Does defense vs position matter for NFL running back rushing yards?
In our 2024 and 2025 NFL regular-season count, backs finished 4.26 rushing yards above their own earlier average against the most generous five defenses and 5.47 below it against the stingiest five, a 9.73-yard gap. But the generous side was +6.77 in 2024 and only +1.07 in 2025, below the middle group's +3.65. The stingy side was below average in both years. This is a historical association, not a forecast or a test against prop lines.
Checked on November 14, 2026: in our held 2024 and 2025 NFL regular-season logs, backs finished about 10 rushing yards further above their own earlier average against the five most generous run defenses than against the five stingiest. The difference was uneven by year. The generous group was +6.8 yards in 2024, then only +1.1 in 2025, below that year’s middle group at +3.7. The stingy group stayed below its own earlier average in both years, at -6.2 and -4.7.
That is the finding, not a fixed adjustment for the next prop. We build the Steezanomics Lab, and this study uses its running-back allowance definition with frozen historical data. The 2024 and 2025 source files were frozen on November 12, 2026. The current NFL defense vs position hub shows 2026, so its live ranks are not the historical ranks used below.
The result, including the middle
For each qualifying back-game, we subtracted that back’s average rushing yards in earlier regular-season appearances that year from his rushing yards in the measured game. A result of 68 yards against an earlier mean of 71.5 is about -3.5 yards. The opponent group was set before the game from the defense’s earlier allowance, not its final-season rank.
| Season | Opponent allowance group | Yards vs own prior mean | Back-games | Backs | Games | Above / equal / below |
|---|---|---|---|---|---|---|
| 2024 | Most generous | +6.77 | 89 | 45 | 54 | 45 / 0 / 44 |
| 2024 | Middle | -2.33 | 368 | 67 | 166 | 147 / 0 / 221 |
| 2024 | Stingiest | -6.17 | 92 | 52 | 53 | 33 / 0 / 59 |
| 2025 | Most generous | +1.07 | 70 | 46 | 44 | 30 / 0 / 40 |
| 2025 | Middle | +3.65 | 376 | 59 | 163 | 176 / 0 / 200 |
| 2025 | Stingiest | -4.65 | 79 | 46 | 49 | 31 / 0 / 48 |
| Both years | Most generous | +4.26 | 159 | 66 | 98 | 75 / 0 / 84 |
| Both years | Middle | +0.69 | 744 | 85 | 329 | 323 / 0 / 421 |
| Both years | Stingiest | -5.47 | 171 | 69 | 102 | 64 / 0 / 107 |
The pooled generous-minus-stingy gap, calculated before rounding the group means, is 9.73 yards. The 2024 gap was 12.94; the 2025 gap was 5.72. In the pooled stingy group, 64 back-games finished above the player’s prior mean and 107 below it. Even in the generous group, 75 were above and 84 below. These are not over/under records against sportsbook lines.
The middle group matters here: in 2025 it finished higher than the generous group. The more consistent pattern is that the stingy side was below its own earlier mean in both seasons. The generous side did not hold a uniform boost.
What we knew before each game
We used the same position definition as NFL Defense vs Running Backs: add the named running backs’ and fullbacks’ rushing yards allowed by a defense, then divide by all its earlier team games, including games with zero RB or FB rushing yards. The source is our held nflverse regular-season play-by-play. Player outcomes and prior averages come from the held game logs behind the Lab. We did not substitute the current live table for a historical rank.
Before each date, we ordered all 32 defenses by their earlier RB rushing yards allowed per game, using the stored two-decimal rates. Rank 1 means fewest allowed. The fifth-lowest allowance and fifth-highest allowance set the stingy and generous cutoffs; any defense tied at a cutoff would be included. In this qualifying sample each extreme contained five defenses on a date. We required all 32 defenses to have at least five earlier games before classifying that date. The live table’s display floor of three games is separate from this study rule.
A back or fullback needed five earlier logged appearances in that season and an earlier mean of at least five carries per appearance. We did not require any number of carries, yards or snaps in the measured game. Earlier appearances with little or no rushing work remain in the player’s prior mean. Both the player mean and defense allowance use dates strictly before the measured game; games on the same date enter only afterward. The baseline resets each season. Partial 2026 was checked separately and is not pooled into these results.
The source population is recorded player appearances, with unique season-roster IDs and a frozen name/position mapping. It is not a reconstruction of who was known to be active before every game. Two 2025 log rows with an unresolved Nathan/Nate Carter name did not join to a season ID; neither could meet the five-earlier-appearance rule. We did not guess an identity.
| Sample step | 2024 back-games | 2025 back-games |
|---|---|---|
| Matched RB or FB regular-season appearances | 1,587 | 1,611 |
| Excluded: fewer than five earlier appearances | 633 | 656 |
| Then excluded: prior mean carries under five | 378 | 400 |
| Then excluded: at least one defense below five earlier games | 27 | 30 |
| Qualifying appearances | 549 | 525 |
That leaves 1,074 back-games by 86 backs across 360 NFL games. The two extremes together have 330 back-games by 80 backs across 186 games. 55 of those 80 backs appear on both sides, and 14 games contain an observation from each extreme. These rows are repeated players and sometimes shared games, not 1,074 independent experiments. The backs and games in separate table rows overlap too.
