The 2024 Best Available Bracket

What would the strongest possible field of 68 have looked like?

Last year, I made a video essay detailing the stark benefits that one-bid leagues would gain from adopting the double ladder conference tournament format.

If you haven’t watched that video, please do that; it’s one of the best things I’ve ever created. But if you don’t have that kind of time on your hands and also don’t know what I mean when I say “double ladder”: it’s the format in the thumbnail—the one the WCC uses, where the top two seeds get multiple byes all the way to the semifinals.

The WCC shouldn’t be using this format because there’s no good reason for them to favor their top two teams. They’re pretty much never a one-bid league, and even in the rare case they are, Gonzaga is always good enough to earn an at-large bid, making their cakewalks to the WCC Tournament title pretty much meaningless, bragging rights aside. The double ladder format should be used by conferences whose best teams absolutely need the autobid. Take the WAC, for instance.

Screen captured from RealGM

Grand Canyon looked like an at-large contender for a while, but consecutive losses at Tarleton and Abilene Christian in late February doomed them to the wrong side of the bubble, with no quality opponents remaining on their schedule against whom they could recover their status. The Lopes 100% needed the autobid, and…they got it. The double ladder afforded them the luxury of only needing to win two games, and they won two games.

I go into this in the video essay, but the double ladder is the best possible option for leagues like the WAC because it allows most or all league members to participate in the conference tournament, which is fun, while still heavily favoring the teams most likely to make the most noise (and money) in the NCAA Tournament, which makes for a more compelling postseason.

As a companion piece, I created the original Best Available Bracket, detailing how last year’s NCAA Tournament bracket would change if every one-bid league was represented by its worthiest member.

Today, I’m back to rearrange some more bids for 2024…but not just for one-bid leagues. This year, several mediocre teams from multi-bid leagues won their way onto the bracket, taking the places of teams that were obviously better than them. In addition to reassigning mid-major and low major autobids, I’m booting the bid-stealers and replacing them with the teams they locked out.

The goal is to find the best possible field of 68. Let’s get started.

Clarifications

Before I dig into this, I want to clarify three things:

1. My intent is not to rain on anyone’s parade. If your favorite mid-major made the NCAA Tournament and I say in this piece that they “shouldn’t have”, please understand that I don’t mean that as a personal slight against the team. I just think they had a less impressive regular season than the team I picked over them.

2. I love conference tournaments. An easy takeaway from reading this piece might be that I despise them and think we should just give the autobid to the regular season champion, but I don’t agree with that at all. For one, it’d be much less fun, but more pertinently, the regular season champ isn’t always the worthiest representative. Out of 22 one-bid leagues this year, I think the conference tournament 1-seed should not have gotten the autobid in ten of them.

3. Most importantly, this piece is scientifically crafted to age like milk. It is a near certainty that at least one team I remove from the field in this exercise will pull off a major upset. I just want to say up front that the results a team achieves after Selection Sunday change nothing about whether they were the most deserving team in the regular season. Suggesting otherwise is an obvious logical fallacy. Additionally, I’m still rooting for most of the upsets regardless of whether I included the team in my field (just please don’t touch UNC).

With that out of the way, let’s get into the nitty gritty.

Methodology

Choosing a “worthiest” representative is kind of arbitrary. You could award it based on any number of résumé considerations, on efficiency metrics, or even just on their position in the conference standings.

My approach last year was to weight résumé and metrics pretty much evenly, and that’s how I’m doing this again today. I’m going to take a snapshot of each league at the time their conference tournament began and pick the team I think had the most impressive combination of high metrics and quality wins in the regular season. Essentially, I’m looking for the team I think would be the scariest for a high-seeded opponent to face in the Round of 64.

As an example, let’s look at a league where different approaches will give you vastly different results: the ASUN.

The ASUN’s highest ranked team on KenPom when the conference tournament began was actually their 3-seed, Lipscomb, who sat at #162, well above 2-seed Stetson (#217) and 1-seed Eastern Kentucky (#200).

However, the ASUN went 1-27 in Quad 1 this season and the 1 belonged to Stetson, who won at UCF. Lipscomb won a Quad 2 game (at Florida State) but they also got a shot at UCF in Orlando and lost by 15.

