Monday, December 2, 2024

NIT Bracketology, Dec. 2

The season has started! We're not even a month in, but a lot has changed. This is my updated projection. The biggest note is that I think the College Basketball Crown is going to have enough decent teams to fill their bracket, so it's possible some of the conferences involved might still send teams to the NIT. (We'll see.) In this projection that happens with Northwestern and TCU taking the Big Ten and Big 12 automatic bids respectively.

Also, we really need conference play to start so that the ACC, SEC, and Atlantic 10 in particular can start getting sorted out. It was like putting Jenga blocks together at the end of this bracket. (We could really use some additional West Coast teams stepping up as well.)

The only non-protected team in a top 4 seed in the bracket therefore is Boise St. (because the Big East doesn't have a team). While I have the Broncos projected behind San Diego St. in the conference standings, they were by far the highest in the at-large pecking order.

NIT Bracket

1. San Diego St. (MWC)
U. Oregon St. (WCC)
4. UC San Diego (BW)
U. St. Bonaventure (A10)
3. San Francisco (WCC)
U. Colorado St. (MWC)
2. SMU (ACC)
U. Saint Joseph's

1. Northwestern (B1G)
U. Loyola Chicago (A10)
4. Bradley (MVC)
U. Wichita St. (AAC)
3. Missouri (SEC)
U. Washington St. (WCC)
2. Florida St. (ACC)
U. George Mason (A10)

1. Oklahoma (SEC)
U. Rhode Island (A10)
4. Louisiana Tech (CUSA)
U. Wake Forest (ACC)
3. VCU (A10)
U. Vanderbilt (SEC)
2. NC State (ACC)
U. Florida Atlantic (AAC)

1. Boise St. (MWC)
U. Santa Clara (WCC)
4. TCU (B12)
U. Notre Dame (ACC)
3. North Texas (AAC)
U. UNLV (MWC)
2. LSU (SEC)
U. Murray St. (MVC)

College Basketball Crown

Butler (BE), Villanova (BE), Xavier (BE), Nebraska (B1G), Iowa (B1G), Rutgers (B1G), West Virginia (B12), Arizona St. (B12), Utah (B12), Washington (B1G), Georgetown (BE), Providence (BE), DePaul (BE), Kansas St. (B12), UCF (B12), Colorado (B12)

Wednesday, October 30, 2024

A New Preseason NIT Projection

I put out an offseason NIT bracket projection earlier this offseason, but that was before the news that the NCAA was changing the NIT selection process for 2024-25. Given those changes it only makes sense to re-release my preseason NIT bracket projections!

This projection still includes sending all of the Big East, Big 12 and Big Ten teams that appear to be in good shape for the postseason to the College Basketball Crown. This is despite the fact that those conferences will clearly be in Ken Pomeroy's top 12 and earn one of the preferred seeds in the NIT. (I wish I could be a fly on the wall when an Athletic Director has to explain to Chris Collins that they're giving up a home game to fly to Las Vegas for a tournament in a few weeks instead.)

I used KenPom's current conference rankings to determine the automatic bids. The at-large teams were determined by the preseason composite ratings I had calculated earlier that mash up KenPom, T-Rank and Evan Miya. While this doesn't match the NIT process exactly, it's a good approximation in the offseason.

The results are below. The only manual adjustment I made really was excluding Virginia from the bracket entirely. The Cavaliers have been rocked by the sudden retirement of Tony Bennett and it's unclear if UVA will actually be able to live up to its mediocre preseason expectations.

The other big change? Home games for Seattle (!), UAB, Louisiana Tech, San Francisco, and Bradley – These all come through being the projected second best team in a top 12 KenPom conference. (There are no automatic bids for projected conference champs with 125 rating or higher because I don't project top seeds losing in a conference tournament until it happens.)

Technically every 5-8 seed is unseeded and I could bump them around to get better regional matchups and such, but I want to give readers some sense of where a team stands in the pecking order prior to the season starting. (I might change the format of these posts in the future to help explain this fact.)

