The bubble: who is in and who is out
The bubble is the cluster of college teams whose NCAA tournament fate is genuinely undecided. Here is how it forms, what the committee weighs, why other teams' results matter, and how to read the arguments.
Every March, a few dozen college basketball teams spend their weekends doing something that has little to do with winning the next game. They are waiting. Their records are decent, sometimes very good, but not so good that the NCAA tournament is guaranteed, and not so poor that it is out of reach. They are on the bubble, and for several weeks their season is a slow argument conducted in public by television analysts, bracket writers and fans who have suddenly learned what a strength-of-schedule number is.
The word is a good one. A bubble is something that can hold its shape or burst, and nobody can say which until the moment it happens. A team can be sitting comfortably in the field on Monday and out of it by Saturday without losing a game, simply because other teams around it won. That quality, the dependence on results that have nothing to do with you, is what makes the bubble different from every other kind of ranking in sports.
This article explains the bubble from the ground up: how the tournament field creates it, the tools the selection committee uses to sort the teams on it, why a conference tournament in a far-off league can sink a team that never plays in it, and a few habits for judging bubble arguments honestly. It is a companion to our explanations of how the NCAA tournament field is chosen and of how the selection committee works, and it concentrates on the teams in the middle, because that is where the interesting decisions are made.
Where the bubble comes from
The men's and women's tournaments each take 68 teams. A little more than 30 of those places go automatically to conference champions, a mechanism we cover in automatic bids and conference tournaments. The rest, roughly three dozen, are "at-large" bids handed out by the selection committee to the best remaining teams, whatever conference they play in.
That second group is the whole story. If every place were automatic, there would be no bubble, only a list of champions. If the committee could take as many teams as it liked, there would be no bubble either, only a long list of acceptable ones. The bubble exists because the number of at-large places is fixed and smaller than the number of teams that can make a respectable case.
Picture the full Division I field of more than 350 teams being sorted by the committee into layers. At the top are the teams that are obviously in: the top seeds, the strong champions of big conferences, the teams with ten or more wins against strong opposition. At the bottom are the teams that are obviously out: the ones with losing records, the ones with no good wins at all. The bubble is the thin layer between them, the last handful of places and the first handful of teams left over. It is usually somewhere between a dozen and two dozen teams wide in the weeks before selection, and only a handful of real decisions in the end.
There is a nice way to see how small the real decision is. Suppose the committee thinks 28 or 29 of the 37 at-large teams are safe by the last Saturday. Then the whole drama of Selection Sunday concerns the final eight or nine places, and a pool of candidates that is usually larger than that, so the bubble is a contest in which most of the contenders will be told no.
The vocabulary of the bubble
Bubble talk has a language of its own, and understanding it makes the arguments easier to follow.
- Lock: a team so well placed that no plausible set of results would drop it out.
- Last four in: the final four at-large teams selected. These are the teams that play in the First Four, usually for an 11 seed, as we explain in the First Four.
- First four out: the first four teams left out, the ones who came closest.
- Next four out: the next group down, teams that were considered but never close.
- Bid thief: a team that wins its conference's automatic bid when it would not have been chosen as an at-large team. Every bid thief removes one at-large place from the pool.
- Resume: the full record of what a team has done, including its wins, its losses, where those games were played, and who the opponents were.
- Eye test: a judgment based on watching a team play rather than on its numbers.
Media analysts publish projected brackets through the winter, updating them daily in March, and the lists of "last four in" and "first four out" are a staple of that coverage. The committee itself does not publish such a list in advance, and in some seasons it has released a preliminary look at the top seeds in February, so the bubble talk you hear is informed guesswork about what the committee will do. The guesses are often good. They are never certain.
The tools: NET, quadrants and the team sheet
For decades the committee leaned heavily on a formula called the Ratings Percentage Index, or RPI, which dated to the early 1980s. It rewarded winning percentage and the winning percentages of opponents and opponents' opponents. It was easy to understand and easy to game: a team could lift its number by scheduling certain opponents, and the number cared little about how a team actually won. Beginning with the 2018-19 season the NCAA replaced it with the NCAA Evaluation Tool, known as the NET.
