All scores →
Loading…
Loading…
Loading…

Tie-breakers in the NBA standings: how seeds are decided

By 4Basketball Staff · · 10 min read

Why NBA teams finish with identical records, and the ordered tests the league uses to split them, from head-to-head results to conference record, plus how multi-team ties are handled and why the order makes sense.

Every spring, in the last week of the regular season, the standings page fills up with ties. Two teams share the same record, or three do, and the difference between sixth and seventh, or between third and fourth, depends on a tie-breaker that most fans have not read since the last time it mattered. It matters more than it sounds. The sixth seed goes straight to the playoffs. The seventh plays an extra round. The higher seed in any series hosts more of its games. A single tie-breaker can decide which side of those lines a team lands on.

This article explains how the league breaks ties. We begin with why ties occur at all, then describe the seeding system they feed into. We lay out the ladder of tests used for a two-team tie, explain how the process changes when three or more teams finish level, work through an example, and look at the special cases of the draft lottery and the Finals. We also try to explain why the tests are in the order they are, because the logic is easy to follow once it is spelled out.

One warning applies throughout. The NBA has revised its tie-breaking procedures several times, and it may do so again. We describe the structure that has been in place in recent seasons and the reasoning behind it. For the exact current wording, the league's official rules are the final authority, and we suggest checking them before settling an argument about a particular season.

Why ties happen

A tie in the standings is inevitable, not an accident. Thirty teams play 82 games each, and each team's total of wins is a whole number between zero and 82. In a conference of fifteen teams that finish across a few dozen possible win totals, the chance that two or more share a record is high.

The schedule makes it more likely and less fair. The NBA does not have every team play every other team the same number of times. Teams play opponents in their own division four times, other teams in their conference three or four times, and teams in the other conference twice. So two teams with identical records can have reached them against different slates, and a tie-breaker has to decide which team has the better claim. Our article on how the schedule is made explains why the balance is uneven, and it is the reason that tie-breakers focus on the games teams played against each other and against comparable opponents.

What the seeds decide

Before the ladder, it helps to know what is at stake. In each conference, teams are ranked by record, and the ranking decides who plays whom.

  • Seeds one through six qualify directly for the playoffs.
  • Seeds seven through ten enter the play-in tournament, which determines the last two playoff spots in each conference. Our guide to how the play-in works covers the format.
  • Teams eleven through fifteen are out of the race.

Within the playoffs, the higher seed hosts more games in a series. For the structure of the postseason, see how the playoffs work.

One change to the seeding system is important background. For many years, the three division winners in each conference were guaranteed the top three seeds, and later the top four, regardless of record. A division winner with a weak record could hold a better seed than a team with more wins in a tougher division. The league ended that practice in the mid-2010s. Today, the teams in each conference are seeded by their records, whether or not they won a division. A division title remains a distinction, and it still plays a small part in the tie-breaker, as we will see, but it no longer overrides the standings.

The two-team ladder

When exactly two teams finish with the same record, the league works down a list of tests in order. The first test that separates the teams settles the question, and the rest are not consulted. In recent seasons, the list has run roughly as follows.

  1. Division leader. If one of the tied teams won its division and the other did not, the division winner gets the tie-breaker. This test applies only to teams in the same conference, and it arises rarely, because it only matters when a division winner and a non-winner tie.
  2. Head-to-head record. The team with the better winning percentage in games between the two teams comes out ahead. If they split a four-game series 2-2, this test does not separate them, and the ladder continues.
  3. Division record. If the teams are in the same division, the better winning percentage in games against division opponents breaks the tie.
  4. Conference record. The team with the better winning percentage in games against its own conference comes next.
  5. Record against playoff teams in its own conference. The team that did better against opponents that made the playoff field in its conference is ahead.
  6. Record against playoff teams in the other conference. The same test, applied to the other conference.
  7. Point differential. If nothing else has separated the teams, the one with the better difference between points scored and points allowed over the whole season comes out ahead.

If a pair of teams were still level after all of these, a random drawing would decide. That outcome is vanishingly rare, and in practice a tie rarely survives past the head-to-head test or the conference record.

Notice the shape of the list. It begins with the most direct evidence, the two teams' results against each other, and moves outward to progressively broader and noisier evidence. The first tests ask who actually beat whom. The middle tests ask who played better against comparable opponents. The last test uses scoring margin, which has no meaning for any particular game but is a good guide to a team's overall strength. A team that wins by a lot and loses narrowly is generally better than its record suggests, and the margin reflects that. For more on why margin predicts well, our piece on Elo ratings in basketball explains the thinking.

Why this order

The order is not arbitrary, and it is worth understanding because it shows the principles that govern fair ranking.

Head-to-head comes early because it is the most direct comparison. Two teams that finish level have, by definition, done equally well against the league as a whole. The one thing that distinguishes them most clearly is what happened when they met. If one team won the series, it has demonstrated, in the only games that involved both, that it was better. Nothing else is as relevant.

Head-to-head can be misleading, which is why it is not the only test. Teams in different conferences play only twice, and teams in the same conference play three or four times. With so few games, a head-to-head record can reflect luck, injuries or the date of a game. If it comes out even, the process looks for broader evidence.

