33-18 for my model, and Boston won by 21 a game
Three of my four spotlight teams went unbeaten this week, and Boston did it by an average of 21.3 points a game. My model had a solid week, apart from one Saturday game in Minnesota that it got badly wrong.
The model’s week
| This week | Winners called | Margin miss (avg) | Season so far |
|---|---|---|---|
| Mon Nov 9 – Sun Nov 15 | 33 of 51 (65%) | 12.9 pts | 132 of 197 |
My model picked 33 of 51 games correctly, 65%, which is a little under its season mark of 132 of 197. The average miss on the point margin was 12.9, against 11.7 for the whole season, so the week was slightly rougher than usual.
The best call was Houston over Utah on Friday, Nov. 13. I had the Rockets at 88% and they won 125-118. The worst was Saturday, Nov. 14, when I gave Minnesota 88% at home against Chicago and the Bulls won 117-106. That one was off by 28.5 points on the margin, and I can’t explain it away. The closest margin was Sunday’s Knicks win in Toronto: I picked the Raptors at 50%, so the winner was wrong, but my margin was off by only 0.2 points.
Team spotlight
Los Angeles Lakers
This week 3-0 (@ ATL W 124-114, @ MIL W 123-112, @ PHX W 131-123) · season 12-2 · net rating +8.1 (6th of 30)
Eight straight wins is the story, but the rating is more measured than the record. A net rating of 8.1 ranks sixth in the league, so the Lakers are playing well without dominating. A 9.7-point average margin this week fits that picture. My model likes their next game, at home against Charlotte on Tuesday, Nov. 17, at 77%.
New York Knicks
This week 2-2 (vs MIL W 117-85, vs CLE L 107-115, @ IND L 104-106, @ TOR W 112-108) · season 7-6 · net rating +6.3 (8th of 30)
A 2-2 week hides a plus-6.5 average margin, and most of that comes from the big win over Milwaukee. The two losses were close, a combined 10 points. A 7-6 record sits below a net rating that ranks eighth, so the results have trailed the team’s play.
The Knicks meet Toronto again on Tuesday, Nov. 17, and my model picks them at 62%, even after missing on this week’s game between the two.
Houston Rockets
This week 4-0 (vs DEN W 110-100, vs GSW W 115-108, vs UTA W 125-118, @ WAS W 147-119) · season 11-3 · net rating +10.7 (4th of 30)
Four wins in four games, each by at least seven points, and a 13.0 average margin. The 147-119 win in Washington pushed that average up, but the team was already ahead in the other three. A net rating of 10.7 ranks fourth, so this looks like a team playing to its level.
Monday, Nov. 16, in Miami is a tougher test. My model leans Houston at 60% but sees the margin as even. Wednesday at Indiana looks cleaner at 79%.
Boston Celtics
This week 3-0 (@ PHI W 123-101, @ TOR W 98-93, vs DAL W 131-94) · season 11-2 · net rating +13.9 (1st of 30)
No team had a stronger week. Boston won three times by an average of 21.3 points, and the 131-94 win over Dallas ended it. That is well above their season net rating of 13.9, which already ranks first in the league. Six wins in a row now.
My model has them at 73% against Orlando on Monday, Nov. 16. The Wednesday game at Denver is much closer, nearly a coin flip.
Who moved
Chicago made the biggest jump in the chart. The Bulls averaged +4.7 this week against -11.3 for the season, a swing of 16 points, and they went 2-1, which includes the win that cost my model its worst pick. Orlando and Miami also climbed, both going 3-0. At the other end, New Orleans lost all four games by an average of 31.0, far below its season margin of -9.1. Dallas fell 18.3 points from its season norm, and Toronto went 0-4 after a season that had been slightly positive.
The week ahead
Next week’s table opens Monday, Nov. 16 with Orlando at Boston and Houston at Miami. The live picks page has every game as soon as my model publishes it.
| Day | Game | Model's pick | Win chance | Projected margin |
|---|---|---|---|---|
| Mon Nov 16 | ORL @ BOS | BOS | 73% | BOS by 9 |
| Mon Nov 16 | HOU @ MIA | HOU | 60% | even |
| Tue Nov 17 | NYK @ TOR | NYK | 62% | NYK by 6 |
| Tue Nov 17 | CHA @ LAL | LAL | 77% | LAL by 15 |
| Wed Nov 18 | HOU @ IND | HOU | 79% | HOU by 10 |
| Wed Nov 18 | MEM @ CLE | CLE | 51% | MEM by 4 |
| Wed Nov 18 | GSW @ DAL | GSW | 50% | GSW by 4 |
| Wed Nov 18 | BOS @ DEN | DEN | 56% | BOS by 2 |
| Sun Nov 22 | OKC @ MIA | OKC | 50% | MIA by 2 |
Every game, updated every 30 minutes until tip-off: today’s predictions. Team ratings and form: the team dashboard.
I’ll keep showing every pick, right or wrong, and I’ll see whether Chicago’s jump lasts.
How these numbers work: the picks come from my XGBoost models, which use team statistics (recent form, ratings, rest, travel), not player availability. Each pick locks at tip-off.