NBA 2026-27 Projections
Projected wins for all 30 teams entering 2026-27, built from the final 2025-26 standings with standard one-third regression to the mean — plus a head-to-head matchup tool and the honest math on why last season’s 64-win team projects to 56, not 64. The real schedule arrives in mid-August; until then everything here is schedule-neutral.
Head-to-head matchup probability
Neutral floor, no home edge (there’s no schedule yet) — log5 of the two teams’ projected strength.
Probabilities, not predictions — and remember these are regular-season strength estimates, not playoff forecasts.
Power ratings & projected wins, 2026-27
Final 2025-26 record, regressed one-third toward .500 and projected over a neutral 82-game schedule.
| Team | 2025-26 | Proj. W | Proj. L | ||
|---|---|---|---|---|---|
| 1 | Oklahoma City Thunder | 64-18 | 56.3 | 25.7 | |
| 2 | San Antonio Spurs | 62-20 | 55.0 | 27.0 | |
| 3 | Detroit Pistons | 60-22 | 53.7 | 28.3 | |
| 4 | Boston Celtics | 56-26 | 51.0 | 31.0 | |
| 5 | Denver Nuggets | 54-28 | 49.7 | 32.3 | |
| 6 | New York Knicks | 53-29 | 49.0 | 33.0 | |
| 7 | Los Angeles Lakers | 53-29 | 49.0 | 33.0 | |
| 8 | Cleveland Cavaliers | 52-30 | 48.3 | 33.7 | |
| 9 | Houston Rockets | 52-30 | 48.3 | 33.7 | |
| 10 | Minnesota Timberwolves | 49-33 | 46.3 | 35.7 | |
| 11 | Toronto Raptors | 46-36 | 44.3 | 37.7 | |
| 12 | Atlanta Hawks | 46-36 | 44.3 | 37.7 | |
| 13 | Philadelphia 76ers | 45-37 | 43.7 | 38.3 | |
| 14 | Orlando Magic | 45-37 | 43.7 | 38.3 | |
| 15 | Phoenix Suns | 45-37 | 43.7 | 38.3 | |
| 16 | Charlotte Hornets | 44-38 | 43.0 | 39.0 | |
| 17 | Miami Heat | 43-39 | 42.3 | 39.7 | |
| 18 | Portland Trail Blazers | 42-40 | 41.7 | 40.3 | |
| 19 | LA Clippers | 42-40 | 41.7 | 40.3 | |
| 20 | Golden State Warriors | 37-45 | 38.3 | 43.7 | |
| 21 | Milwaukee Bucks | 32-50 | 35.0 | 47.0 | |
| 22 | Chicago Bulls | 31-51 | 34.3 | 47.7 | |
| 23 | New Orleans Pelicans | 26-56 | 31.0 | 51.0 | |
| 24 | Dallas Mavericks | 26-56 | 31.0 | 51.0 | |
| 25 | Memphis Grizzlies | 25-57 | 30.3 | 51.7 | |
| 26 | Sacramento Kings | 22-60 | 28.3 | 53.7 | |
| 27 | Utah Jazz | 22-60 | 28.3 | 53.7 | |
| 28 | Brooklyn Nets | 20-62 | 27.0 | 55.0 | |
| 29 | Indiana Pacers | 19-63 | 26.3 | 55.7 | |
| 30 | Washington Wizards | 17-65 | 25.0 | 57.0 |
Benchmark: NumberBench projection — final official 2025-26 standings (1,230 games; OKC 64-18 best overall, Detroit 60-22 best in the East; Knicks beat the Spurs in the Finals), regressed one-third to the mean, the same season-transition fraction our NFL and MLB models use. Schedule-neutral until the NBA releases the 2026-27 slate in mid-August.
How it’s calculated
Projected win% = .500 + (last season’s win% − .500) × ⅔. That one line is the whole model — deliberately. With no schedule, no summer-league box scores worth trusting, and rosters still moving, a transparent regression of real results beats a black box of guesses. The matchup tool combines two projected win percentages with the log5 formula, the standard method for head-to-head probability.
Model output, not a forecast — and not betting advice. Offseason trades and injuries are not priced in; this is a results-based starting line that the August schedule release and training camp will move.
Worked example
Oklahoma City finished 64-18, a .780 win rate. Regressing one-third of the way to .500: .500 + (.780 − .500) × ⅔ = .687, which over 82 games projects to 56.3 wins — still the league’s best, but nearly eight wins shy of a repeat. That gap is the honest price of health, close-game luck, and 29 other teams trying.
Frequently asked questions
Why does a 64-win team project to only 56 wins?
Regression to the mean. Extreme records are part true strength, part good fortune — health, close-game luck, opponents' shooting variance. Historically about a third of the distance between a team's record and .500 evaporates the next season on average, so the model regresses every team one-third of the way back. Oklahoma City's 64-18 (.780) becomes a projected .687, about 56 wins — still the best in the league, just honestly so.
Where is the strength of schedule?
It doesn't exist yet. The NBA releases its schedule in mid-August, so every team here is projected against a neutral, average-difficulty 82 games. Once the real schedule is out, the same ratings plug into a game-by-game projection exactly like our NFL page.
How does the matchup tool work?
It uses the log5 formula — the standard way to combine two teams' win percentages into a head-to-head probability. A .687 team versus a .500 team comes out to 68.7%; two equal teams are a coin flip. No home edge is applied since there's no schedule yet, so read it as a neutral-floor probability.
Are the 2025-26 records official?
Yes — final standings, all 1,230 games played: Oklahoma City led the league at 64-18, Detroit won the East at 60-22, and the New York Knicks beat the San Antonio Spurs in the Finals for their first championship since 1973.
Why do the defending champions project seventh?
The Knicks went 53-29 — a strong record, but five teams won more regular-season games, and playoff success doesn't feed a regular-season projection. Championships are won in April through June; this page only projects the 82-game season.
When will this page update?
When the 2026-27 schedule drops in mid-August, these ratings become game-by-game probabilities with home edge and rest factored in. Until then, treat the win totals as schedule-neutral central estimates.