The other 49ers get no more sacks with Nick Bosa on the field
So what does the defense get when he comes back? Across 3,154 dropbacks, the sack column holds about a sixth of the answer.
The case for a great pass rusher has two halves. He gets his sacks, and he frees everyone else to get theirs, because no line can double-team him and block the other three men straight up.
The second half does not appear in Nick Bosa's numbers. Across his five seasons of 8 or more sacks, the rest of the 49ers sacked quarterbacks on 5.1 percent of dropbacks with him on the field and 5.2 percent with him off it. n = 3,154.
That matters this week because the 49ers are about to find out what they have been missing. Bosa practiced on a limited basis Wednesday for the first time since he strained his calf on Sept. 24. He has missed five games, and he has played five of San Francisco's last 24. The defense has 15 sacks in seven games, 5.6 per 100 dropbacks, which ranks 20th. Las Vegas comes to Santa Clara on Sunday. The natural place to look for his return is the sack column, and the sack column is the wrong place.
The team gains about half of a star's sacks#
The raw tie flatters his teammates. Third and fourth downs are when everyone's sacks come, and they are 35 percent of the dropbacks Bosa plays and 15 percent of the ones he sits. Compare the same season and the same kind of down, and the rest of the 49ers fall by about 1.2 sacks per 100 dropbacks when he is on the field.
I ran the same test on every rusher in the league with a season of 8 or more sacks since 2019, which is 228 seasons, 103,925 dropbacks with the rusher on the field and 41,206 with him off it.
| Change per 100 dropbacks, on the field against off | 228 seasons of 8+ sacks | Bosa, five seasons |
|---|---|---|
| His own sacks | +2.2 | +2.5 |
| His teammates' sacks | -1.1 | -1.2 |
| The team's sacks | +1.1 | +1.3 |
| His teammates' quarterback hits | -2.6 | -1.4 |
Bosa is the league in miniature. Teammates' sack rates rose in 72 of the 228 seasons. The team's rate rises by about half of the star's own rate, for the league and for him, and that half is the number to repeat at a bar.
For the league, on a sample that size, this is a real effect. For Bosa alone it is only suggested: his teammates' change has a 90 percent range of -3.1 to +0.7, and zero sits inside it.
Here is what this does not show. It does not show that he fails to help the men beside him. A quarterback goes down once on a play, so a sack credited to Bosa may be one a teammate was about to get. Real help and that displacement could cancel each other in this table, and the data cannot separate them yet. What the table does measure is the net, and on his usual workload the net is about four-tenths of a sack a game.
The roster moves that rate more than he does. With Bosa on the field, his teammates' rate has run from 2.6 per 100 in 2022, the year he had 18.5 sacks and won Defensive Player of the Year, to 7.0 in 2019. Without him for nearly a whole season, it was 5.2 in 2020 and 3.0 in 2025. The five games he has just missed came in at 5.2. San Francisco's 15 sacks in 14 games without him last season were a real collapse, and 2020 says his absence does not have to produce one.
Where the points went#
Now the other column. Expected points is a model's estimate of how much each play moves a team's scoring chances, and it is model output, not a count. In the same comparison, the 49ers gave up about 14.5 fewer expected points per 100 dropbacks with Bosa on the field.
The extra sacks explain about a sixth of that. A sack is worth about 1.9 expected points against the average dropback that is not one, so 1.3 more sacks is about 2.5 points. The rest shows up on plays where nobody sacked anyone. On 2,785 throws, quarterbacks completed about 5 percentage points fewer with him on the field and gained about three-quarters of a yard less per attempt. The Next Gen Stats pressure flag fired on 32 percent of dropbacks with him on and 23 percent with him off, or 7.4 points more after the same adjustment.
Sacks miss this routinely. In 2023, sixteen players had more sacks than Bosa's 10.5. PFF's pass-rush grades had him third of 129 edge rushers that year.
