Minutes fall at both ends. Points fall at one.
a game, against each player's own season average · by the team's final margin · NBA 2025-26 regular seasonShow the numbers
| Final margin, the player's team | Games | Minutes a game | Against own average | Under 30 minutes | Points a game | Against own average | Reached own points average | Points a minute |
|---|---|---|---|---|---|---|---|---|
| Lost by 20+ | 384 | 29.2 | −3.8 | 180 of 384 (46.9%) | 15.6 | −4.3 | 96 of 384 (25.0%) | 0.54 |
| Lost by 11 to 19 | 583 | 32.7 | 0.0 | 121 of 583 (20.8%) | 18.1 | −1.5 | 229 of 583 (39.3%) | 0.55 |
| Within 10 | 2,286 | 34.9 | +1.8 | 249 of 2,286 (10.9%) | 21.1 | +1.1 | 1,250 of 2,286 (54.7%) | 0.60 |
| Won by 11 to 19 | 771 | 32.5 | −0.4 | 162 of 771 (21.0%) | 20.5 | +0.5 | 401 of 771 (52.0%) | 0.63 |
| Won by 20+ | 638 | 29.1 | −3.8 | 332 of 638 (52.0%) | 19.5 | −0.4 | 288 of 638 (45.1%) | 0.67 |
Why?
Blowout risk is the worry that a big lead, or a big deficit, sends a star to the bench for the fourth quarter. It gets said about every lopsided game, one game at a time. I wanted to know what a whole season of final scores says about it.
Here is every game of the 2025-26 NBA regular season, by how far apart it finished:
281 of 1,230 games ended 20 or more apart
NBA 2025-26 regular season · games, by final margin in pointsShow the numbers
| Final margin | Games |
|---|---|
| 1–4 | 244 |
| 5–9 | 286 |
| 10–14 | 243 |
| 15–19 | 176 |
| 20–24 | 111 |
| 25–29 | 74 |
| 30–34 | 53 |
| 35–39 | 23 |
| 40+ | 20 |
281 of the 1,230 games (22.8%) were decided by 20 or more points, and 96 by 30 or more. The widest was 55.
The group is every regular: the 69 players who averaged 30 or more minutes over 40 or more games, 4,662 games between them. Each game is set against that player's own season averageThe player's points, or minutes, a game across the whole regular season. Each game's number minus it is derived., the same way the back-to-backs count measured tired legs on the same season.
The minutes go either way
Figure 1 is the answer in one picture. In games won by 20 or more, regulars played 3.8 minutes fewer than their own average. In games lost by 20 or more, 3.8 fewer as well. In games within 10, they played 1.8 more. A 20-point win and a 20-point loss cost the same minutes.
Count the short nights and it is the same. Every regular averaged 30 or more minutes; they played under 30 in 52.0% of the big wins and 46.9% of the big losses, against 10.9% of the games within 10.
The points did not follow. In the big wins, regulars scored 0.4 fewer than their own average, about a normal night. In the big losses, 4.3 fewer. Blowouts cost minutes either way. Only the losses cost points.
The kind that cost almost no points was also the more common one: regulars played in 638 wins by 20 or more and 384 losses by 20 or more.
Another way to count it: how often a regular scored at least that player's own season average.
Losing by 20 is where the points went missing
% of games a regular scored at least that player's own season average · by the team's final marginEvery game: 48.6%, 2,264 of 4,662
Show the numbers
| Final margin, the player's team | Reached own points average | Games | % |
|---|---|---|---|
| Lost by 20+ | 96 | 384 | 25.0% |
| Lost by 11 to 19 | 229 | 583 | 39.3% |
| Within 10 | 1,250 | 2,286 | 54.7% |
| Won by 11 to 19 | 401 | 771 | 52.0% |
| Won by 20+ | 288 | 638 | 45.1% |
| Every game | 2,264 | 4,662 | 48.6% |
Within 10, regulars reached their own average in 1,250 of 2,286 games (54.7%). In wins by 20 or more, 288 of 638 (45.1%), close to the 48.6% of every game. In losses by 20 or more, 96 of 384 (25.0%): about 1 game in 4.
Two kinds of Shai Gilgeous-Alexander night
No regular played in more blowout wins than Shai Gilgeous-Alexander: 23 wins by 20 or more, with Kon Knueppel next on 18. In those 23, Gilgeous-Alexander scored points in minutes a game. In the 25 games within 10, 34.8 points in 37.3 minutes. The two side by side:
Shai Gilgeous-Alexander's two kinds of night
a game, 2025-26 regular season · wins by 20+ against games within 10 · dashed: the season's averageShow the numbers
| Games | Minutes a game | Points a game | Points a minute | Under 30 minutes | |
|---|---|---|---|---|---|
| Won by 20+ | 23 | 29.5 | 28.9 | 0.98 | 14 of 23 |
| Within 10 | 25 | 37.3 | 34.8 | 0.93 | 0 of 25 |
| Season | 68 | 33.3 | 31.1 |
Gilgeous-Alexander played under 30 minutes in 14 of the 23 big wins, and in none of the 25 close games. Per minute it was the same scorer: 0.98 points a minute in the big wins, 0.93 in the close games. The points that went missing were the minutes that went missing.
