Once the regular season and the Play-In Tournament are both done, 8 teams from each conference — 16 total — make the NBA playoffs. The actual seeding format is fairly simple and shows up all over American pro sports playoffs: the top seed plays the bottom seed, the second-best plays the second-worst, and so on down the line. Here’s exactly how that bracket gets built.
How the bracket is set: 1v8, 2v7, 3v6, 4v5
The playoff bracket follows a fixed formula: each conference ranks its 8 playoff teams 1 through 8 by regular-season record, and the first-round matchups are:
- No. 1 seed vs. No. 8 seed
- No. 2 seed vs. No. 7 seed
- No. 3 seed vs. No. 6 seed
- No. 4 seed vs. No. 5 seed
Every round is best-of-7 — first team to 4 wins advances.
The bracket never gets reseeded
Most regular NBA viewers won’t even ask this, but I’ve genuinely had a few casual friends who don’t follow sports much ask me: doesn’t the league re-rank the remaining teams by record after every round? The answer is no — once the first-round matchups are set, the path to each round after that, and who you’ll face there, is locked in. This is actually part of why you’ll sometimes hear that a team is deliberately managing its own seeding — they’re trying to make sure the path to the later rounds runs through opponents they feel better about. Say a team is locked into the No. 6 seed, but only landed there because its franchise player missed a huge chunk of the season with an injury and has looked outstanding since coming back late in the year. The team currently sitting at No. 3 might actually prefer to drop to No. 4, just to guarantee it never has to face that surging No. 6 seed before the conference finals. That, in turn, can make the race for the No. 1 and No. 2 seeds even more intense, since the No. 2 seed might end up facing that No. 6 seed as early as the second round.
The two conference champions then meet in the NBA Finals, also best-of-7.
There used to be a rule guaranteeing division winners at least the No. 4 seed — it got scrapped
Here’s a rule fewer people remember: starting in 2006, the NBA guaranteed that a division winner would never be seeded worse than No. 4 in its conference, regardless of actual record — the idea being to protect the value of winning a division. But the rule kept causing controversy: division winners with clearly worse records sometimes ended up seeded above teams with better records that just happened to be stuck in a stacked division. In 2006 itself, the New Jersey Nets won the Atlantic Division at 49-33 and took the No. 3 seed, while the Cleveland Cavaliers, at a better 50-32, were only seeded No. 4. Taken to a genuinely extreme scenario, you could theoretically have had an entire division where every team’s record was too weak to belong in the playoffs at all, and yet the rule would still hand that division’s “winner” a top-4 seed — a disaster for the quality of the postseason. That “worse record, higher seed” problem kept resurfacing, and the league finally scrapped the guarantee after the 2015-16 season, going back to pure record-based seeding.
Why a fixed bracket instead of reseeding
The logic behind a fixed bracket is that it makes the whole path to the finals transparent from the moment the first round is set — every team knows, in theory, who it has to get through, and front offices and fans alike can plan around that. Reseeding after every round would make scheduling and home-court assignments far more complicated.