The NBA bracket has one feature that catches everyone out
On the surface, the NBA postseason is the friendliest bracket in major sport. Sixteen teams, two conferences, four rounds, every round best of seven. It is close to the clean symmetrical grid people picture when they hear the word bracket.
Then there is the play-in tournament, which sits in front of the bracket and decides two of the sixteen places, and which almost no generic template accounts for. It is not part of the bracket and it is not separate from it either: it feeds two slots that stay empty until it finishes.
The second thing worth knowing before you build anything is that the NBA does not reseed. Once the bracket is set, the paths are set. A team that beats the top seed in the first round does not get moved into an easier half; it inherits that half. This has a large effect on how a bracket should be filled in and on how a pool plays out.
Get those two things right and the rest is straightforward.
The format, in order
Thirty teams, split into two conferences of fifteen. The regular season decides the order.
Seeds one to six qualify directly
The top six records in each conference are in the playoffs, with their seeds fixed. Nothing else can change that, which is why the difference between finishing sixth and seventh is far larger than one place suggests.
Seeds seven to ten go to the play-in
Four teams per conference, competing for two places, in a short sequence that is easy to describe and easy to get wrong.
The seventh seed plays the eighth. The winner takes the seventh playoff seed and is done. The ninth seed plays the tenth, and the loser is eliminated immediately. Then the loser of the seven eight game plays the winner of the nine ten game for the final playoff place, the eighth seed.
So the seventh and eighth seeds get two chances to qualify and the ninth and tenth get one. That asymmetry is deliberate and it is worth reflecting in any bracket that shows the play-in, because it is the part people ask about.
Then four rounds, all best of seven
The first round pairs one against eight, two against seven, three against six and four against five within each conference. Then the conference semifinals, then the conference finals, then the Finals between the two conference champions.
Home court advantage follows the better regular season record in every series, in a two two one one one pattern.
That is the whole thing. Twenty teams involved, sixteen in the bracket, four rounds, one champion.
Why no reseeding changes how you should fill it in
This is the detail that separates a thoughtful NBA bracket from a list of favourites, and it is genuinely counterintuitive.
In a reseeding format, the remaining teams are re-paired each round so the best surviving team always faces the worst. Upsets get punished: beat the top seed and you are rewarded with the next strongest opponent.
The NBA does the opposite. The bracket is fixed from the start, so an upset opens a path. If the eighth seed beats the first seed, then whoever comes through the four five series faces that eighth seed rather than the top team in the conference. One upset in the right place makes an entire half of the bracket softer.
Two consequences for your picks.
- Think in halves, not in rounds. The real question is not who wins each series but who comes out of each half of each conference. A team you rate fourth best in the league might have an easier route to the Finals than the team you rate second, purely because of where they sit.
- An upset pick is worth more than it looks. Correctly calling a first round upset does not just win you one series, it changes the two rounds after it. In a pool with round weighting, this is where large leads come from.
It also means the two strongest teams in the league can meet in a conference semifinal and one of them is out in May. That is not a flaw in your bracket, it is the format.
Building the bracket
The quickest route is a template that already has the shape, including the play-in. The NBA playoff bracket page has it, and you can open the editor and enter teams straight away.
If you are assembling it yourself, these are the decisions that matter.
- Decide whether to show the play-in at all. Two valid choices. Include it, and your bracket represents the full postseason but is more complex and cannot be completed until the play-in ends. Or start at the sixteen team field and treat the play-in as a qualifier that happens before your bracket opens. For a public bracket, including it is more useful. For a prediction pool, excluding it avoids the awkwardness of people predicting a team that never makes it.
- Enter teams by seed. The pairings are entirely determined by seeding, so entering one to eight per conference in order and letting the bracket pair them is faster and less error prone than placing teams by hand.
- Keep the conferences visually separate. East on one side, West on the other, Finals in the middle. It matches how people think about the postseason and makes the no-reseeding structure legible.
- Mark the series lengths. All best of seven, which is unusual enough among sports brackets to be worth stating.
- Use team colours or logos. Sixteen names in plain text is hard to read at a glance, and a bracket that is hard to read does not get shared.