The NET is a ranking of every Division I team. It combines game results, the strength of the schedule, where games were played, and a measure of offensive and defensive efficiency, with scoring margin counted but capped so that running up the score does not help. You can follow the current men's and women's lists on our rankings page.
The NET also drives the quadrant system, which sorts every game a team has played into one of four groups according to the opponent's NET rank and the location. A road game against a top-75 opponent is Quadrant 1. A home game against a team ranked below 160 is Quadrant 3 or worse. The thresholds the NCAA has published look like this:
| Quadrant | Home | Neutral | Away |
|---|---|---|---|
| 1 | 1 to 30 | 1 to 50 | 1 to 75 |
| 2 | 31 to 75 | 51 to 100 | 76 to 135 |
| 3 | 76 to 160 | 101 to 200 | 136 to 240 |
| 4 | 161 and below | 201 and below | 241 and below |
The ranges are the ones the NCAA introduced with the NET, and the committee can revise them, so treat the exact numbers as a snapshot of the idea. The idea itself is lasting: a win over a good team on the road is worth much more than a win over a poor team at home, and the committee wants to be able to see at a glance how often a team did the difficult things.
All of this arrives on a single sheet, the team sheet, that each member of the committee receives for every team under consideration. It shows the NET rank, the record in each quadrant, the strength of schedule, the nonconference schedule, and a list of results sorted by quality. Since the NET's arrival, the sheets have also carried several other well-known ratings: results-based measures that reward what a team has accomplished and predictive measures, of the kind you see from Ken Pomeroy and others, that estimate how strong a team really is. Committee members have said, and the NCAA's published principles reflect, that these outside numbers are reference points rather than rulings. The committee votes; the numbers inform.
What actually gets a team in
A strong bubble case is rarely one number. It is a pattern, and the pattern usually has four parts.
Quality wins
Quadrant 1 victories are the currency of the process. A team that is 8-10 in Quadrant 1 is often in better shape than a team that is 20-2 outside it, because the committee is trying to answer one question: how would this team do against the other teams in the field? A record built mostly against opponents who would not be in the field answers that question poorly.
Avoiding bad losses
The mirror image of a quality win is a bad loss. A defeat in Quadrant 3 or 4 is a serious mark, because it suggests that the team is capable of losing to someone nobody would call a tournament team. One such loss is forgivable for a team with a large pile of good wins. Two or three, with few good wins to balance them, usually mean a trip to the NIT; we describe that event in the NIT: history and why it exists.
The schedule you chose
Teams cannot control their conference schedule, but they do control their non-conference one. A team that schedules ambitiously and splits the games is often praised; a team that schedules nothing but home games against low-ranked opponents and goes undefeated can find that the committee shrugs. The point is not the record but the opportunity: did the team seek out chances to prove itself?
The shape of the season
The committee has said it looks at how a team has played over the whole season, but late form matters, and it also asks whether a key player was missing for part of the year. If a team went 6-8 without its best player and 14-3 with him, that is information, and the committee is allowed to use it. Injuries, along with other situations like suspensions, are one of the few places where the committee openly departs from pure record.
Notice what is not on the list: the conference name. The committee is meant to look at the team, not its league. In practice teams from stronger conferences have more chances to collect Quadrant 1 wins simply by playing their conference schedule, which is one reason the power leagues take so many bids. A mid-major team with a good record has to find those games elsewhere, a problem we discuss in mid-major basketball.
Why other people's games decide your fate
The strangest feature of the bubble is that a team can improve its position by watching. Consider three kinds of results that happen far from your own gym.
Bid thieves. On the last Saturday before Selection Sunday, a conference whose regular-season champion has already earned a place might see a lower seed win the tournament. That lower seed gets the automatic bid, but its regular-season champion still gets an at-large bid. This leaves the pool unchanged. The painful case is the opposite one: a team that would not have been chosen at large wins the tournament. The result is one fewer place for the bubble, and a bubble team in another state is out because of a game it did not play. The more automatic bids go to unexpected winners, the more crowded the bubble becomes.
Quality of your wins changes. A Quadrant 2 win in January can become a Quadrant 1 win in March if the opponent climbs the NET, and vice versa. A team that beat a mediocre opponent before Christmas may find itself grateful for a late-season surge by that opponent.