The NBL, the same shape
The NBL's 2025-26 regular season had 165 games, and 46 of them (27.9%) were decided by 20 or more. An NBL game is 40 minutes, not 48, so a regular here is 25 or more minutes a game, the same share, over 17 or more of 33 games: 31 players, 921 games.
The NBL: minutes at both ends, points at one
a game, against each player's own season average · by the team's final margin · NBL 2025-26 regular seasonShow the numbers
| Final margin, the player's team | Games | Minutes a game | Against own average | Points a game | Against own average | Reached own points average | Points a minute |
|---|---|---|---|---|---|---|---|
| Lost by 20+ | 105 | 27.0 | −2.5 | 11.7 | −2.8 | 32 of 105 (30.5%) | 0.43 |
| Lost by 11 to 19 | 131 | 29.9 | +0.7 | 14.0 | −0.8 | 61 of 131 (46.6%) | 0.47 |
| Within 10 | 441 | 30.6 | +1.4 | 15.5 | +0.5 | 221 of 441 (50.1%) | 0.50 |
| Won by 11 to 19 | 124 | 28.9 | −0.4 | 16.7 | +1.4 | 69 of 124 (55.6%) | 0.58 |
| Won by 20+ | 120 | 25.4 | −3.3 | 15.8 | −0.1 | 59 of 120 (49.2%) | 0.62 |
Regulars played 3.3 fewer minutes in wins by 20 or more and 2.5 fewer in losses by 20 or more. Points: −0.1 in the big wins, −2.8 in the big losses. They reached their own average in 59 of 120 big wins (49.2%) and 32 of 105 big losses (30.5%). The same shape, on a smaller sample.
Which came first?
A final margin is the result of the points as much as a reason for them. A star's quiet night can be why a game was lost by 20, and a hot first half can be why a win got out of hand. The minutes fell the same either way; the scoring rate didn't. It rose with the margin, band by band:
| Final margin | Points a minute |
|---|---|
| Lost by 20+ | 0.54 |
| Lost by 11 to 19 | 0.55 |
| Within 10 | 0.60 |
| Won by 11 to 19 | 0.63 |
| Won by 20+ | 0.67 |
The middle bands point the same way. In losses by 11 to 19, regulars played their usual minutes (0.0 against their own average) and still scored 1.5 fewer points. There, the shortfall wasn't minutes.
These numbers can't separate the two. They are final margins only: Punt holds no NBA play-by-play, so it can't see when a game got out of hand, or the score when a regular sat down. And they are one season: 2025-26 is the only finished NBA season in Punt's database, so no gap here has been tested against another year. I want to count the same games again with play-by-play, to see whether the bench came before the margin or after it.
Summary
- Minutes v own average
- −3.8won by 20+−3.8lost by 20++1.8within 10
- Points v own average
- −0.4won by 20+−4.3lost by 20+
- Reached own points average
- 54.7%within 10, 1,250 of 2,28625.0%lost by 20+, 96 of 384
The working
Definitions, sources and when the numbers were read
- NBA season
- The 2025-26 regular season: 1,230 games, 21 October 2025 to 12 April 2026, every club's 82. Every 2025-26 game in Punt's database reads as regular season, play-in and playoffs included, so the season is cut at 12 April 2026, the night every club reached 82. The NBA Cup final (16 December 2025) is left out too: it is not one of the 82, the same cut as the back-to-backs article.
- NBL season
- The 2025-26 regular season: 165 games, 18 September 2025 to 20 February 2026, every club's 33. Finals left out.
- A game
- A box-score row with minutes above zero. A traded player is one player across clubs.
- Regulars
- NBA: 30+ minutes a game over 40+ games, 69 players. NBL: 25+ minutes over 17+ games, 31 players; 25 of the NBL's 40 minutes is the same share as 30 of the NBA's 48, and 17 of 33 games is about half a season, as 40 of 82 is.
- Final margin
- The player's team's final score minus the opponent's. Bands: lost by 20 or more, lost by 11 to 19, within 10 either way, won by 11 to 19, won by 20 or more. Final scores only: Punt holds no NBA play-by-play, so the article can't see when in a game the margin opened up, or who was on the court after it.
- Against own average
- Each game's minutes, or points, minus the player's own regular-season average. Derived.
- Reached own average
- Points in the game at or above the player's own season average. Every game: 2,264 of 4,662 (48.6%) in the NBA, 442 of 921 (48.0%) in the NBL. Derived.
- Under 30 minutes
- A game in which a regular played fewer than 30 minutes. NBA only. Derived.
- Points a minute
- A band's points divided by its minutes. Derived.
- Shai Gilgeous-Alexander
- 68 regular-season games, 31.1 points in 33.3 minutes a game. 23 wins by 20 or more, 25 games within 10, 2 losses by 20 or more. Most wins by 20 or more of the 69 regulars; Kon Knueppel next on 18.
- Read
- 26 October 2026, from Punt's database, in a read-only transaction.
- Sources
- NBA box scores: BALLDONTLIE. NBL results: the competition's own record at nbl.com.au.
- Not in this article
- When in a game a regular sat, and why (the score, foul trouble, an injury, rest); quarter-by-quarter scores; the play-in and the playoffs; any NBA season but 2025-26, the only finished one in Punt's database; any stat but minutes and points; prices. No game still to be played is predicted here.
- Sources and terms
- Every source Punt uses, with its licence, is on Credits. The terms are defined in the Glossary, and the methodology page sets out where every number comes from and the rules it follows.