Build it somewhere it can be updated and shared as a link. A postseason runs for two months, and a bracket that cannot be updated is useful for about a day.
Filling it in without being boring or reckless
The NBA is the most predictable of the major postseasons, which creates its own problem: if everyone picks the favourites, a pool is decided by tiebreaker.
Best of seven is a long series. It gives the better team time to adjust, removes most of the randomness a short series introduces, and historically produces far fewer first round upsets than a single game format would. Top seeds usually advance.
So the skill is not in spotting chaos, it is in judging a small number of genuinely close calls correctly.
- Look at the four five and three six series. These are usually the closest matchups in the bracket and the most valuable to get right, because the teams are near enough in quality that record tells you little.
- Check the injuries, and the schedule. A best of seven punishes a thin rotation. A team whose best player is carrying something into April is a different team from the one that produced the record.
- Weigh matchup over quality in specific cases. Over seven games, a team built to exploit a particular weakness can beat a better team. This matters more in basketball than in most sports because rotations are short and a mismatch can be targeted every possession.
- Do not fear picking chalk in round one. Picking four favourites in a conference is often correct. The differentiation comes later.
- Remember the play-in team is playing well. A team that has just won one or two elimination games to get in has momentum and a healthy disregard for seeding. They are still likely to lose to a top seed in seven, but they are not the pushover the eighth seed used to be.
Running an NBA bracket pool
An NBA pool has one structural difficulty: the postseason lasts about two months, and keeping people interested that long takes more than a prize.
Weight the rounds heavily
With four rounds and mostly predictable first rounds, flat scoring means everyone is tied after round one and the pool is dull. Something like one, two, four and eight points per correct series by round keeps the later rounds decisive and rewards the people who thought about the conference finals rather than just the openers.
Consider a series length bonus
Asking for the result in games, four one, four two, four three, and awarding a bonus for getting it exactly right is where basketball knowledge shows. It also creates separation without needing upsets.
Set the deadline at the first game of round one
If you include the play-in, you have a choice: lock entries before the play-in begins, with two slots unknown, or after it ends, which is cleaner and gives people less to predict. Either works. Announce which one before entries open.
Publish standings after every series
Over two months this is the whole difference between a pool people follow and a pool people forget. A live bracket plus a leaderboard, updated as series finish, is what gives the group chat something to react to.
Pick a tiebreaker in advance
Total points in the deciding game of the Finals is conventional and fine. What matters is that it exists before anybody needs it.
Sharing and embedding it
A bracket is a social object, and how you distribute it matters as much as how you build it.
A shared link is the default. One address that always shows the current state, which you can paste anywhere, and which updates for everyone when you update it once.
An embed puts the bracket on your own page, which is what you want if you are running a pool for a club, a workplace, a league or an audience. It keeps the bracket next to the rules and the standings, and it means people come to your page rather than someone else's.
An exported image is for social posts that will not render an embed, and it is a snapshot rather than the bracket. The most common way a good bracket becomes useless is somebody screenshotting it in April and sharing that picture for the next two months.
If you embed it, put it on a page with something to read. A bracket alone gives a visitor nothing before or after, and nothing for anyone to find the page by. The format explainer, your predictions or the pool rules are all fine company.
Variations worth considering
The straight prediction bracket is not the only option, and for some audiences it is not the best one.
- A live results bracket. Not predictions at all, just what has happened, updated as series finish. It is the most useful version for a general audience and the easiest to maintain interest in, because there is always something new.
- Predictions and results side by side. Two brackets, one of them yours from April, one of them reality. The gap between them is the content, and it gives you something to post about every week.
- Round by round picks. Instead of predicting everything in April, reopen picks each round. It is a weaker pool because there is no long shot glory, and it is much better at keeping people involved, because a bad first round does not eliminate anyone from caring.
- A voting bracket. Ask who people want to win rather than who will. It suits an audience that does not follow the league closely, and the result is interesting in a different way.
The right choice follows the audience. Dedicated fans want the full prediction bracket with series lengths. A workplace or a broad following will engage far more with round by round picks or a straight vote.
