Forecast NBA 2026-27
● Updated · 7 Nov
Boston gains 4.0 projected wins since the 7 Nov update: from 53.2 to 57.2.
Derrick White
Since the 7 Nov update, 24 teams have moved more than 0.1 wins.
Yang Hansen is the player who rises the most: his projected impact improves by 5.9 points per 100 possessions from 2025-26.
The player who drops the most is Shai Gilgeous-Alexander: his projected impact falls by 2.7 points.
Impact: the points per 100 possessions a player adds (+) or costs (−) his team compared with an average NBA player, who is 0.
The team that rises the most from 2025-26 is Washington: from 17 wins to 33.8 projected.
The team that drops the most is Charlotte: from 44 to 22.9.
Projections run from 17.7 to 60.2 wins; actual 2025-26 records ran from 17 to 64. A projection is an average, so it sits closer to the middle.
The East is led by Boston: 57.2 wins, 0.5 more than Detroit.
Eastaverage 41
Behind Houston, Oklahoma City, Minnesota and LA Lakers are within 1.8 wins of each other.
Westaverage 41
The forecast uses the rosters of 7 Nov, with the summer signings and trades. ESPN and NBA.com were last checked on 7 Nov, and the forecast recalculates itself when a roster changes.
What changes if a player plays more or fewer games?
Choose one or more players from the same team and set how many of the 82 regular-season games each one plays.
Rookies
No rookie projects to be above an average player this season; the closest is Yaxel Lendeborg (Golden State, No. 11 pick): −0.7.
The next game is Philadelphia at Chicago on 8 Nov: Philadelphia start with a 63% chance.
Tyrese Maxey has a 51% chance of reaching 50 points in a game of the 2026-27 regular season.
What are the chances that Tyrese Maxey scores 50 or more points in a game this regular season?
51%With the minutes and games we project for Tyrese Maxey (about 67, at 34 minutes) and his 2025-26 per-minute rate, reaching 50 points in a game would happen in about 5 of every 10 seasons like this one.
He has done it 4 times in 449 career games, playoffs included; his best: 54 points.
Data as of: box scores 6 Nov 2026 · projected minutes 7 Nov 2026 · schedule 7 Nov 2026
Prediction record
No predictions registered yet.
With the final rosters, Forecast missed by 7.3 wins per team on average; repeating the previous season's record missed by 8.4.
average error in wins · lower is better · 210 team-seasons, 2019-2025
With October rosters, Forecast's error has not been measured yet and may be larger.
For players, it calls the direction (up or down) right 67.7% of the time (2008-2023).
Game accuracy for this season starts counting on 8 Nov.
The Analyst answers these questions with SportiveWorld's data
How it's calculated and its limits
forecast interval
The seed is an ordering by expected wins, without play-in or tiebreakers. Injuries are received but the model does not consume them: they do not move the number.
It missed by 7.3 wins per team on average; repeating the previous season's wins missed by 8.4. It was measured season by season, with the model fitted only on earlier seasons (2019-2025, 210 team seasons). Evaluated with the rosters that actually played (information known after the fact): with October rosters, the real coverage and error have not been measured yet.
Players
Impact: the points per 100 possessions a player adds (+) or costs (−) his team compared with an average NBA player, who is 0. A player's impact is measured in points per 100 possessions compared with an average player, who is 0 (−13 to +13).
- right direction: 67.7 % · 2008-2023
- inside the range: 80.2 % · 2008-2023
- Of the 20 players it expects to improve most, 63% really improved (2008-2023 seasons).
- Of the 10 it expects to decline most, 89% really declined (2008-2023 seasons).
Measured over 16 real seasons (2008-2023, n=5,851): out of every 100 players, 80 finished inside the range we would have given.
Game
In games, the error of its probabilities (Brier: the lower, the better) is 0.227; counting only home advantage it would be 0.247. It was measured over 10,749 games, from 2017 to 2025, season by season and without seeing the future. The game model does not take rest, back-to-back games, travel or pace into account.
No resolved games with a sealed pre-game call yet: the prospective record starts with the season.
What Forecast covers in this competition
- Game: Available — Does not read the day's injury report. Publishes a point per game, no band: calibration is measured in the track record. No resolved games with a sealed pre-game call yet: the prospective record starts with the season.
- Season: Available — The seed is an ordering by expected wins, without play-in or tiebreakers. Injuries are received but the model does not consume them: they do not move the number. The wins interval states in each snapshot which method it is and its measured coverage; the evaluated one (residuals, 80% with ex-post rosters) is served after the re-issue.
- Playoffs & title: Not yet modeled — No evaluated model: no probability is claimed.
- Players: Available — No projected stat line (points, rebounds, assists).
- Awards: Not yet modeled — No model: nothing to claim.
- Track record: Available — No resolved games with a sealed pre-game call yet: the prospective record starts with the season.
Data dates: source 7 Nov · input 7 Nov · model 7 Nov · published 8 Nov.
This is reading the data, not a bet: we work with no bookmakers and show no odds. And a game turns on an injury, a hot night or a bad call.