Projected NIT Bracket:

1. Oklahoma (SEC 1)
8. Loyola Chicago
4. Boise St. (MWC 1)
5. Notre Dame
3. LSU
6. North Texas
2. SMU
7. Utah St.

1. Georgia (SEC 2)
8. Syracuse
4. San Francisco (WCC 1)
5. San Diego St.
3. Dayton (A10 1)
6. UNLV
2. Louisville (ACC 3)
7. Western Kentucky

1. Miami FL (ACC 1)
8. Vanderbilt
4. Seattle (WAC 1)
5. Nevada
3. UAB (AAC 1)
6. Virginia Tech
2. Missouri (SEC 3)
7. George Mason

1. N.C. State (ACC 2)
8. Saint Joseph's
4. Bradley (MVC 1)
5. Colorado St.
3. Louisiana Tech (CUSA 1)
6. Georgia Tech
2. South Carolina
7. Washington St.

Other Teams With Automatic Bids (But College Basketball Crown): Northwestern (Big Ten 1), Providence (Big East 1), TCU (Big 12 1)

Other Possibilities: Wichita St., Florida St., Santa Clara, Stanford, South Florida, Massachusetts, Yale, Florida Atlantic, Sam Houston St.

Tuesday, October 22, 2024

The NIT Selection Criteria Is Changing Again (For The Better!)

The NCAA is changing the NIT Selection criteria again, but this time it's for the better!

The changes announced today include 16 "exempt" teams. Those teams are:

  1. The top two teams from the SEC not selected for the NCAA Tournament
  2. The top two teams from the ACC not selected for the NCAA Tournament
  3. The top team not selected from the NCAA championship from the top 12 conferences (based on the Ken Pomeroy Rating) — and yes, the SEC and ACC do get to double dip here

The "top" team doesn't mean "top" in the standings though, which could have some interesting consequences. "The top teams from each conference will be determined based on the average of the teams’ ESPN Basketball Power Index (BPI), Kevin Pauga Index (KPI), NCAA Evaluation Tool (NET), Ken Pomeroy Rating (KenPom), Strength of Record (SOR), Torvik ranking and Wins Above Bubble (WAB) ranking. All rankings are already included in the “Team Sheets” used by the NIT Committee during the selection process."

The bonus for these 16 teams is that by being "exempt" they're guaranteed to host a first-round game.

The NCAA also announced that any "regular-season conference champions that are not otherwise selected to the NCAA championship can earn an automatic bid to the NIT as along as that regular-season champion has an average of 125 or better across BPI, KPI, NET, KenPom, SOR, Torvik and WAB rankings".

This will definitely shake up the field moving forward and appears to be a direct response to the College Basketball Crown.

I used Torvik WAB from last season to fill in for WAB, and if this system was in place last season I believe these would've been the "exempt" teams:

  • Pittsburgh (ACC)
  • Wake Forest (ACC)
  • LSU (SEC)
  • Ole Miss (SEC)
  • Oklahoma (Big 12)
  • St. John's (Big East)
  • Ohio St. (Big Ten)
  • Georgia (SEC)
  • Virginia Tech (ACC)
  • Utah (Pac-12)
  • UNLV (Mountain West)
  • Richmond (Atlantic 10)
  • Memphis (AAC)
  • Indiana St. (Missouri Valley)
  • San Francisco (WCC)
  • Western Carolina (Southern Conference)

These teams would also qualify under the regular-season champion exception:

  • South Florida
  • Princeton
  • UC Irvine
  • Appalachian St.
  • High Point

Toledo would have just missed at 129. Quinnipiac would've been a bit further off at 149.

Why is Memphis the automatic qualifier from the AAC over South Florida? A four point difference in their average rating, which mostly comes down to how KenPom viewed the two programs last season.

Iowa, Seton Hall, Bradley and VCU would've made excellent at-large teams based on their rankings in all of these metrics.

In addition, the race for the "exempt" bid in conferences like the Atlantic 10, Big Ten (sorta), and AAC would have been fascinating to watch down the stretch. Most seasons the Mountain West, Missouri Valley and WCC could also fall into that group. It will give teams a lot of incentives.

