Advanced stats show Cowboys strengths and weaknesses
Nov 7, 2010; Green Bay, WI, USA; A Dallas Cowboys fan wears a bag over their head during the fourth quarter against the Green Bay Packers at Lambeau Field. The Packers defeated the Cowboys 45-7. Mandatory Credit: Jeff Hanisch-USA TODAY Sports | IMAGN IMAGES via Reuters ConnectFire the coaches! Cut the players! Churn the roster! Send a message! - It's hard to avoid hyperbole in the days following a season-opening loss, but we take a dispassionate look at the Giants game anyway. By the numbers, of course.
But the numbers come with their own baggage, because we have no way of knowing which numbers are outliers and which numbers are indicative of the coming regular-season performance. In other words, if this had been a Week 8 game, we'd have the previous seven games worth of data to compare against to determine what is an outlier and what is a pattern and would likely be much calmer about a mid-season loss than staring into a dark abyss to start the season.
-49.4 Passer Rating Differential (PRD): That's the difference in passer rating between the Giants (134.2) and the Cowboys (84.8). We know that passer rating differential (PRD) is one of the stats most closely linked to winning in the NFL. And one of the interesting things about PRD is that it can be plugged into a very simple formula to predict win totals in the NFL. The PRD formula (Projected Wins = PRD * 0.16 + 8.5) has shown a fairly close correlation with the Cowboys' actual wins over the last few years, as the table below illustrates.
body .sbnu-legacy-content-table td,body .sbnu-legacy-content-table th,body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important;}body .sbnu-legacy-content-table td,body .sbnu-legacy-content-table th { padding: 4px 6px !important;}| Dallas Cowboys Passer Rating Differential, 2020-2025 | ||||||
| Year | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 |
| PRD | -12.8 | 27.1 | 1.9 | 22.0 | -17.2 | -12.8 |
| Projected Wins | 6.5 | 12.8 | 8.8 | 12.0 | 5.7 | 6.5 |
| Actual Wins | 6 | 12 | 12 | 12 | 7 | 7 |
| Projection error | 0.5 | 0.8 | 3.2 | 0.0 | 1.3 | 0.5 |
The mean absolute projection error (the average of the absolute deviation between the projected and the actual win total) is quite low at 1.0 for the Cowboys over the last six years using our little formula, a strong indication that the formula is a good projection tool.
So what does this mean for the 2026 Cowboys?
Sadly, the -49.4 PRD produces a projected 0.6 wins for the 2026 Cowboys.
If you're wondering whether that means the Cowboys are on pace for a 1-16 season, congratulations : you've just discovered the problem with extrapolating one game across an entire season.
Of course, extrapolating one game of PRD across an entire season is statistically ridiculous - which is exactly the point.
73% Third-down efficiency: The Cowboys' third-down efficiency on offense is one of the highlights of the game. The Cowboys were 8-of-11 on third down, which should bode well for the rest of the season.
Too bad the defense was 9-of-11 (82%) on third down.
+0.271 EPA/play: Per RBSDM.com, the Cowboys offense ranked fifth overall in EPA/play and eighth in success rate (percent of plays with positive EPA) with 48.3%
Unfortunately, the defense ranked 32nd with an EPA/play of 0.363 and 31st in success rate with 59.1%
The defensive problems weren't limited to third downs. In fact, the second-down numbers may be even more revealing. John Owning summarizes:
This is the Cowboys defense's performance by down in Week 1, per Sumer Sports.
- John Owning (@JohnOwning) September 15, 2026
1st down: 31 plays, 95 yards allowed (3.06 yds/play), and a -0.105 EPA/play for the opposing offense - this was actually a strong down for the defense, giving up only 4 first downs.
2nd down: 26... https://t.co/X3xwA4dEfO
1st down: 31 plays, 95 yards allowed (3.06 yds/play), and a -0.105 EPA/play for the opposing offense - this was actually a strong down for the defense, giving up only 4 first downs.
2nd down: 26 plays, 210 yards allowed (8.08 yds/play), +0.476 EPA/play for the offense, and 13 first downs allowed - a big step down in performance.
3rd down: 11 plays, 89 yards allowed (8.09 yds/play), +0.886 EPA/play for the offense, and 9 first downs allowed out of just 11 plays - meaning they converted a first down on roughly 82% of their third-down snaps against Dallas.
I think the issues on 2nd down were just as big, and maybe even bigger, than their issues on third downs
100% red zone percentage. The Cowboys were 3-of-3 on their red zone drives, up from their 56.9% average last year on a topic that had been declared a special area of attention this year.
Bummer that the Giants were also 100% in the red zone, but on four drives, not three.
1.10 Yards Per Route Run (YPRR) for Cowboys receivers. YPRR is a football metric that measures a pass catcher's efficiency by dividing their total receiving yards by the number of passing routes they ran, essentially giving us the average number of receiving yards a player makes every time he runs a route. We looked at the metric in great detail this offseason, go to this post if you need more info.
