Milwaukee Bucks at Boston Celtics, NBA forecast
Win probability
Starts from the pregame forecast, then moves with every score and the time left. Above the line favors BOS; hover for each play.
Game center
From the ESPN box score.
Top performers
Bucks- Damian LillardPoints33
- Damian LillardAssists7
- Giannis AntetokounmpoRebounds10
Celtics- Jaylen BrownPoints30
- Derrick WhiteAssists8
- Jayson TatumRebounds8
Last plays
- Q40.0End of Game108–119
- Q40.0End of the 4th Quarter108–119
- Q417.0Andre Jackson Jr. makes driving dunk108–119
- Q420.1Andre Jackson Jr. defensive rebound106–119
- Q424.1Drew Peterson misses 26-foot three point jumper106–119
- Q439.1Celtics defensive team rebound106–119
- Q442.0Tyler Smith misses 26-foot three point jumper106–119
- Q453.2Andre Jackson Jr. defensive rebound106–119
- Q457.1Neemias Queta misses 7-foot two point shot106–119
- Q41:16AJ Green makes 25-foot three point shot (Andre Jackson Jr. assists)106–119
Turning points
The 5 plays that swung the win chance most, each judged at its own moment: the chance with the score just before it against just after.
- Q311:03Jrue Holiday makes 25-foot three pointerMIL 58 – 56 BOS · BOS 55% → 64% +9BOS 55% → 64%+9 pts
- Q37:13Al Horford makes 26-foot three point jumper (Jrue Holiday assists)MIL 67 – 70 BOS · BOS 68% → 76% +8BOS 68% → 76%+8 pts
- Q35:03Jrue Holiday makes 26-foot running pullup jump shotMIL 75 – 75 BOS · BOS 57% → 67% +10BOS 57% → 67%+10 pts
- Q31:58Derrick White makes 25-foot three pointerMIL 80 – 82 BOS · BOS 62% → 72% +10BOS 62% → 72%+10 pts
- Q31:24Derrick White makes 27-foot three pointerMIL 80 – 85 BOS · BOS 72% → 81% +9BOS 72% → 81%+9 pts
The pick
Reconstructed from what was known before the game: the ratings at that point and the closing line.
Celtics 77%
Forecast Elo ratings alone
How often that comes true: NBA favorites the forecast gave 75–80% have won 75.8% of 293 graded games.
- Market
- BOS -9.5
- ML MIL +310 / BOS -400
- Pre-line forecast
- BOS 74%
- Elo + market-implied
- Elo ratings
- BOS -7
- BOS 75%
- Ratings vs. market
- −1.3 pts
- BOS: Elo 75.3%, market 76.6%
- Fair odds
- BOS -328
- MIL +328 · the forecast as odds, no margin
- Pregame projection
- MIL 111.5 – 121.0 BOS
- Posted spread, posted total 232.5
How the number is built
Elo inputs in rating points (higher favors that team; 25–28 Elo is about a point of spread in football and basketball), then how the pieces combine.
| Factor | MIL | BOS | Why |
|---|---|---|---|
| Team rating | 1505 | 1633 | Elo entering the game; 1500 is league average |
| Home field | — | +36 | Worth 36 Elo in the NBA |
| Rest | −40 | — | MIL on the second night of a back-to-back |
| Lineups | +2 | −8 | MIL: Usual rotation · BOS: Weaker than its recent lineups (from the box score) |
| Elo ratings | BOS +194 Elo | BOS 75.3% | |
| Market-implied ratings | BOS 73.9% | Team strength as past closing lines priced it, same game context | |
| Pre-line forecast | BOS 74.1% | 10% Elo, 90% market-implied (log-odds blend) | |
| Forecast | BOS 76.6% | Follows the market (vig removed): in testing, adding the ratings made it less accurate | |
Tale of the tape
This season coming into the game, regular season, each number with its league rank (the longer the bar, the higher). Home and road: the host's record at home against the visitor's on the road. Offense and defense are points per 100 possessions, which sets aside how fast a team plays.
| MIL | Stat | BOS |
|---|---|---|
| 8th of 30116.0 | Points scored | 126.02nd of 30 |
| 24th of 30119.0 | Points allowed | 109.79th of 30 |
| 17th of 30−3.0 | Margin | +16.34th of 30 |
| 8th of 30112.5 | Offense rating | 127.21st of 30 |
| 24th of 30115.5 | Defense rating | 110.714th of 30 |
| 12th of 30103.4 | Pace | 99.228th of 30 |
| 7th of 251–1 | Road vs home | 1–01st of 30 |
Who's playing
Who sat out, from the box score, and what each absence cost against the lineups the team had been playing.
Bucks
+2 EloFull strength: nobody who matters sat out.
Celtics
−8 EloFull strength: nobody who matters sat out.
Rating history
Each team's Elo over the last two seasons. Hover or drag to compare on any date.
Recent form
Last five results, and whether the forecast hit.
Head to head
MIL 2–2 BOS in their last 4 meetings; the forecast called 2 of 4.