If this system was in place for Selection Sunday 2024 it would've given UNLV, Richmond and San Francisco home games that they didn't have last season. In addition, Western Carolina, which missed the bracket entirely, would've been given a home game. Memphis would've also been invited as an exempt team, but they were on record as planning to decline an NIT invite.

In addition to Western Carolina, High Point would also have been invited to the NIT last season.

This would have unfortunately most likely taken a home game away from Princeton, unless the number of exempt teams declining allowed the Tigers to host at Jadwin. (This is essentially what happened last season anyways.)

Overall, these changes look like they could represent a positive step forward for the NIT. It will create better representation across college basketball and lead to some interesting teams receiving invites. The College Basketball Crown still hangs over this operation, with the likelihood that the best teams from the Big East, Big Ten and Big 12 will head to Vegas, but the NIT should have a very competitive field.

Thursday, October 17, 2024

Preseason NIT Bracketology

The 2024-25 college basketball season is coming up fast! KenPom, T-Rank and EvanMiya have all released rankings of all 364 teams competing in Division I this season. I used those to put together this projected NIT bracket.

A few initial thoughts:
  1. No one has any idea what to do with the College Basketball Crown. The new tournament is supposed to kick off in the spring of 2025. The original release said that it would include teams from the Big Ten, Big East and Big 12 "along with additional at-large participants". I have no idea how to interpret that statement. Given that the SEC and ACC decided to stick with the NIT for now I think we can safely assume that top teams from those leagues aren't going to play in an upstart new tournament. This projection only includes teams from those three leagues in my CBC tournament.
  2. The ratings aren't in sync on a number of high profile teams. Rutgers is one of the teams I am projecting to the CBC for now. The Scarlet Knights have high-profile 5-star recruits and a team ranked in the initial AP Top 25. KenPom and EvanMiya aren't buying it though. Those two systems rate Steve Pikiell's team 63rd and 72nd respectively. T-Rank is much higher at 19th. Splitting the difference still leaves Rutgers as a bubble team for the NCAA Tournament. We've seen teams with that type of talent miss the NCAA Tournament entirely, so I don't think this is that controversial of a projection. Other teams that are in a similar boat near the bubble are Miami FL and Dayton.
  3. The lack of Big Ten, Big East and Big 12 teams might create bracketing issues for the NIT. If teams from those leagues do opt out of the NIT, there will definitely be some issues bracketing the tournament. The reason is because it will likely result in a flood of ACC teams. Here I had to break procedure because I ended up with too many ACC teams for them to fit nicely into brackets. That could be a big challenge down the line.
  4. I am assuming that automatic NIT bids still exist for the ACC and SEC. This doesn't really matter once you remove the other power conferences, but I'm assuming that those schools will still be guaranteed home games. The dissolution of the PAC-12 (at least for this season) means that this would only really affect four teams anyways. (The "auto bids" are on the 1 and 2 seed lines in this projection.)
Projected NIT Bracket:

1. Miami FL
8. Loyola Chicago
4. Utah St.
5. San Francisco
3. San Diego St.
6. Notre Dame
2. South Carolina
7. Syracuse

1. Dayton
8. Wichita St.
4. Vanderbilt
5. UAB
3. SMU
6. Colorado St.
2. LSU
7. Georgia Tech

1. Oklahoma
8. Saint Louis
4. Virginia
5. George Mason
3. Boise St.
6. Washington St.
2. Missouri
7. Bradley

1. Georgia
8. Virginia Tech
4. Nevada
5. Saint Joseph's
3. Louisville
6. North Texas
2. NC State
7. UNLV

College Basketball Crown:

1. Northwestern
2. Rutgers
3. USC
4. Nebraska
5. TCU
6. Providence
7. Penn St.
8. UCF
9. Minnesota
10. Utah
11. Washington
12. Arizona St.
13. Butler
14. West Virginia
15. Seton Hall
16. Georgetown

Other Possibilities: Florida St., Santa Clara, Stanford, South Florida, Massachusetts, Louisiana Tech, Yale, Florida Atlantic, Western Kentucky, Sam Houston St.