Two games that run against the average
We selected the first qualifying 2025 Saquon Barkley appearance in each extreme, before looking at whether he beat his prior mean. Both rows were visible in Barkley’s public rushing-yards game log when checked on November 14, 2026.
| Opponent group | Game | Defense before game | Barkley’s earlier mean | Measured result | Difference |
|---|---|---|---|---|---|
| Stingy | Nov 16, 2025 vs Detroit | 71.78 RB rushing yards allowed per game; rank 4 of 32 over nine earlier games | 64.33 yards over nine earlier appearances | 83 yards on 26 carries | +18.67 |
| Generous | Dec 28, 2025 at Buffalo | 112.87 allowed; rank 29 of 32 over 15 earlier games | 71.47 yards over 15 earlier appearances | 68 yards on 19 carries | -3.47 |
Barkley beat his earlier mean against Detroit’s stingy ranking and missed it against Buffalo’s generous one. One player’s games cannot validate or refute the group averages. The study uses unique game IDs from the frozen logs.
What changes with more history
The primary result requires five earlier player appearances and five earlier games for every defense. We repeated it with eight and ten earlier appearances and team games, keeping the prior mean carries rule and no measured-game workload filter. These are smaller, later-season samples.
| Rule, both years | Opponent group | Yards vs own prior mean | Back-games | Backs | Games | Above / equal / below |
|---|---|---|---|---|---|---|
| Eight earlier | Most generous | +2.08 | 109 | 57 | 68 | 52 / 0 / 57 |
| Eight earlier | Middle | +1.47 | 530 | 78 | 244 | 238 / 0 / 292 |
| Eight earlier | Stingiest | -4.06 | 115 | 58 | 75 | 47 / 0 / 68 |
| Ten earlier | Most generous | +4.74 | 82 | 51 | 53 | 40 / 0 / 42 |
| Ten earlier | Middle | +2.96 | 391 | 72 | 187 | 185 / 0 / 206 |
| Ten earlier | Stingiest | -3.44 | 79 | 49 | 53 | 33 / 0 / 46 |
| Original: 5+ carries in measured game | Most generous | +12.30 | 162 | 76 | 102 | 98 / 0 / 64 |
| Original: 5+ carries in measured game | Middle | +7.72 | 721 | 111 | 327 | 401 / 0 / 320 |
| Original: 5+ carries in measured game | Stingiest | +0.39 | 158 | 73 | 104 | 78 / 1 / 79 |
The eight-game and ten-game generous-minus-stingy gaps are 6.14 and 8.18 yards, respectively. The generous group is only 0.61 and 1.78 yards above the middle group under those rules; the stingy group stays below the middle. A bigger history requirement does not turn the generous side into a steady boost.
We changed the primary filter after seeing the first results: the original rule required five carries in the measured game, which selected larger current workloads against earlier averages that included small-workload games, so the revised primary uses only earlier workload for eligibility. Under that original carry-selected rule the generous-minus-stingy gap was 11.91 yards, but even the middle group was +7.72 against its own prior mean; those three rows are shown only as a sensitivity, not as the headline result.
How far to take it
The gap moves with the season, history requirement and workload rule. A separate rough resampling with a stricter player-name match put the gap around 3 to 18 yards. That range belongs to the stricter-match version of the data, not a formal interval calculated for the exact 1,074 rows shown here. Repeated backs and shared games make a simple independent-row interpretation unsafe.
This is an association in historical recorded appearances. A defense’s allowance may move with its opponents, injuries, scheme or game script; a back’s workload and offense can change too. Subtracting each back’s prior mean does not hold those conditions constant. It does not establish why the stingy group was lower or what would happen next week.
Nor did we compare results with a sportsbook’s rushing-yards line or price. Beating an earlier average is different from clearing a posted line. These counts give no betting return or next-game probability. Our NFL rushing-yards research guide shows how to read a back’s carries and yardage beside matchup context. The separate NFL funnel study asks whether early defense ranks held up later in 2026, a different question from this player-by-player 2024/2025 comparison.
For this week’s context, open NFL Defense vs Running Backs, read the current season and game count, then check the player’s workload in the Lab. The Lab and our tools are free and stay free. The historical gap here remains a reason to investigate, not a number to add to a line.
The allowance calculation follows the Lab’s held play-by-play definition, including named rushes on two-point attempts. Removing those attempts in a separate check changed no 2024/2025 primary bucket memberships and left the pooled gap the same.
Quick answers
How large was the NFL running back rushing yards gap?
Across the qualifying 2024 and 2025 regular-season games, the mean result minus each back's earlier-season average was +4.26 yards against the most generous five defenses, +0.69 against the middle and -5.47 against the stingiest five. The generous-minus-stingy gap was 9.73 yards, but it changed by year and does not predict the next game.
Did generous defenses lift backs in both seasons?
Not consistently. The generous group averaged +6.77 yards relative to each back's earlier mean in 2024, then +1.07 in 2025 while the 2025 middle group averaged +3.65. The stingy group averaged -6.17 and -4.65 in those years.
Were the defense ranks known before the measured game?
Yes. We used only strictly earlier regular-season dates to calculate each defense's RB and fullback rushing yards allowed per team game and each back's own average. All 32 defenses needed five earlier games; each included back needed five earlier logged appearances and a prior mean of at least five carries. We did not filter on his carries in the measured game.
Does the 9.73-yard gap show that an NFL rushing yards prop would have won?
No. The comparison is with each back's prior average, not a sportsbook's line or price. The same backs recur, games can contain multiple backs, and role, injuries, offense and opponent mix were not held constant. We did not test betting returns or next-game probabilities.
Is the live NFL defense table the historical study ranking?
No. The linked Lab table shows the current season. We recomputed the same RB and fullback rushing-yard allowance definition before each 2024 and 2025 game, using only earlier dates. Rank 1 means the fewest yards allowed.
Want tonight's board?
The Lab has the full slate: every line with L5, L10, L20 and season hit rates, game logs, matchup context and the best price at your books. A short daily card of picks is posted before the games and graded in public afterwards, win or lose. The Lab is free, and it stays free.