During the course of NIT Bracketology I’m often asked a number of questions about the process. Using data I pulled from on NIT participants for the past five seasons I took at some trends I think would be most useful for current observers.
How many automatic bids are going to be handed out?
There have been 61 automatic bids to the NIT in the past five seasons, an average of five per. The Ivy League moving to a conference tournament means there is the possibility of one more being given out. Assume that the 8 seeds will all be automatic bids (that’s happened every season) and all of the 7 seeds will be as well (that’s happened every season besides 2012). Last season there were 15 automatic bids to the NIT, the most in this sample. That’s partly because teams like Monmouth, Valparaiso, San Diego State and Saint Mary’s didn’t receive at-large bids to the NCAA Tournament. But those teams being upset in their respective conference tournaments still meant that fewer at-large teams were going to sneak into the bottom of the bracket.
Is there anything else I should know about automatic bids?
About one fifth (12, or 2.5/season) of the teams that received automatic bids were given home games in the first round. The average RPI of automatic bid team is 101 with a 71 percent win percentage, so these teams aren’t pushovers.
Can [insert team here] get in with a .500 record?
It’s within the rules for NIT selection that a team could even get in even with a record under .500, but no team with a record of .500 or below has been given an at-large berth to the NIT in the past five seasons. It’s possible that Pittsburgh or Georgetown could stretch the bounds of reality this season, but I wouldn’t recommend pushing it. Two teams have made it with a record just 1-game above .500: Iowa at 17-16 in 2012 and St. John’s at 16-15 in 2013. Both teams ended up winning their first round games.
Alright, what about if our RPI is high?
That Iowa team in 2012 also has the highest RPI of any at-large team to make the NIT in the past five seasons at 125. The Hawkeyes were 17-16 (8-10 in the Big Ten) and lost to Campbell in non-conference, but also beat four ranked teams during Big Ten play. They are definitely the exception. Beyond that, the highest RPI for at-large consideration is right around 100, which makes sense considering that there are 100 teams between the NCAA Tournament (68) and NIT (32).
Is that true for teams from every conference?
Nope. That RPI guidance is much more useful if you’re from a high-major conference or the Atlantic 10. Otherwise the magic number is closer to 90. The only team from outside those conferences to get an at-large bid with an RPI above 90 in the past five season was Illinois State back in 2012.
Can you give us a sense where we might be seeded based on RPI?
As much as the committee wants to say that more goes into team selection, historically the RPI does a reasonably good job of projecting a team’s seeding. The average RPI for at-large bids by seed the past five seasons demonstrates that concept:
Seed. Average RPI:
Why does it drop for six seeds?
That’s where the committee likes to put strong mid-majors. For instance two mid-majors with RPIs in the 40s (Toledo in 2014 and Princeton in 2016) were given six seeds. Of the 14 at-large bids given to six seeds in the past five seasons 10 were given to mid-majors. UT-Arlington, Memphis and BYU are all teams that might find themselves in that area this season.
Do mid-majors get at-large bids?
Sure, sometimes. Of the 99 at-large bids given out during the sample, 30 went to teams outside the high-major conferences and the Atlantic 10. That means 30 percent of bids go to mid-majors. It’s heavily dependent on the season of course. Forty-five percent of at-large bids went to mid-majors in 2015, but that was certainly an anomaly.
I hope this has been helpful. Please let me know if there are any other questions I can answer about the history of the selection process in the comments.