NBA 2026-27 · season tips off October 20
Every team's projected record with a range, its playoff and play-in odds, and how our number compares with the market's win total. Built from rosters and the real schedule, played out 20,000 times. Pick any team for its full breakdown.
Eastern Conference · sorted by projected wins
Playoffs means a top-six seed; play-in is seeds seven to ten. Market is the win total implied by prediction-market prices. Gap is ours minus the market.
| # | Team | Wins | Range | Playoffs | Play-in | Market | Gap |
|---|---|---|---|---|---|---|---|
| 1 | Boston Celtics | 57 | 51 to 62 | 100% | 0.2% | 53.6 | +3.5 |
| 2 | Detroit Pistons | 54 | 48 to 59 | 98% | 1.7% | 49.3 | +4.9 |
| 3 | New York Knicks | 49 | 43 to 54 | 85% | 15% | 51.8 | -2.4 |
| 4 | Cleveland Cavaliers | 49 | 43 to 54 | 85% | 15% | 48.9 | +0.3 |
| 5 | Toronto Raptors | 46 | 40 to 51 | 62% | 36% | 48.4 | -2.1 |
| 6 | Atlanta Hawks | 45 | 40 to 51 | 59% | 39% | 44.3 | +1.7 |
| 7 | Orlando Magic | 43 | 37 to 48 | 35% | 60% | 45.1 | -2.0 |
| 8 | Charlotte Hornets | 42 | 37 to 48 | 32% | 63% | 37.7 | +5.1 |
| 9 | Miami Heat | 42 | 36 to 47 | 28% | 65% | 47.8 | -5.4 |
| 10 | Philadelphia 76ers | 40 | 34 to 45 | 14% | 71% | 49.9 | -9.8 |
| 11 | Milwaukee Bucks | 32 | 26 to 37 | 0.3% | 15% | 27.8 | +4.6 |
| 12 | Indiana Pacers | 30 | 25 to 36 | 0.1% | 8.9% | 47.1 | -16.2 |
| 13 | Brooklyn Nets | 30 | 24 to 35 | 0.1% | 6.6% | 26.5 | +3.6 |
| 14 | Chicago Bulls | 29 | 23 to 34 | 0.0% | 4.1% | 31.4 | -2.5 |
| 15 | Washington Wizards | 23 | 18 to 28 | 0% | 0.2% | 36.3 | -12.8 |
Western Conference · sorted by projected wins
Playoffs means a top-six seed; play-in is seeds seven to ten. Market is the win total implied by prediction-market prices. Gap is ours minus the market.
| # | Team | Wins | Range | Playoffs | Play-in | Market | Gap |
|---|---|---|---|---|---|---|---|
| 1 | Oklahoma City Thunder | 63 | 58 to 68 | 100% | 0% | 63.9 | -0.4 |
| 2 | San Antonio Spurs | 59 | 54 to 64 | 100% | 0.0% | 61.4 | -2.3 |
| 3 | Houston Rockets | 51 | 45 to 56 | 95% | 4.7% | 51.5 | -0.2 |
| 4 | Minnesota Timberwolves | 45 | 39 to 50 | 64% | 33% | 50.5 | -5.1 |
| 5 | Phoenix Suns | 44 | 38 to 49 | 54% | 41% | 40.1 | +4.0 |
| 6 | Golden State Warriors | 44 | 38 to 49 | 52% | 43% | 36.2 | +7.6 |
| 7 | Denver Nuggets | 43 | 38 to 49 | 51% | 44% | 49.4 | -5.6 |
| 8 | LA Clippers | 42 | 36 to 47 | 35% | 54% | 27.9 | +14.2 |
| 9 | Portland Trail Blazers | 41 | 36 to 47 | 32% | 56% | 42.1 | -0.2 |
| 10 | Los Angeles Lakers | 38 | 32 to 43 | 10% | 53% | 47.5 | -9.3 |
| 11 | Memphis Grizzlies | 37 | 31 to 42 | 6.8% | 46% | 32.0 | +5.2 |
| 12 | Dallas Mavericks | 32 | 27 to 37 | 0.6% | 14% | 35.8 | -3.2 |
| 13 | Sacramento Kings | 30 | 25 to 36 | 0.2% | 7.8% | 22.4 | +8.6 |
| 14 | New Orleans Pelicans | 28 | 23 to 33 | 0.0% | 2.6% | 30.6 | -2.1 |
| 15 | Utah Jazz | 25 | 20 to 31 | 0.0% | 0.7% | 39.9 | -14.0 |
Where we and the market disagree most
A gap is a difference of opinion, not a call that we are right. These are the teams where a roster-first view and the market see the season most differently.
Rosters, then the schedule, then 20,000 seasons
Team ratings fit from 6,153 completed games across the last three seasons are kept alongside as a comparison on each team page.
Tested on seasons it never saw
Better than the simple baseline, and still about 8 wins off for a typical team. Treat every number here as a wide estimate. We will grade this edition against the final standings and publish the results, misses included.
Common questions
A projection of the 2026-27 NBA regular season for all 30 teams: a win total with a range, playoff and play-in odds, and a comparison with the betting market's win total. Each team also has its own page with the last three seasons, offseason moves and the projected rotation.
We start from each roster: every player's on-court impact from his most recent full season, weighted by projected minutes, turned into a team strength. We then play the real 2026-27 schedule 20,000 times, game by game, and count wins. The middle of those seasons is the projection; the range covers the middle 80%.
Tested on past seasons it was not trained on (60 team-seasons), the typical miss was 8.3 wins per team, against 9.0 for last season's record pulled toward average. That is better than the simple baseline and still a wide miss, so read every number as an estimate with real uncertainty.
The market number on this page comes from prediction-market prices, read as the win total the market sees as a coin flip. Ours is built only from rosters and the schedule. Where the two disagree we show the gap; a gap is a difference of opinion, not a prediction that we are right.
Injuries during the season, trades after the build, and in-season form. Roster impact uses box-score plus-minus, which does not separate a player from his teammates, and players with no NBA history count as average.
This edition was built October 19, 2026, before opening night on October 20. We grade it against the real standings as the season plays out and publish where it was close and where it missed.
Built Oct 19, 2026. Estimates are ours and can be wrong.