If you want to argue that winning the conference means something, you could consider Eastern Kentucky, but I’m dismissing them out of hand. They don’t have the best résumé or the best metrics.

If you think metrics should be the deciding factor, you give this bid to Lipscomb, and that would be a reasonable selection to me.

But I think Stetson, the actual autobid winner, is the play here. Their body of work scares me more than Lipscomb’s does, and the difference in metrics between the two teams isn’t quite wide enough to make up for it. I wouldn’t pick Stetson here if they were #317 on KenPom instead of #217, but if they were that low, they wouldn’t have such a good résumé in the first place.

Does this all make sense? I hope so, cuz I’m about to do the same thing for every other one-bid league—that is, every league that received only one NCAA Tournament bid and had that team get a 10-seed or worse.¹

Autobid reassignments

Let’s start with the ones I think got it right.

Bullseyes

Missouri Valley: (2) Drake

Anyone paying close enough attention to bracketology over the past couple weeks should not be surprised that I think Drake, not (1) Indiana State, was the most deserving team in the Missouri Valley.

At the start of the conference tournament, Indiana State was a good amount higher in the NET (#30 to Drake’s #48), but that’s pretty much the only thing the Sycamores had going for them. In KenPom, a much better and more highly respected efficiency metric, Indiana State was #45 and Drake was #54; the difference was negligible.²

Indiana State’s lone Quad 1 win was at Bradley. Drake did that, beat Nevada on a neutral, and also beat Indiana State at home, giving them three Quad 1 wins to just one loss. Drake’s two Quad 3 losses and Indiana State’s one Quad 4 loss are pretty much a wash.

Washington State is gonna be real nervous about Drake on Thursday. If you ask me whether I think they’d be more nervous if Indiana State was in their place, my answer would be no.

SoCon: (1) Samford

Samford won the SoCon by three games and was its only team to have a winning record in Quad 2 (where they went undefeated: 2-0). They were also the league’s best team in efficiency metrics by a mile. No contest.

WAC: (1) Grand Canyon

Another gimme. Grand Canyon was the only team in the WAC who came anywhere close to competing for an at-large. The league’s second-place finisher, Tarleton, wasn’t even eligible for the NCAA Tournament because they’re still transitioning to Division I. No further explanation necessary.

Horizon: (1) Oakland

(2) Youngstown State was higher on KenPom when the conference tournament started (#129 to #141), but Oakland’s body of work clears easily. They started the season by taking Ohio State and Illinois both down to the wire on the road, then eventually completed a Quad 1 upset at Xavier. They also had a Quad 2 win: at Youngstown State.

Youngstown State went winless in the top two quadrants and had two Quad 4 losses (Oakland had none).

America East: (1) Vermont

This was kind of a down year for Vermont, and they still won the America East by four games and were its only team to go .500 in the top two quadrants (at 0-2 in Quad 1 but 2-0 in Quad 2).

(3) Bryant earned the league’s only Quad 1 win, at Florida Atlantic, but when the conference tournament started, they were #179 on KenPom to Vermont’s #99. Bryant also went 0-3 in Quad 2 and 4-5 in Quad 3 (Vermont went 9-3).

Summit: (1) South Dakota State

The leader on KenPom at conference tournament time was actually (4) St. Thomas, but they’re ineligible for the NCAA Tournament until 2027 via the transition rule, so they’re out. At #151 on KenPom, South Dakota State was easily the highest eligible team, with (2) Kansas City coming the next closest at #209.

There’s not a single good résumé in this conference, as the Summit League went 0-16 in Quad 1 and their only Quad 2 win came when (5) North Dakota State won at South Dakota State. The Jackrabbits were the only team in the league with a winning record in Quad 3. That’s the bar here.

MAC: (2) Akron

Akron lost the MAC by one game to (1) Toledo, but they had the better résumé and metrics. Akron won two Quad 2 games (at South Dakota State and vs. Bradley) to Toledo’s one (at Oakland), also holding a moderate lead on KenPom at conference tournament time: #117 to #133.

ASUN: (2) Stetson

I went into this one in the Methodology section. No point in repeating myself.

