A basketball at rest on a polished court, arena lights reflecting in the floor behind it.

NBA predictor

Eighty-two games a team, basketball almost every night, and a leaderboard that moves while you sleep.

An NBA prediction league is the one competition where the schedule itself is the edge. Back-to-backs, rest days and the NBA Cup all sit inside the same 82 games. Fixtures and scoring run automatically, in your branding.

Free to start. No credit card.

The 2026-27 NBA season, at a glance.

Teams and games
30 teams across two conferences, 82 games each. Two of those 82 are not fixed at schedule release: they're determined by how far a team goes in the NBA Cup.
Opening night
Tuesday 20 October 2026, with a three-game opening slate.
Regular season ends
Sunday 11 April 2027, with all 30 teams playing and start times staggered so no team gains an information advantage.
Emirates NBA Cup
Group play on Fridays from 30 October to 27 November 2026, knockouts 4 to 11 December, and the championship game on 11 December at Hinkle Fieldhouse. Group games count as regular-season games too.
Play-In Tournament
Six games deciding the seventh and eighth seeds in each conference. Seeds 7 through 10 in both conferences are still alive after game 82.
All-Star break
19 to 21 February 2027 in Phoenix. The one week in six months with no games to predict.
Playoffs
Four rounds, best of seven, sixteen teams. A series rather than a single match, which is a different prediction problem entirely.
UK viewing
Most games start between midnight and the early hours in the UK, so results are in by breakfast. Weekend afternoon games in the US are the exception, landing in the UK evening.

What a round looks like when there are games every night.

  1. You decide what a round is.

    The NBA doesn't have rounds. It has a calendar with roughly a dozen games most nights. A round can be a calendar week, a single night's slate, or a fixed number of matchups you nominate. Most organisers run weekly rounds with a single deadline, which turns a continuous season into something with a rhythm.

  2. Fixtures and locks arrive on UK time.

    Start times, lock countdowns and reminders show in your players' time zone. A 7pm game on the US east coast is midnight here, so a round that looks open on a Tuesday evening is about to start locking. Fixture-level locks handle that without anyone doing the arithmetic.

  3. Results score overnight and the leaderboard is there in the morning.

    Games finish while your players are asleep. Scores post as each one ends, and by the time anyone opens their phone on the train the standings have already moved. The NBA builds a morning-after habit, the same way the overnight games always have.

The arguments every NBA predictor organiser has.

Basketball breaks the assumption that the team you picked on Monday is the team that plays on Tuesday. Almost every scoring problem in an NBA competition comes back to who is actually on the floor.

Load management makes the team sheet the pick.

Stars sit out games they're fit for. A matchup that looked one-sided when the round opened can invert when an inactive list drops an hour before the start. If your locks sit at the start of each game, the players who stay up for the injury news hold a real advantage over the people who filled their card on Sunday night. That's either a feature or a fairness problem, and you should decide which before October rather than in January.

Back-to-backs are the most reliable edge in the competition.

Teams play on consecutive nights, often in different cities. The second night of a road back-to-back is measurably harder than the first, and it's visible in the schedule months in advance. In a competition where nothing else is predictable, the travel calendar is the one repeatable signal, and the players who read it will win your season.

Margin actually works here, which it doesn't in football.

NBA scores run into the hundreds and margins spread wide, so a margin prediction is a real weekly scoring mechanic rather than a season tiebreaker. This is the one competition where you can score margin every round and have it separate the field honestly.

NBA Cup group games count twice, and people will not realise.

Friday group-stage games in November are regular-season games and tournament games at the same time. Decide whether your competition scores them once, as ordinary fixtures, or runs the Cup as a separate short competition on top. Running it separately gives you a five-week sub-competition inside the season, which is a genuinely good hook for an audience that won't commit to 82 games.

There are no draws, so half your football rulebook is dead weight.

Overtime resolves every game. The draw rule your Premier League predictor needs does nothing here, and the settings that actually matter are your round definition, your lock timing relative to the inactive list, and whether margin is scored weekly.

Formats that suit a season with games every night.

Weekly slate predictions

Group a week of games into one round with a single deadline. The format that turns a continuous 82-game season into something an ordinary office can actually keep up with.

Margin predictions

Wide, stable margins make this the one competition where predicting the number is as playable as predicting the winner. Score it weekly rather than saving it for the tiebreak.

NBA Cup competition

Five weeks from the end of October to a single championship game in December. A short, self-contained competition for people who won't sign up for a six-month season.

Playoff bracket

Sixteen teams, four rounds, best of seven. Fill it in in April, then watch it survive until June. The bracket that brings back everyone who stopped predicting around the All-Star break.

Why run your NBA predictor on Fanzava.

You define the round, so the season is manageable.

A nightly calendar becomes a weekly competition, or a nightly one, or a slate you nominate. The format bends to your players rather than the other way around.

Your logo, your colours, your domain.

White label on every paid plan. Pro and above add a custom domain and a custom email sender, so a competition about an American league still arrives entirely in your branding.

Locks that work while you sleep.

Each game locks at its own start time, in the middle of the UK night, without anyone staying up to close it. The overnight slate is scored by the time the office opens.

Six months of contact, and a bracket at the end of it.

October to June, with the NBA Cup in November and the playoffs in April. Three distinct hooks in one season, all running on the same hub with the same player list.

Questions NBA organisers ask.

How do you run predictions when there are games every night?
You define what a round means. Most organisers group a calendar week of games into one round with a single deadline, which gives a continuous season the weekly rhythm a football gameweek has naturally. You can also run nightly rounds for a basketball-literate audience, or nominate a fixed slate of marquee matchups each week for a casual one.
How should we handle load management and late withdrawals?
By deciding where your locks sit. Inactive lists drop shortly before each game, so a lock at the start rewards people who check and a weekly lock rewards people who plan. Neither is wrong. What causes complaints is not saying which one you've chosen before the season starts.
Should the NBA Cup be part of the main competition?
Group-stage Cup games are also regular-season games, so they'll appear in your fixtures either way. The interesting option is running the Cup as a separate five-week competition on top: it's short, it's self-contained, it finishes with one championship game in December, and it gives you a way to recruit people who won't commit to an 82-game season.
Do margin predictions work in basketball?
Better than in any other sport here. Scores run into the hundreds and margins spread wide enough that predicting the number genuinely separates the field. You can score margin every round here rather than holding it back as a season tiebreaker, which is what it has to be in football.
What does it cost?
Free for up to 25 players with no credit card. Paid plans are priced in pounds sterling and sold to businesses, with a VAT number given at checkout; the pricing page lists each plan's price and player cap. An NBA season runs October to June, so plan for eight or nine months rather than a short competition.

Ready to turn 82 games a team into a competition your office can actually keep up with?

Free to start. No credit card.

Want a weekly football competition? See the Premier League predictor →