Here's the YPRR for the Giants game, and how that compares to last year's numbers.
body .sbnu-legacy-content-table td,body .sbnu-legacy-content-table th,body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important;}body .sbnu-legacy-content-table td,body .sbnu-legacy-content-table th { padding: 0px 6px !important;}| Yards per Route Run, 2025 vs 2026 | ||||||||
| POS | Player | Wk 1 Targets | Wk 1 Receptions | Wk 1 Yards | Wk 1 Routes | Wk 1 YPRR | 2025 YPRR | Change |
| WR | KaVontae Turpin | 1 | 1 | 12 | 5 | 1.60 | 1.65 | -0.05 |
| CeeDee Lamb | 8 | 5 | 44 | 29 | 1.52 | 2.37 | -0.85 | |
| Jonathan Mingo | 1 | 1 | 10 | 7 | 1.43 | 0.61 | 0.82 | |
| George Pickens | 6 | 3 | 28 | 32 | 0.88 | 2.35 | -1.47 | |
| Ryan Flournoy | 5 | 2 | 22 | 26 | 0.85 | 1.74 | -0.89 | |
| TE | Luke Schoonmaker | 1 | 1 | 8 | 5 | 2.40 | 0.96 | 1.44 |
| Brevyn Spann-Ford | 2 | 2 | 14 | 8 | 1.75 | 1.29 | 0.46 | |
| Jake Ferguson | 2 | 2 | 6 | 27 | 0.22 | 1.20 | -0.98 | |
| HB | Javonte Williams | 5 | 5 | 31 | 20 | 1.55 | 0.40 | 1.15 |
In total, YPRR for all receivers dropped from 1.51 last year to 1.10 in Week 1, a significant drop in efficiency, especially for last year's receiving yard leaders Pickens, Lamb, Flournoy, and Ferguson. PFF shows Pickens with two drops, Lamb and Flournoy with one each, and those drops were key to each receivers efficiency drop.
FTNFantasy registered only three drops, but calculated the cumulative EPA lost on those three drops at 8.3 points - more than the Cowboys' eight-point margin of defeat.
How many expected points did each team lose on drops in week 1?
- SFdata9ers (@sfdata9ers) September 16, 2026
1. PIT 11.7 (2 drops)
2. DAL 8.3 (3)
3. DET 6.8 (2)
4. ARI 5.1 (3)
5. IND 4.0 (3)
Drops charted by @FTNFantasypic.twitter.com/XRX9D5l1PH
6.3% Pressure Percentage: The Cowboys front four generated zero sacks, four QB hits, and five hurries on 142 combined pass rush opportunities. That 6.3% pressure percentage ranks them 28th in the league, far off the league average of 10.1%.
- The defensive interior went 0-1-3 for a pressure percentage of 5.1% (26th, 7.8% league average)
- The edge rushers went 0-3-2 for a pressure percentage of 7.8% (26th, 12.4% league average)
Not only is that a bad look this year, it's a significant decline versus last year where the DI led the league with a 10.4% pressure percentage, and the edge rushers and their 11.8 pressure percentage at least ranked them 17th overall.
So far in 2026, the roster bets on the defensive line are not paying off.
Again, one game is nowhere near enough to declare those bets failures. But after one game, the numbers certainly aren't providing evidence that the Cowboys' defensive-line strategy has worked.
-24 Hidden Yardage: Bill Parcells used to say that a 100-yard advantage in field position (the sum of starting field positions on all drives), or hidden yardage," was worth seven points.
body .sbnu-legacy-content-table td,body .sbnu-legacy-content-table th,body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important;}body .sbnu-legacy-content-table td,body .sbnu-legacy-content-table th { padding: 0px 6px !important;}| 2026 Hidden Yards Differential | ||||
| Metric | Pre Wk 1 | Pre Wk 2 | Pre Wk 3 | vs. Giants |
| Hidden Yardage differential | -200 | +101 | -93 | -24 |
| Result | W: 17-7 | W: 34-13 | L: 24-27 | L: 20-28 |
The Giants had a small hidden yardage advantage of 24 yards versus the Cowboys, but hidden in that number is a 45-yard field position gain on the Prescott interception just before the half.
On paper, and going by Parcells' calculation, that INT/45-yard field position gain cost the Cowboys 3.2 points. But the turnover directly turned into seven Giants points, making its real-world impact considerably larger than the raw hidden-yardage differential suggests.
So what do the numbers tell us?Not that the Cowboys are a 0-17 team. Not that the offense is broken. And certainly not that one game is enough to rewrite the entire season. But the numbers do tell us that the Giants exposed a very real problem on defense.
- Offense: generally productive, particularly on third down and in EPA/play.
- Passing game: inefficient despite offensive success overall, with drops and a key INT contributing.
- Defense: catastrophic on efficiency, particularly on second and third down.
- Pass rush: nowhere near the level expected from the roster construction.
- Field position: basically neutral/slightly negative rather than the main explanation.
The Cowboys' offense was productive by EPA and third-down efficiency, while a handful of offensive execution failures turned a potentially winnable game into a loss.
But the defense was near the bottom of the league in almost every efficiency measure that matters, particularly on second and third down. The pass rush, in particular, has plenty of work to do.
The good news is that Week 1 gives us questions, not answers. In seven weeks, we'll have enough data to start separating the outliers from the patterns. For now, the numbers suggest there are some things to fix - and plenty of football left to figure out which ones actually matter.