Ohio Valley: (3) Morehead State

This year’s Ohio Valley was the ultra rare case of a double ladder hurting the league’s best team. Morehead State tied with (1) Little Rock and (2) UT Martin atop the standings at 14-4, but lost the three-way head-to-head tiebreaker and were relegated to the second rung of the ladder, needing to win three games to reach March Madness. Good thing they did!

When that conference tournament began, Morehead State was #123 on KenPom and Little Rock was next in line all the way down at #192. Nobody in this league won a game in the top two quadrants except for (8) Southern Indiana, whose Quad 2 win came from—you guessed it—winning at Morehead State.

The metrics keep this bid in Morehead.

Southland: (1) McNeese

This one almost goes without saying. McNeese finished the regular season #66 on KenPom, over 100 spots ahead of (2) Texas A&M–Corpus Christi at #170. They earned the conference’s only Quad 1 win (at VCU) and two of its five Quad 2 wins (at UAB and at Michigan)—two of four if you discount one of the other three being Southeastern’s win over McNeese themselves.

Patriot: (1) Colgate

This one’s also a no-doubter. Colgate won this league by six games and began the conference tournament at #142 on KenPom, lightyears ahead of the second highest team, (6) Lehigh at #270. Nobody in the Patriot League won a Quad 1 game, but Colgate tied for the league lead with one Quad 2 win (at Vermont) and was the only team to win multiple Quad 3 games.

That’s 11 autobids I think are “correct”, leaving 11 more I think aren’t.

Missed the Mark

Ivy: (1) Princeton over (2) Yale

Despite being the 12th-ranked conference in both KenPom and the NET, the Ivy League didn’t win a single Quad 1 game this season, going a collective 0-16 in their toughest games of the year. That gives Princeton the best Quad 1 record in the league by default, as they were the only Ivy League team not to play a single game in the quadrant.

That’s hardly their fault, of course. The Tigers certainly tried to give themselves opportunities for good wins. They scheduled their first five games away from home (one neutral and four true road) and won all five, including at Hofstra and Duquesne. Both of those games were Quad 1 at one point, but both teams regressed enough throughout the season that they ended up in Quad 2.

Yale didn’t really do anything like that. Their best Quad 2 win was, well, their home game against Princeton. They also lost three more lower-quadrant games than Princeton and had somewhat worse metrics; Princeton was #56 on KenPom at Ivy Madness time and Yale was #87.

C-USA: (2) Louisiana Tech over (3) Western Kentucky

Louisiana Tech’s season was largely defined by the games they almost won. They played three Quad 1 games and lost all three of them but kept all of them within single digits, including losing by just three points at Grand Canyon. Western Kentucky didn’t play a Quad 1 game all year and their only Quad 2 win of the year was…at Louisiana Tech.

(1) Sam Houston is basically a non-starter on account of having a 1-3 Quad 2 record and three Quad 4 losses; Louisiana Tech had a 1-1 Quad 2 record and just one Quad 4 loss.

Add in the stark difference in efficiency metrics—Louisiana Tech was #90 in KenPom at conference tournament time while Sam Houston was #142 and Western Kentucky was #152—and it’s clear who belongs here.

Big West: (1) UC Irvine over (4) Long Beach State

UC Irvine was so far ahead of the pack in efficiency metrics that résumé almost doesn’t matter at all here.

Even if it did, UC Irvine and Long Beach State both went winless in Quad 1 (though UC Irvine did lose at San Diego State by a single point) and both had the same Quad 2 win at USC (bad year for them). Long Beach State had one more Quad 2 win (at Michigan), but they also had two more losses in Quad 3 and five more in Quad 4, while UC Irvine was the only team in the league not to lose multiple Quad 4 games.

I could say all this, or I could just point out that Long Beach State pre-fired their coach before the regular season was even over.

Sun Belt: (1) Appalachian State over (2) James Madison

This was a toss-up for me. You could still argue James Madison here and I wouldn’t really disagree with you.

The only real blemish on James Madison’s résumé is a Quad 3 loss at Southern Miss, while Appalachian State had four Quad 3 losses and a Quad 4 loss before the conference tournament. The wide majority of James Madison’s schedule was against Quad 4, but they didn’t lose a single game in it: 22-0. Regardless of the quality of opponent, that’s not easy to pull off. James Madison was also marginally higher on KenPom (#61 to App State’s #73).

