01Line score
| Team | Q1 | Q2 | Q3 | Q4 | Final |
|---|---|---|---|---|---|
| Hawks | 22 | 26 | 24 | 25 | 97 |
| Knicks | 35 | 29 | 26 | 36 | 126 |
02How the lead moved
- NYK +32 · ATL +2
- 1 / 3
- NYK 12-0 run, Q4 8:04-5:59
- NYK 10-0 run, Q4 2:18-0:18
- NYK 12-2 run, Q1 3:19-0:03
03Four factors and why they won
| Factor | ATL | NYK | Edge |
|---|---|---|---|
| Effective FG%Shooting, with 3s worth 1.5 makes | 52.4% | 63.9% | NYK |
| Turnover %Turnovers per 100 plays | 10.0% | 10.5% | ATL |
| Offensive rebound %Share of own misses rebounded | 11.1% | 26.7% | NYK |
| Free throw rateFree throws made per shot | 12.0% | 31.6% | NYK |
Why the Knicks won
- Made more 2s, 34-24, worth about 20 points.
- Made more free throws, 25-10, worth about 15 points.
- Won the offensive glass 8-5, worth about 3 points.
04Top performers
05Where they shot from
Hawks
| Zone | FG | Share | Season share | FG% vs season |
|---|---|---|---|---|
| At the rim | 14-17 | 20% | 30% | 82% vs 65% |
| Paint | 8-14 | 17% | 20% | 57% vs 45% |
| Midrange | 2-10 | 12% | 8% | 20% vs 42% |
| Corner 3 | 8-16 | 19% | 12% | 50% vs 39% |
| Above-the-break 3 | 5-26 | 31% | 31% | 19% vs 36% |
Knicks
| Zone | FG | Share | Season share | FG% vs season |
|---|---|---|---|---|
| At the rim | 23-32 | 41% | 29% | 72% vs 66% |
| Paint | 7-14 | 18% | 17% | 50% vs 45% |
| Midrange | 4-7 | 9% | 11% | 57% vs 45% |
| Corner 3 | 3-5 | 6% | 14% | 60% vs 42% |
| Above-the-break 3 | 8-21 | 27% | 29% | 38% vs 35% |
Share is the percent of the team's shots from that zone. Season numbers are the team's full regular season, also for playoff games. Blue means 5 or more points above its season number, red means 5 or more below. See every shot on the game shot chart
How this works
Where the numbers come from. The line score and every scoring change come from ESPN's play-by-play, loaded the morning after the game. The margin chart plots the home team's lead after each score against game time. Team totals and player lines come from Courtside's box scores.
Runs. A run starts with one team's basket and ends with its last basket before the other team scores more than 2 points. It needs at least 8 points, and we rank runs by net points, so a 14-0 run beats a 15-2 run.
Why they won. Simple rules, not a model. Each extra turnover forced and each extra offensive rebound is one more possession, worth about 1.1 points because NBA teams score about 1.1 points per possession. Each extra 3 made is worth 3, each extra 2 made is worth 2 and each extra free throw made is worth 1. We show the three biggest gaps in the winner's favor worth at least 2 points. They overlap (an extra possession often becomes one of those extra makes), so they don't add up to the final margin.
Four factors are Dean Oliver's: effective FG% (a 3 counts as 1.5 makes), turnovers per 100 plays, the share of your own misses you rebound, and free throws made per shot. Game Score is John Hollinger's one-number box score: points plus credit for makes, rebounds, assists, steals and blocks, minus missed shots, fouls and turnovers. About 10 is a solid game and 30 is a great one.
What we checked. Across 1,322 loaded games, the ESPN line score added up to our final score in 100.0% of them and the play-by-play reached the final score on its own in 99.2% (for the rest, a few late scores were missing, so the chart jumps to the official final at the buzzer). To test the possession value, we fit final margin on the turnover, offensive rebound, shooting (effective FG%) and free throw gaps over 1,315 games: each turnover was worth 1.01 points and each offensive rebound 1.05 (the fit explains 95% of margins). That is a little under the 1.1 we use, so those two bullets can run about 5 to 10% high.
What it can't see. A live-ball turnover that becomes a fast-break dunk costs more than one out of bounds; the rules value them the same. Garbage time, foul trouble, injuries and matchups are not in these numbers. Shot zones only count shots ESPN logged with a location, and the season numbers cover the whole regular season, including games played after this one.