Division and conference records follow because they are the best like-for-like comparisons. A team's record against its own division or conference is the record against teams that both candidates faced, in comparable numbers. It is a fairer comparison than the overall record, which is already tied.

Record against playoff teams checks the quality of the opposition. Beating good teams is worth more than beating bad ones, and this test rewards the team that did more against the better half of the league.

Point differential is last because it ignores which games were won. It is a valuable measure of strength, but it can reward a team that ran up scores in blowouts, so the league uses it only when everything that depends on actual outcomes has failed to separate the teams.

The progression, then, moves from specific to general and from outcomes to margins. That is a sensible way to build a ranking from inconclusive evidence.

Ties among three or more teams

Multi-team ties are more complicated, because a ladder designed for pairs does not work directly when three teams are level. The league's approach follows two ideas.

First, it applies similar tests to the group as a whole. Among the tied teams, it asks which of them has the best record in games against the other tied teams. Imagine three teams that finish level, which we will call A, B and C. A beat B twice and C twice, B beat C twice and lost twice to A, and C lost every game against the other two. Then the combined record among the three puts A first, B second and C third. If the records among the group do not separate them, the process moves on to division record, conference record and the later tests, applying them to the tied teams as a group.

Second, once one team is separated from the group, the rest start over. This restart principle is important. Suppose A is placed first by the group test. The remaining teams, B and C, are now a two-team tie, and the league returns to the beginning of the two-team ladder to decide between them. They do not carry forward the group comparison. The reasoning is that a tie among two teams is a different question from a tie among three, and each should be resolved using the evidence that is most relevant to it.

The multi-team process can produce counterintuitive results. A team may win a tie-breaker against one rival and lose it to another, so the result depends on the group's composition. That is a known feature of any fair ranking system: when the comparisons are cyclical, there is no perfect answer. The restart rule gives a clear procedure that all teams know in advance.

A worked example

To see how it works, take two hypothetical teams in the same conference that finish the season with identical records and are tied for the sixth and seventh seeds. Call them the Harbors and the Peaks. They are in different divisions and neither won its division.

The league begins with the first test. Neither team is a division leader, so it does not apply. The second test looks at the head-to-head record. The teams met three times, and the Harbors won two. That is a better winning percentage, so the Harbors take the sixth seed and the Peaks fall to the seventh. The rest of the ladder is not needed.

Now change the story. Suppose the teams met only twice and split the series. The head-to-head test does not separate them. They are not in the same division, so the division test does not apply. The next test is conference record. The Harbors finished 33-19 in conference play and the Peaks 31-21, so the Harbors win. Again the process stops.

The point of the example is that the seed can depend on games that look unrelated to the matchup: a Tuesday game against a conference rival in January may turn out to be the one that decided the sixth seed. The standings page makes this visible. When a tie is in play, the games that count most are the ones against the team you are tied with and the ones against teams in your own conference.

The stakes are large. The sixth seed goes straight to the playoffs. The seventh plays the play-in, with an extra chance to be eliminated. Fans who study the tie-breakers in April can see exactly which games they should be watching.

Special cases

Several other ranking situations use their own procedures.

The draft lottery. Teams that miss the playoffs are ranked by record to set their odds in the draft lottery. When two or more lottery teams have the same record, the tie is resolved by a random drawing, because there is no reason that one deserves a better draft position than the other. The drawing is held before the lottery, and the result determines which of the tied teams gets which set of odds. Our article on the NBA draft lottery explains how the lottery uses the standings.

The Finals. In the championship series, home-court advantage goes to the team with the better regular-season record, regardless of conference. If the two finalists have the same record, a separate procedure decides, since the conference-based tests used within a conference do not apply when the teams come from different ones.

Ties between teams in different conferences. For cross-conference ties, such as for the best overall record, the league uses a different ladder, because conference record and division record cannot be compared when the teams played different schedules. The logic is the same: start with head-to-head, then move to broader measures.

The in-season tournament. The NBA Cup, covered in our explainer, uses its own tie-breakers for group play, which rely on results within the group rather than the full-season ladder.

What to take away

A tie-breaker is a rule for answering a hard question with limited evidence: if two teams are equal on the main measure, which one is better? The NBA's answer is a ladder that moves from the most direct comparison to the most general. First the two teams' games against each other, then games within the division and the conference, then games against the strongest opponents, and finally the margin of victory over the whole season. Multi-team ties follow the same spirit and use a restart rule to keep the process consistent.

For a fan, the practical benefit is knowing which games matter in the last weeks. If your team is tied or close in the standings with a rival, count the head-to-head games remaining. If you are chasing a conference rival, remember that conference record is a tie-breaker, so a game against a conference opponent counts twice, once in the standings and once as a potential tie-breaker.

When you check the standings page, you can read the ties yourself. Look at who won the season series, then at conference records, and you will often predict the order before the league announces it. The standings page and the playoffs page track the races, and the schedule shows which games remain. And if you want to see how the same ideas apply in the college game, our article on how the tournament field is chosen shows another way to rank teams that did not play each other.

Search more on this

Keep reading

Load management: what it is and why teams do it

Load management means deliberately limiting a player's playing time or games to protect long-term health. This guide covers where it came from, the logic and limits of the evidence, and how the league has pushed back.

4Basketball Staff · Aug 21, 2026 · 11 min read