If you count, you'll see that the CBC bracket includes seven teams from the Big Ten, five from the Big 12 and four from the Big East. That seems about the right breakdown given how many Big 12 teams are likely to make the NCAA Tournament. Four of the CBC participants are projected to be under .500 if you look at projected regular season records on T-Rank right now (Arizona St., WVU, Seton Hall and Georgetown).

Is this really what's going to happen? Your guess is as good as mine right now. I'll be keeping an ear out for any news I hear about the college basketball's rapidly shifting postseason as we try and venture into this new world.

Thursday, July 11, 2024

WAB Is Going To Be On NCAA Team Sheets!

The NCAA Team Sheets are going to have Wins Above Bubble! That was one of the excellent pieces of news that came out of the meetings of the NCAA's Men's Basketball Committee the past few days. Today it was announced that T-Rank (from Bart Torvik's excellent site) and Wins Above Bubble (apparently the NCAA's NET-based version) are going to be added to the NCAA Team Sheets, which will make them much more accessible for the NCAA's Selection Committee.

This is a big deal because Wins Above Bubble (WAB) often paints a different picture for teams that play a non-traditional power conference schedule (see mid-majors). It's especially valuable because it's so difficult for elite mid-majors to get certain types of games (home games against power conference teams for instance). Don't believe me? Here are the top teams to miss the NCAA Tournament according to T-Rank WAB the past five seasons (well minus 2020 because there was no tournament):
  • 2024: Indiana St. (28th), Princeton (42nd), Seton Hall (43rd)
  • 2023: Oklahoma St. (51st), Sam Houston St. (52nd), Santa Clara (53rd)
  • 2022: SMU (38th), Wake Forest (44th), Texas A&M (47th)
  • 2021: Arizona (37th), Louisville (39th), SMU (50th)
  • 2019: UNC Greensboro (35th), North Carolina St. (37th), TCU (42nd)
Last season's Indiana St. team is the prime example of the types of teams a WAB metric can benefit. The Sycamores were 28-6 when they lost to Drake in the Missouri Valley Conference tournament final. Three of those losses were to Alabama, Michigan St. and Drake on the road — games most bubble teams would be expected to lose. WAB properly contextualizes those losses and also just how difficult it is to go 10-0 in Q3 games, which is what ISU did while playing its MVC regular season schedule.

UNC Greensboro in 2019 is another perfect example. The Spartans under Wes Miller were 28-6 on Selection Sunday and eventually ended up with an NIT No. 1 seed despite being ranked 100th in T-Rank and 87th in KenPom. Why? Because none of those six losses came against a team ranked lower than 62nd (Furman) on T-Rank that season. While a 2-6 record against Quad 1 doesn't look that impressive it matters than five of those games were against Quad 1-A and that they went 24-0 against Q2 and above. Wins Above Bubble is able to normalize those types of resumes.

Hopefully this will lead to a bigger tent in at-large selection conversations. Teams that might not have the "traditional" resume can at least get a shot. (Note: For much of last season this would've included a team like Syracuse.) It will also help better sort out resumes as conferences continue to balloon and the traditional home-and-homes disappear from conference play. (Look how many ACC teams appear on that list above.)

It's not just mid-majors that will benefit from this change, but it's a good step towards equity.

Tuesday, May 14, 2024

How ESPN Might Save The NIT

ESPN’s relationships with the NCAA, SEC and ACC might end up saving the NIT. At least that is my initial takeaway after reading a memo that the SEC’s Charlie Hussey, deputy commissioner/COO, and Garth Glissman, associate commissioner for men's basketball sent to SEC Athletic Directors on March 13, 2024.

The memo was sent in preparation for a meeting of SEC athletic directors on March 20. It outlines the pros and cons of signing on with the new unnamed (at the time) competing postseason tournament for 2025 and beyond. That tournament, announced on April 3 as the College Basketball Crown, will feature 16 non-NCAA Tournament teams in Las Vegas the week after the Elite 8.

I obtained a copy of that memo from the University of Mississippi through a FOIA request related to the NIT. The document was part of a 987-page production of email correspondence related to the NIT during the week of Selection Sunday.

