BOS
GSW
Boston Celtics @ Golden State Warriors
Any-day matchup · ratings as of
Our pick
BOSBoston Celtics
57% to win
A probability, not a promise. A 57% pick still loses about 4 times in 10.
Advantages Why Celtics can win
- Net rating (points per 100 possessions): Celtics 4th (+7.5), Warriors 19th (-0.7).
- Recent form: average margin over the last five games, Celtics +8.0, Warriors -1.8.
- The Celtics' shooting (55.1% eFG%, 10th in the NBA) meets a Warriors defense that ranks 24th.
- Recent scoring margins: worth about 2.8 points to the Celtics in our ratings.
- The Celtics' offense (116.6 points per 100 possessions, 2nd in the NBA) meets a Warriors defense that ranks 17th.
Disadvantages Why Celtics may not
- Upset odds: the Warriors win 4,310 of 10,000 simulations (43%). A 57% pick still loses about 4 times in 10.
- Home court: the Warriors play at home, worth about 2 points in today's NBA.
- Turnovers forced per 100 possessions: Warriors 5th (15.7), Celtics 27th (12.8).
- Range of outcomes: in the worst 10% of simulations the Celtics lose by 14 or more.
- Body clock: the Celtics cross 3 time zones.
The line and our numbers
| Market | AZMango | |
|---|---|---|
| Spread | No line yet | BOS -2.6 |
| Total points | n/a | 228.6 |
| Win chance | BOS n/a · GSW n/a | BOS 57% · GSW 43% |
| Fair moneyline | BOS -132 · GSW +132 |
No betting line for this pairing. Odds change all the time, so check a sportsbook before acting on any number here.
10,000 simulated games, possession by possession
- Typical score: Warriors 113, Celtics 116.
- 80% of simulations land between Celtics by 19 and Warriors by 14.
Side by side
| BOS | GSW | |
|---|---|---|
| 3-1 | Record | 2-3 |
| +7.5 4 | Net rating | -0.7 19 |
| 116.6 2 | Offensive rating | 112.2 16 |
| 109.1 5 | Defensive rating | 112.9 17 |
| 55.1% 10 | Effective FG% | 55.0% 12 |
| 52.2% 4 | Opponent eFG% | 55.6% 24 |
| 12.7 2 | Turnovers per 100 | 15.5 27 |
| 29.0% 6 | Offensive rebound % | 26.3% 11 |
| 36.5% 8 | 3-point % | 35.7% 16 |
| 98.1 30 | Pace | 102.3 17 |
Small numbers are league ranks (1 = best of 30; pace 1 = fastest). Ratings are points per 100 possessions (2025-26 and 2026-27 seasons).
Last 5 meetings
| Date | Season | Score | Winner |
|---|---|---|---|
| 2026-03-18 | 2025-26 | GSW 99 @ BOS 120 | BOS |
| 2026-02-19 | 2025-26 | BOS 121 @ GSW 110 | BOS |
| 2025-01-20 | 2024-25 | BOS 125 @ GSW 85 | BOS |
| 2024-11-06 | 2024-25 | GSW 118 @ BOS 112 | GSW |
| 2024-03-03 | 2023-24 | GSW 88 @ BOS 140 | BOS |
No betting line exists for a game that isn't scheduled, so this uses our own power ratings (in the backtest they pick about 66% of winners, against about 68% for the closing line), with both teams at full strength.
Basis: power ratings only (no betting line yet). 10,000 simulated games, checked against a walk-forward backtest of 10,227 NBA games since 2018-19. Model version 1. How it works