The 49ers' gap with Bosa on the field is nearly three times the median rusher's
Fewer expected points allowed per 100 dropbacks with the rusher on the field than off it, same team and season, same down group. The 31 rushers with three or more seasons of 8+ sacks, 2019-25. Model output. The top seven and the median are shown. Bosa's 90 percent range runs from 3 to 25.
Read the top four first, then the Bosa bar against the bottom one. Parsons, Donald, Garrett and Bosa is roughly the list you would have written without a computer, which is pleasant and proves little. The spread across all 31 rushers is about as wide as the noise on Bosa's own bar. Von Miller is last at +17.4, which means either that his teams improved whenever he sat down or that this measure is noisy. I am going with noisy.
The confound that worries me most is the company he keeps. When Bosa rests, other starters may rest with him, and then a backup defense is being charged to his absence. I checked the one I could. On snaps with Fred Warner on the field, the gap is 15.4. With win probability between 20 and 80 percent, so no garbage time, it is 16.7. Limiting it to the first three quarters, or comparing only within a single game, pushes it higher. Those tests still do not cover the rest of the lineup. Leadership does not appear in the participation file.
So I hold two numbers. Taken at face value, 14.5 per 100 dropbacks is about 4 points a game, on the 28 dropbacks a game he has been on the field for. Pulled toward the median as hard as that noise suggests, it is about 7 per 100, or 2 points a game. The estimate is 2 to 4 points a game, against a defense allowing 19.9. His two games this season, 76 snaps and no sacks, are too few to move it either way.
What to watch, and the call#
Do not grade his return by the sack column in the next two weeks. The Raiders and Cowboys have allowed 3.6 and 3.3 sacks per 100 dropbacks, fourth-lowest and second-lowest in the league. Bosa could change those games and finish them with nothing in the box score. Completion rate and yards per attempt will say more, and they will still say very little in two games.
The projection for the total is this. Ten games at 38 dropbacks is about 383 dropbacks. His teammates' baseline is about 5.2 per 100, between this season's 5.6 and the 5.1 from the career split. His net adds about one more per 100 on a full workload, for about 6.2. That is 24 more sacks, and 39 for the season.
The call: San Francisco finishes the 2026 regular season with fewer than 41 sacks. Check it Jan. 11, the day after the finale in Arizona.
The case against is on the 49ers' own record. They had 28 sacks in the last 10 games of 2019, 32 in 2021 and 33 in 2023, all with Bosa. They need 26. I have added nothing for DeForest Buckner, the interior rusher acquired from Indianapolis last month, and subtracted nothing for further missed games. Those are the two largest errors in that 39, and they point in opposite directions.
Data: nflverse play-by-play and participation files, regular seasons 2019-25, which give the players on the field on each play. Dropbacks include sacks and scrambles; spikes and two-point tries are out. There is no 2026 participation file yet, so 2026 figures are counts over whole games. The sample: rusher-seasons with 8 or more sacks, at least 250 dropbacks on the field and 60 off (n = 228). For Bosa that is 2019 and 2021-24: 2,381 dropbacks on, 773 off, 84 games. The chart uses rushers with three or more such seasons. The adjustment: on-against-off differences within team-season and down group (first and second down; third and fourth with under 5 to go; third and fourth with 5 or more), combined with Mantel-Haenszel weights. Bosa's ranges come from resampling his 84 games 2,000 times. Pressure: the Next Gen Stats flag in the participation file, present on about 88 percent of 49ers dropbacks through 2022 and nearly all since. PFF's grade ranks include the postseason. The shrink: the spread across the 31 rushers is 7.0 and the sampling error on Bosa is 6.8, so most of his distance from the median is treated as noise. Weakness: rest snaps are not random, and only one other starter was controlled for. Sack credit cannot separate displacement from help.
Sources: nflverse play-by-play, participation and snap count files; NFL.com career statistics for Nick Bosa; ESPN team statistics for San Francisco, Las Vegas and Dallas; NFL practice report, Nov. 4; PFF.
