NBA 2026-27
LIVENBA title, conference, playoff and play-in odds for all 30 teams, refreshed daily.
Elo + net rating prior · blended (weight 0.60 of 0.60 max, Polymarket futures, 30 teams) · updated 2026-10-13 · preseason forecast
Live forecast
The Spurs have a 19.4% chance to win the NBA title
20,000 simulations · field 38%
The race for the title
Share of simulated seasons each team wins the NBA Finals.
More
The races still open24 teams
Every team the model puts between 15% and 85% to reach the 16-team playoff bracket, closest to a coin flip first.
Why this matters
Trail BlazersNorthwest30% to finish top six · 39% to land in the play-in · 39.6 projected wins
SunsPacific31% to finish top six · 37.8% to land in the play-in · 39.6 projected wins
MagicSoutheast31.3% to finish top six · 37.9% to land in the play-in · 40.4 projected wins
HornetsSoutheast35.9% to finish top six · 31.8% to land in the play-in · 40.8 projected wins
WarriorsPacific31.6% to finish top six · 40.1% to land in the play-in · 40.0 projected wins
HawksSoutheast35.3% to finish top six · 37.6% to land in the play-in · 41.3 projected wins
ClippersPacific24% to finish top six · 33% to land in the play-in · 37.4 projected wins
PacersCentral23.1% to finish top six · 28.7% to land in the play-in · 37.1 projected wins
MavericksSouthwest16.4% to finish top six · 33.3% to land in the play-in · 35.7 projected wins
HeatSoutheast53.2% to finish top six · 33.1% to land in the play-in · 44.9 projected wins
JazzNorthwest16.1% to finish top six · 27.1% to land in the play-in · 34.4 projected wins
RaptorsAtlantic53.5% to finish top six · 32.9% to land in the play-in · 44.9 projected wins
LakersPacific53.2% to finish top six · 33.2% to land in the play-in · 44.1 projected wins
CavaliersCentral55.8% to finish top six · 31.5% to land in the play-in · 45.3 projected wins
WizardsSoutheast15.2% to finish top six · 21.5% to land in the play-in · 33.1 projected wins
RocketsSouthwest58.8% to finish top six · 29.4% to land in the play-in · 45.2 projected wins
76ersAtlantic64.1% to finish top six · 25.3% to land in the play-in · 47.2 projected wins
PistonsCentral67.4% to finish top six · 23.8% to land in the play-in · 47.9 projected wins
NuggetsNorthwest66.3% to finish top six · 25.2% to land in the play-in · 46.6 projected wins
TimberwolvesNorthwest66% to finish top six · 25.5% to land in the play-in · 46.5 projected wins
GrizzliesSouthwest7.4% to finish top six · 24.5% to land in the play-in · 32.4 projected wins
BullsCentral7.3% to finish top six · 24.5% to land in the play-in · 33.1 projected wins
CelticsAtlantic69.3% to finish top six · 23.3% to land in the play-in · 48.1 projected wins
PelicansSouthwest7% to finish top six · 24.5% to land in the play-in · 32.4 projected wins
The field the model expects20 teams
The ten teams from each conference the model has reaching the playoffs most often: six straight in and four in the play-in, two of whom survive. A play-in tag marks a team the model expects to find there rather than in the top six. Ordered by those odds, not by the standings.
HawksPlay-in
HornetsPlay-in
MagicPlay-in
WarriorsPlay-in
Trail BlazersPlay-in
SunsPlay-in
ClippersPlay-inEvery club, every outcome6 divisions
Projected final wins with the P10 to P90 range, and the odds of each landing spot, division by division.
Eastern Conference
Western Conference
Get the data: season table as CSV · raw JSON on GitHub
Where these numbers come from
Ratings blend last season's closing Elo and net rating, shrunk toward average, with the championship futures market, then replayed forward 20,000 times against every remaining game on the real 1,200-game schedule, the play-in and the sixteen-team bracket. Last generated 2026-10-13 (preseason).
How the model works
TestedEach drawn season conserves its games, fills every playoff slot once and never crowns an eliminated club under property tests; the season odds have no published backtest. About the tiers
Each team's rating is a points-of-margin number. It blends last season's closing Elo (regressed a quarter of the way back to average, then converted at 28 Elo points per point of margin) with last season's net rating (keeping 60% of it), and, before and early in the season, the championship futures market, which is where roster changes live: trades, signings and injuries that last year's numbers cannot see. The market's say fades as the season is played, from 60% of the blend today to nothing by the end. This season's real margin folds in at a weight that climbs with games played.
The game model. A game is a normal margin: the home side wins with probability Φ((rating gap + 2.3 points of home edge) / 12). The 2.3-point edge is the current-era NBA home advantage; Neil Paine measures it at about 41 Elo points, between 1.5 and 2.5 points of margin. Neutral-site games carry no edge.
Every one of the 1,230 games left is simulated 20,000 times, with each simulated season drawing league-wide, division and per-team rating errors so the output carries the model's own uncertainty. Ties are broken on head-to-head record, then a coin flip. The postseason is the real one: seeds 1 to 6 go straight in, 7 to 10 play in (7 v 8 for the 7 seed, 9 v 10 with the loser out, then the loser of 7 v 8 hosts the winner of 9 v 10 for the 8 seed), then four best-of-seven rounds with home court to the better record. Market: blended (weight 0.60 of 0.60 max, Polymarket futures, 30 teams).
What it does not know. No injuries, no player ratings, no rest or travel. Before tip-off this forecast is a prior, and it tightens as games arrive.