Why the play-in exists, and what it changed
A short piece of context, because it explains why the bracket has a bit bolted onto the front of it and why that bit is not going away.
Before the play-in, the eighth seed in each conference was simply the eighth best record, and by March a large number of teams were mathematically out of contention with a month of the season left. That produced two problems the league cared about. Teams with nothing to play for had an incentive to lose rather than win, because the draft rewards it. And a very long stretch of the calendar had games that mattered to nobody.
Extending the field to ten teams per conference, with the last two places decided in elimination games, pulls the cut line down the standings. Suddenly a team sitting tenth in April is playing for something real, and the incentive to lose deliberately arrives later and applies to fewer teams.
It also produced a genuinely good few days of basketball, which is why it survived its trial period. Single elimination games between teams with something at stake are the most watchable thing the regular season can produce.
For a bracket builder the effect is practical. The postseason field is no longer known when the regular season ends: two seeds per conference remain open for several more days. Any bracket you publish before the play-in finishes has four empty slots, and any pool you run has to decide whether people are predicting into that uncertainty or not.
Keeping the bracket useful during the playoffs
Most brackets are built once, in April, and then quietly go stale. Keeping one current through two months is not much work and it is where most of the value is.
- Update after every series, not every game. A best of seven can swing several times, and a bracket that changes daily is noise. When a series ends, advance the winner and note the games. That is a natural weekly rhythm.
- Record the series score, not just the winner. Four two tells a story that a tick does not, and if your pool has a series length bonus you need it anyway.
- Keep the eliminated teams visible. A bracket that deletes losers loses its history. Part of what people want from a bracket in June is to look back at what happened in April.
- Post the update somewhere. Advancing the bracket and telling nobody is half a job. A short note with the link, after each round, is what keeps a pool alive.
If you are running the bracket for an audience rather than a group of friends, the update itself is content. A bracket page that changes every few days gives people a reason to come back, which a static one never does.
Home court, and how much it is actually worth
Every NBA series gives home court to the team with the better regular season record, in a two two one one one pattern: two games at the higher seed, two at the lower, then alternating with game seven at home for the favourite.
It is worth something and it is worth less than people assume when filling in a bracket. The team with home court wins a clear majority of series, but most of that is because the team with home court is usually the better team. Strip out quality and the advantage of the venue itself is real but modest.
Where it matters most is game seven, and only game seven. A series that reaches a deciding game at the higher seed's arena is a meaningful edge, which is the argument for taking home court seriously when you are choosing between two teams you rate as close to equal. If you cannot separate them on merit, take the one with game seven at home.
Where it matters least is the first two games. A team that loses the opener at home has not lost the series, and brackets that treat an early road win as decisive tend to overreact.
One practical note for a pool with a series length bonus: the two two one one one pattern makes four two and four three the most common outcomes in close series, and sweeps rarer than they feel. If you are guessing a length, six games is the safest bet in a series between two good teams.
Common mistakes
- Ignoring the play-in. A sixteen team template built without it does not match the postseason, and two of your seeds will be wrong.
- Assuming the bracket reseeds. It does not, and a pick sheet built on that assumption misreads every path after round one.
- Treating a round as one game. Every NBA round is best of seven, and a bracket that records single results cannot represent it.
- Flat round scoring. With predictable first rounds, this leaves everyone tied and the pool decided by luck.
- Filling it in before the seeds are final. Seeding often changes on the last night of the regular season, and the play-in changes it again.
- Sharing a screenshot. It cannot be updated, and a two month postseason needs updating.
- No standings. Over two months, a pool nobody can check is a pool nobody remembers.
Where to start
An NBA bracket is the most symmetrical of the major postseason brackets and it has two details a generic template will miss: the play-in tournament that fills the last two seeds in each conference, and the fixed bracket that never reseeds. Both change what a good pick sheet looks like.
Start from the NBA playoff bracket template, decide whether to show the play-in, enter the teams by seed once the regular season and play-in have settled, and weight your pool towards the later rounds so it stays alive into June.
When you are ready, create your NBA playoff bracket and share it as a link so it can be updated as the postseason unfolds.