The memo includes a lot of interesting information both about how the SEC’s main office was thinking about the College Basketball Crown and the NIT. Here are a few takeaways:

1. ESPN’s relationship with the NCAA (as the television broadcaster of the NIT) and the SEC and ACC (through their conference networks) is an important talking point.


Under “Option 1 (NIT)” the memo states that: “ESPN is the NIT's broadcast partner through 2032. If the SEC continues to send its top non-NCAA Tournament teams to the NIT, ESPN is willing to broadcast some of the SEC teams' NIT games on the SEC Network. ESPN offered the same arrangement to the ACC and its network.” 
 
This is important because the College Basketball Crown will take place on the networks of FOX (FOX or FS1 according to the memo), which would mean losing games to networks outside of the conferences’ typical television footprint. 
 
In fact, under the “Analysis” headline the memo states: “ESPN Relationship: As part of ESPN's recently signed agreement with the NCAA to expand its coverage of NCAA championship events, ESPN acquired the rights to broadcast the NIT through 2032. The SEC's relationship with ESPN could be adversely impacted if the SEC opts to participate in the FOX Event and therefore diminishes the value of ESPN's rights to broadcast the NIT. Additionally, to incentivize SEC teams' continued participation in the NIT, the NIT and ESPN offered to broadcast NIT games involving SEC teams on the SEC Network.”

2. The decision to join the College Basketball Crown is not a small one.


The memo states that the competition was looking for a 5-year commitment from each conference. Conversely, the NIT was willing to continue with an annual commitment.

3. The NIT was prepared to offer incentives.


Last season the NIT added protected bids for the power conference teams in order to stave off the momentum of an event like the College Basketball Crown. At the same time, the NIT also apparently raised the honorarium it pays to host institutions from 15% to 20%. In order to remain competitive with the College Basketball Crown, the NIT was willing to offer power conference teams further incentives. The memos states: “To further incentivize the SEC to participate in the NIT in 2025 and beyond, the NIT is potentially willing to make the following accommodations to an SEC team participating in NIT: (i) increase the travel party size above 25, (ii) increase the per diem rate, (iii) increase the honorarium paid to host institutions, (iv) increase the travel stipend, (v) provide host team sponsors with a percentage of rotating signage, and (vi) provide charter flight accommodations to the site of the NIT semifinals and championship.”

Those incentives make sense given that the College Basketball Crown was apparently offering: “Covered expenses for each team's travel party (up to 30 individuals) include commercial airfare, hotel rooms, ground transportation, and meals.” The NIT’s incentives make the decision quite competitive for teams from conferences like the SEC that could potentially host multiple games.

4. It does appear the College Basketball Crown’s is a field of “power conference” teams.


The memo states that, “The FOX Event intends to distinguish itself from the NIT by fielding teams largely from "power" conferences (rather than teams from a broader group of NCAA Division I conferences) and providing a premium experience for student-athletes, coaches, and fans.” How the tournament is able to fill out a competitive 16-team field drawing from currently three power conferences remains to be seen.

Given all of that, why wouldn’t the SEC stick with the NIT? Well the “Analysis” portion of the memo highlights three key factors:

  1. “Watered Down” NIT: This is mainly a concern about what will happen to the NIT once the College Basketball Crown starts in 2025, but it does include this interesting sentence, “Further, the quality of teams participating in the NIT and its prestige already has diminished over time, culminating in the NIT's semifinal and championship games recently moving away from its longtime home, Madison Square Garden in New York.” It’s interesting that even the SEC’s office feels that moving away from MSG reduced the prestige of the NIT.
  2. Power Conference Opponents: All you need to know here is that it includes the statement that “participating in the FOX Event would enable SEC teams to avoid the "brand damage" associated with losing to a school from a non-power conference in the NIT”. If that’s really the concern
  3. Opportunity for Innovation: Basically, why not try something new?

Neither the SEC nor the ACC was mentioned in the initial press release that came out about the College Basketball Crown, so it appears that both conferences declined to sign 5-year contracts with the event. This of course does not preclude them from doing so in the future, but it does seem like the NIT has some potential ways to mitigate the damage done from this competing secondary postseason tournament.