Rival bubble teams. Two teams with similar resumes are competing for one place, and when one wins a game against a strong opponent, the other's argument weakens. Fans who root against strangers in the final week are doing something rational.
This is why bubble watching is a communal exercise. There are many teams and many results, each one is minor, and the sum of them decides a few places. A coach's whole season can turn on a team he has never heard of.
Famous bubble stories
Bubble teams are often written off as the weak part of the field, but history is kind to them. VCU in 2011 was one of the last at-large teams chosen, played in the very first First Four, and went to the Final Four. UCLA in 2021 played in the First Four as an 11 seed and also reached the Final Four, losing to Gonzaga on a buzzer-beating shot by Jalen Suggs in overtime. Syracuse was in the same group in 2018, was chosen as one of the last teams and won through to the Sweet 16 before losing to Duke.
These examples teach something about the bubble that is easy to miss. The committee is judging a body of work, not predicting a tournament run, and the difference between the 36th and 40th best at-large candidate is smaller than the difference between how any given team plays on any given night. A team that is barely in is not necessarily weak; it is a good team whose resume has a flaw, perhaps a bad loss or a thin schedule, that the committee decided to forgive. We look at what happens once these teams are in the bracket in Cinderellas and why upsets happen.
The opposite experience is just as common and rarely celebrated: the team with a strong record left out. When a team with a high win total misses the field, the usual explanation is on the team sheet, where a soft schedule and few Quadrant 1 wins can outweigh a gaudy record. Fans tend to feel this as an injustice, and sometimes it is a judgment call that could reasonably have gone either way. Close decisions always contain the possibility of regret.
How to read a bubble argument
You can sharpen your own judgment by asking a short list of questions whenever someone says a team is in or out.
- Where is the team in the NET, and what is its record against Quadrant 1 and Quadrant 2? A team with a mediocre NET rank can still be in with a pile of top wins, but it is rare.
- How many bad losses does it have? Look at Quadrant 3 and 4 losses, not at total losses.
- What does the nonconference schedule look like? Did the team play anyone?
- Can the team still win its conference tournament? If it can, nothing else matters, because an automatic bid ends the argument. If it cannot, a loss in a conference tournament can be costly even to a team that was already in good shape.
- Who else is competing for the same place? Compare resumes directly rather than judging a team in isolation.
- Was a key player missing for significant stretches? If so, the committee may weigh the games with him more heavily.
Be suspicious of three things. First, arguments that rely on a single statistic: no single number decides the bubble. Second, arguments that rely on a team's "name" or past success; the committee is meant to evaluate this season only. Third, anyone who claims to know the final answer before the committee has met. Even the best projections miss a few teams every year, and the misses are usually the teams on the edge.
The bubble from the inside
It helps to imagine the committee room on Selection Sunday week. The members have spent months studying data and watching games, and they gather in Indianapolis to rank the candidates. They start with a first ballot, in which each member lists the teams that member believes belong in the field, and a team named by enough members is placed in. They then repeat the process, narrowing the group of unselected teams and voting on who should be added next. Eventually the last few places are filled by direct comparison between teams the room considers close. Members from the same school or conference leave the room when their teams are discussed, so the people deciding are working under rules meant to keep conflicts out.
The timing adds pressure. Conference championship games are played on the final weekend and the committee waits for them, adjusting its work as the results arrive. That is why the bracket is not finished until the last conference title game has ended: an upset in the last hours can force a change to the at-large order, and the committee builds its choices in a way that can absorb such surprises. The bubble is the reason a bracket cannot be written in advance.
What to take away
The bubble is not a flaw in the system; it is the system doing what it was designed to do. A field of 68 with roughly three dozen at-large places will always have teams that could easily have been chosen and teams that barely were. What separates a thoughtful view of it from a frustrated one is a habit: look at the whole body of work, remember that a quality win is worth more than a pile of easy ones, and accept that the end of the season belongs partly to teams you will never see play.
If you want to follow it yourself, check the rankings for the NET, the men's bracket pages for how past fields were set, and the college basketball section through March. Then watch the weekend before Selection Sunday with one eye on the teams around your favorite. That is where the bubble actually bursts.