James Madison also had the marquee win at Michigan State, but App State’s big win, vs. Auburn at home, is even better. Plus—and this is the real deciding factor for me—App State swept James Madison this season, home and away.

Put yourself in the shoes of a 5-seed here. Who’s scarier than the team that spent the year dominating bottomfeeders? Try the team that beat them twice.

Big South: (1) High Point over (5) Longwood

The only noteworthy knock against High Point is that they’re the only team in the Big South who didn’t play a Quad 1 game; even so, they won just as many as the rest of the league.

High Point led the conference on KenPom by a mile at conference tournament time; they were #111, (2) UNC Asheville was #153, and Longwood was down at #174. UNC Asheville did notch a neutral-site Quad 2 win over Appalachian State, but that’s not enough to tip the scale.

CAA: (4) UNC Wilmington over (1) Charleston

Big mess at the top of this league.

Charleston won the league but lost their only Quad 1 game of the season, at Florida Atlantic, pretty badly. They went 1-2 in Quad 2 and were #103 on KenPom at conference tournament time.

(2) Drexel went 1-1 in Quad 1 (winning vs. Villanova on a neutral but losing at Princeton), but they also went 0-3 in Quad 2 and 4-5 in Quad 3, so they’re out.

(3) Hofstra went 0-2 in Quad 1, losing two Quad 1A games at Duke and St. John’s. They went 1-4 in Quad 2 and were #118 on KenPom.

UNC Wilmington won their only Quad 1 game, and it was even Quad 1A: at Kentucky. They went 1-3 in Quad 2 and a shaky 5-5 in Quad 3, but were the league’s second highest team on KenPom at #112.

Notably, though, UNC Wilmington’s Quad 2 win is at Charleston, and so is one of their Quad 3 wins; they swept them. Considering this sweep, that UNC Wilmington has the best win in the league by a longshot, and that the metric difference is negligible, I’m giving the nod to UNC Wilmington here. Congrats to Charleston on going 14-0 in Quad 4, though.

Big Sky: (4) Weber State over (5) Montana State

When the conference tournament began, Weber State was a very close second in the league on KenPom, ranking #146 to (1) Eastern Washington’s #141 (and Montana State’s #243).

Weber State had easily the most impressive performance against top teams, taking home the Big Sky’s only Quad 1 win, and a Quad 1A one at that: at Saint Mary’s. A week later, they added a Quad 2 win, beating Yale comfortably on a neutral court. Both of these wins are more impressive (to me, at least) than Montana State’s best win, a low Quad 2 win at Cal.

(2) Northern Colorado and (3) Montana went a combined 0-5 in the top two quadrants.

MAAC: (2) Fairfield over (5) Saint Peter’s

When the MAAC Tournament began, Fairfield held a narrow lead in KenPom over (1) Quinnipiac, #167 to #176. They also earned a Quad 2 win at Yale, whereas Quinnipiac got crushed in their only Quad 2 game of the season, at UMass.

Saint Peter’s was third in the league on KenPom at conference tournament time (#199), but the résumé just wasn’t there: 0-3 in the top two quadrants and 1-6 in Quad 3.

MEAC: (1) Norfolk State over (4) Howard

Norfolk State entered the conference tournament with a marginal KenPom lead over (2) NC Central, #236 to #242, with nobody else close to that pair. Norfolk State gets the nod here on account of having the only win in the top two quadrants across the entire league: a Quad 1 win at VCU.

NEC: (2) Merrimack over (6) Wagner

The competition here is between Merrimack and (1) Central Connecticut. Both résumés are pretty much the same flavor of blah (zero Quad 1 wins, zero Quad 2 wins, two Quad 3 wins), so I’ll take the team with the better metrics. Merrimack finished the regular season at #195 on KenPom to Central Connecticut’s #255. Easy choice.

SWAC: (4) Southern over (1) Grambling State

As much as I hate to revoke a program’s first NCAA Tournament berth, Southern was clearly the best team in the SWAC this year. They got the league’s only Quad 1 win (at Mississippi State) and only Quad 2 win (at UNLV) while everybody else went a combined 0-68 in the top two quadrants.

Add in Southern’s league-leading metrics—#257 KenPom, moderately ahead of (3) Texas Southern at #277—and they’re a shoo-in.