Know more about the NIT or the College Basketball Crown? Send me tips at jtemplon@gmail.com.

Tuesday, April 9, 2024

On Dan Hurley and Close Games

Dan Hurley and the UConn Huskies are once again the champions of Division I men's college basketball. UConn is the first school to win back-to-back titles since Florida in 2006 and 2007. The way UConn did it though is also astounding. The Huskies have played 12 NCAA Tournament games in the past two seasons and the closest victory was by 13 points against Miami FL in last season's Final Four. It is as dominant a stretch as could be imagined.

That dominant run is also great for Hurley in particular, because he has historically struggled in close games. But if you never play in them — it doesn't matter.

The chatter about Hurley and close games died down a bit after the Huskies won their first title. This season UConn just didn't play many close games. The Huskies went 2-1 in games decided by four points or fewer. The loss was at Kansas on Dec. 1. (We'll get back to the Jayhawks and Bill Self in a minute.)

Eventually UConn will have to play another close game and what happens then is anyone's guess. Here's a look at Hurley's career coaching record by margin and school. A big thanks to Bart Torvik for the underlying data.
 
SchoolMarginWinsLossesWinning Pct.Pct. Games
Wagner1-47847%25%
Wagner5-89660%25%
Wagner9-1511379%23%
Wagner16+10663%27%
Rhode Island1-4213140%27%
Rhode Island5-8172046%19%
Rhode Island9-15371670%28%
Rhode Island16+371374%26%
UConn1-410628%18%
UConn5-8221559%19%
UConn9-15401080%25%
UConn16+68791%38%

Hurley's teams at UConn have lost 72% of the 1-possession games they've played during his time there. It's almost unbelievable considering how good he's been in every other situation and even moreso considering how good UConn has been against elite competition. These should be 50/50 tossups and the only place that was almost ever true for Hurley was at Wagner.

Maybe though it's just that great coaches are: 1) Less likely to play close games or 2) Worse in them?

To test those theories I looked at five of Hurley's contemporaries: Bill Self at Kansas, John Calipari at Kentucky, Tom Izzo at Michigan St., Mark Few at Gonzaga, and Scott Drew at Baylor. Here are their results at their current (or most recent) school.
 
Coach / SchoolMarginWinsLossesWinning Pct.Pct. Games
Self / Kansas1-4762972%17%
Self / Kansas5-8833471%19%
Self / Kansas9-151312783%26%
Self / Kansas16+2062490%38%
Calipari / Kentucky1-4484552%17%
Calipari / Kentucky5-8743667%21%
Calipari / Kentucky9-151122880%26%
Calipari / Kentucky16+1751493%36%
Izzo / Michigan St.1-4684461%19%
Izzo / Michigan St.5-8734860%20%
Izzo / Michigan St.9-151245071%29%
Izzo / Michigan St.16+1603283%32%
Few / Gonzaga1-4512964%14%
Few / Gonzaga5-8492566%13%
Few / Gonzaga9-151122184%23%
Few / Gonzaga16+2831695%51%
Drew / Baylor1-4594756%19%
Drew / Baylor5-8725059%21%
Drew / Baylor9-151155368%29%
Drew / Baylor16+1502586%31%

I probably didn't need to make this table for it to be obvious, but Bill Self is an absolutely amazing in-game coach. The fact that his Kansas teams have won 72% of their games decided by four points or fewer is just unbelievable.

Most of these coaches though: Have coached in a similar percentage of close games (with the exception of Few) and have won more of those games than Hurley. Few is probably an outlier in terms of the percentage of close games because Gonzaga plays their conference season against the West Coast Conference, which doesn't lend itself to a ton of high-level competition. Self is by far the outlier in terms of close-game winning percentage, but the lowest otherwise is Calipari at 52% (Drew isn't far behind at 56%). 

While it hasn't mattered during the NCAA Tournament in either of the past two seasons. Hurley's close game results are still fascinating. Does anyone have a good explanation? I'm all ears.

NIT Bracketology, Dec. 2

The season has started! We're not even a month in, but a lot has changed. This is my updated projection. The biggest note is that I thin...