2026-27 Season Simulator
50,000 simulated seasons · updated nightly · last run Oct 16, 5:30 AM ET · preseason projection
Championship odds
Share of the 50,000 simulated seasons each team won the title.
- 1OKC16.0%
- 2
SAS9.4% - 3DET8.6%
- 4
DEN8.5% - 5
POR8.2% - 6
PHI6.5% - 7
CLE6.4% - 8
HOU5.6% - 9
BOS5.1% - 10
TOR4.5% - 11
NYK3.3% - 12
CHA2.8%
Offseason movers
How much the 2026-27 roster changes a team's rating vs the minutes it actually got in 2025-26 (points per game, before weighting).
- POR+7.8
Added Ja Morant (from MEM)
- PHI+6.1
Added Jaylen Brown (from BOS)
IND+4.7Added Kelly Oubre Jr. (from PHI)
BKN+3.5Added Julius Randle (from MIN)
LAC-7.9Lost Kawhi Leonard (to TOR)
SAC-6.5Lost DeMar DeRozan (to DEN)
- BOS-4.0
Lost Jaylen Brown (to PHI)
All 30 teams
| Team | Conference seed odds (1 → 15) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
1 OKC | 54.345-64 | 95.3% | 88.3% | 10.5% | 42.4% | 25.1% | 16.0% | +5.4 | |
2 SAS | 51.442-61 | 90.9% | 79.7% | 17.5% | 30.2% | 15.9% | 9.4% | +4.0 | |
3 DET | 49.639-60 | 87.2% | 76.1% | 18.4% | 31.9% | 18.3% | 8.6% | +3.5 | |
4 DEN | 50.240-60 | 87.8% | 74.7% | 21.3% | 28.0% | 14.7% | 8.5% | +3.8 | |
5 POR | 50.040-60 | 87.8% | 74.1% | 21.9% | 27.2% | 13.9% | 8.2% | +3.7 | |
6 PHI | 48.138-58 | 82.9% | 69.8% | 22.7% | 26.9% | 14.5% | 6.5% | +2.9 | |
7 CLE | 47.938-58 | 82.1% | 68.6% | 23.4% | 26.1% | 13.8% | 6.4% | +2.8 | |
8 HOU | 48.338-58 | 82.6% | 66.3% | 27.2% | 21.1% | 10.2% | 5.6% | +2.9 | |
9 BOS | 46.936-57 | 78.7% | 64.1% | 25.7% | 22.5% | 11.3% | 5.1% | +2.3 | |
10 TOR | 46.136-56 | 75.8% | 60.0% | 28.2% | 20.8% | 10.4% | 4.5% | +2.1 | |
11 NYK | 44.734-55 | 70.7% | 53.7% | 30.9% | 17.1% | 8.2% | 3.3% | +1.6 | |
12 CHA | 44.034-54 | 67.5% | 50.0% | 32.7% | 15.3% | 7.2% | 2.8% | +1.2 | |
13  PHX | 44.734-55 | 68.8% | 48.2% | 38.0% | 12.7% | 5.3% | 2.7% | +1.5 | |
14  LAL | 44.434-54 | 67.9% | 47.2% | 38.3% | 11.8% | 4.8% | 2.3% | +1.4 | |
15  ORL | 43.333-54 | 63.7% | 46.0% | 34.0% | 13.7% | 6.2% | 2.3% | +0.9 | |
16  GSW | 43.633-54 | 64.0% | 42.7% | 40.5% | 10.5% | 4.3% | 2.0% | +1.0 | |
17  MIN | 42.632-53 | 59.9% | 38.1% | 41.9% | 9.2% | 3.7% | 1.7% | +0.8 | |
18  ATL | 41.131-51 | 53.3% | 35.3% | 36.4% | 8.9% | 3.7% | 1.3% | -0.1 | |
19  MIA | 40.530-51 | 50.2% | 32.7% | 36.3% | 8.1% | 3.3% | 1.1% | -0.3 | |
20  NOP | 37.628-48 | 35.1% | 17.6% | 40.9% | 3.5% | 1.1% | 0.5% | -1.3 | |
21 IND | 35.726-46 | 27.9% | 15.0% | 30.9% | 3.2% | 1.2% | 0.4% | -2.1 | |
22  MIL | 34.424-45 | 23.0% | 11.8% | 27.3% | 2.4% | 0.9% | 0.2% | -2.6 | |
23  CHI | 33.223-43 | 18.6% | 9.0% | 24.1% | 1.7% | 0.6% | 0.2% | -3.1 | |
24  DAL | 33.123-43 | 17.5% | 7.0% | 28.1% | 1.1% | 0.4% | 0.1% | -3.2 | |
25  MEM | 32.823-43 | 16.7% | 6.6% | 27.2% | 1.1% | 0.3% | 0.1% | -3.2 | |
26 LAC | 32.623-43 | 15.8% | 6.1% | 26.8% | 0.9% | 0.3% | 0.1% | -3.4 | |
27 BKN | 31.121-41 | 12.6% | 5.6% | 18.5% | 1.0% | 0.3% | 0.1% | -3.9 | |
28  UTA | 30.020-40 | 9.3% | 3.2% | 18.2% | 0.4% | 0.1% | 0.0% | -4.3 | |
29  WAS | 27.718-37 | 5.7% | 2.2% | 10.3% | 0.4% | 0.1% | 0.0% | -5.4 | |
30 SAC | 20.212-29 | 0.5% | 0.1% | 1.7% | 0.0% | 0.0% | 0.0% | -8.8 |
Wins range = the middle 80% of simulated seasons. Playoffs = made the 16-team bracket, through the play-in if needed. Percentages are shares of simulated seasons, rounded to 0.1.
Go deeper than the odds
Members get the full NBA research toolkit every night of the season: player game logs and trends, matchup and defense breakdowns, injury news and the daily research board.
How the simulator works
1 Start from last season. Every 2025-26 regular-season game margin is fit at once, so each team gets a rating in points per game that accounts for who it played and where (home court came out at 1.7 points).
2 Adjust for roster moves. Each player is valued by his box-score production per minute (Game Score) above a replacement-level player, converted to points of team margin (fit on 60 team-seasons). A team's current roster — from the official player directory, so trades and signings like LeBron James to Philadelphia count — is compared with the minutes it actually played last season. Players back from injury count too; rookies count as replacement level.
3 Regress toward average. Ratings carry 58% of last season and 50% of the roster adjustment. Those weights come from projecting 2025-26 from 2024-25 the same way: that cut the average rating error to 4.7 points vs 5.1 using last season alone.
4 Play the season 50,000 times. Each simulated season gives every team a random “true talent” error (±2.75 points), then plays all 1,230 games with a game-to-game spread of 12.5 points. Ties are broken at random.
5 Postseason. Seeds 7-10 play the play-in (7 vs 8, 9 vs 10, then loser vs winner); the bracket is best-of-seven with home court to the higher seed.
6 Updated nightly. Once games are played, results are locked in and each rating blends toward the team's 2026-27 form (after 20 games, this season counts as much as the preseason rating).
Limits: box scores miss most defense, so rosters built on defenders may be underrated; the official schedule lists 1200 games until the NBA Cup knockout dates are set, and the remaining 30 are simulated as neutral-site games. These are research estimates from historical data, not predictions of any single game.