The average BPI rank of the teams in the West Region was 44.8, followed by the Midwest (49.2), South (54.1) and East (56.9).
The South’s top 4 seeds had an average BPI rank of 8.0, followed by the Midwest (8.3), East (11.8) and West (12.8).
BPI matchup notes
In the Round of 64, there are only two instances of the lower-seeded team having a better BPI than its opponent:
- (10) Colorado (39th in BPI) vs. (7) Illinois (63rd) in the East
- (9) Missouri (23rd in BPI) vs. (8) Colorado State (28th) in the Midwest
In fact, other than Missouri-Colorado State, there are only four other Round of 64 matchups with teams that are within 10 spots of each other in the BPI rankings (three of them in the West):