Bid-Stealer Elimination

Replacing all of the one-bid league representatives with the worthiest team from each league is a great start, but it doesn’t get us all the way to the Best Available Bracket.

To truly reach the pinnacle, we must also say goodbye to all of the bid-stealers from multi-bid leagues. Each of them is occupying a spot that, for our purposes, would fit better in the hands of the first at-large hopefuls left out of the field. They’re better teams, after all.

We had five bid-stealers this year. One of those was New Mexico, who was on the at-large bubble, but whom the committee said was not going to be selected if they didn’t win the Mountain West Tournament. We’ll never know exactly how far out they would have been, but I’m going to assume it was within the first five and keep them in the field.

Let’s go one-by-one through the other four and replace each of them with the team they booted:

1. Goodbye Oregon, hello Oklahoma!

2. Goodbye NC State, hello Seton Hall!

3. Goodbye Duquesne, hello Indiana State!

4. Goodbye UAB, hello Pitt!

This raises the Missouri Valley from a one-bid league to a multi-bid league, rendering my whole diatribe about Drake deserving the bid over Indiana State pointless. At least they’re both in!

The Seed List

Now that we’ve reassigned all of our bids, let’s re-seed the bottom of the bracket. I’ll start by ranking all of our new bids in the full seed list, as the committee does. To do this, I’ll take the committee’s 1-68 seed list, remove the teams whose bids I reassigned, and slot the new teams in where I think the NCAA would.

That gives us this list:

40. Drake

41. (Virginia)

42. (Colorado State)

43. Oklahoma

44. Seton Hall

45. New Mexico

46. Indiana State

47. Pitt

48. Grand Canyon

49. Appalachian State

50. McNeese

51. Princeton

52. UNC Wilmington

53. Vermont

54. Samford

55. UC Irvine

56. Louisiana Tech

57. Oakland

58. Akron

59. Weber State

60. High Point

61. Morehead State

62. Colgate

63. South Dakota State

64. Norfolk State

65. Fairfield

66. Stetson

67. Southern

68. Merrimack

The Bracket

All that’s left to do is place these seeds into the bracket. Here it is:

Green: New to bracket. | Blue: Moved down a seed line. | Orange: Moved from bye to First Four. | Purple: Moved from First Four to bye.

I left the matchups the same as the real bracket where possible, but even so, this bracket looks way more fun to me.

We’ve got UConn possibly facing an in-state foe, a 2019 title game rematch, and an incredible Mountain Time 6-11 matchup all in the Round of 64, while Auburn has the opportunity for revenge on Appalachian State in the Round of 32.³

My bracket is also harsher toward the higher seeds, giving them almost universally harder matchups, sometimes ridiculously so. Here are the two most notable differences in actual point spreads for Round of 64 games vs. projected point spreads on this bracket based on efficiency metrics:

• [2] Tennessee: -21.5 vs. Saint Peter’s,⁴ -14 vs. High Point

• [3] Baylor: -13.5 vs. Colgate, -8 vs. UC Irvine

And, based on my same efficiency math, Purdue goes from a likely 22.5-point favorite against Montana State or 24.5 against Grambling State to a 17.5-point favorite against South Dakota State.

In a word, the Best Available Bracket is tougher on Goliath.

The people love upsets. This bracket would facilitate them.

1. Normally I restrict this to 11-seeds or lower, but the bid-stealers screwed up the bracket math so much this year that 10-seeds are playing in the First Four. Mostly, it seems incorrect not to call the Missouri Valley a one-bid league under the circumstances we got. Indiana State didn’t make the tournament with a Drake autobid and I’m almost certain Drake wasn’t making it with an Indiana State autobid.

2. Indiana State was +15.37 AdjEM and Drake was +13.72. That’s a difference of 1.65 points per 100 possessions. Multiply that by .67 to account for the 67 possessions that take place in an average game, and that becomes 1.11 points. Barely a point, without even accounting for any possible margin of error in the data. It’s nothing.

3. The committee usually tries to avoid non-conference rematches in the Round of 32, but it’s not a hard and fast rule, and I like chaos.

4. This opened at -18.5, right in line with where efficiency metrics say it should be, but it has moved three points in Tennessee’s favor at time of publication.