NFL Week 3 Advanced Stats: Every Game, Coaching Flags and the Full 32-Team Stack Ranking
All 15 games from Sunday and Monday night run back through the play-by-play: who won the tape, who won the scoreboard, and where every team stands after three weeks.

Every table sorts: click a column header, click again to reverse it, and once more to go back to the original order.
Fifteen games: all 14 from Sunday, Sept. 27, plus Monday night's Eagles at Bears. Every number below is from the 2026 regular season.
Baltimore at Dallas was played at a neutral site in São Paulo, Brazil, the league's third-ever regular-season game there.
How to read this report
EPA (Expected Points Added): how many points a play was worth compared with an average play in the same situation. Positive is good for the offense.
Net EPA: a team's own offensive EPA minus the EPA its defense allowed in the same game. This is the core "who actually played better" number in this report.
Garbage time removed: plays that happen after a game is effectively decided are left out of the EPA and success-rate numbers, because they distort them. Yards, plays, third downs and turnovers are not filtered. Those are whole-game totals.
Success rate: the share of plays that gained enough yardage for the down to stay "on schedule" (roughly 40% of the needed yards on first down, 60% on second, all of it on third or fourth).
CPOE (Completion Percentage Over Expectation): how much more (or less) accurate a quarterback was than expected, given how hard his throws were. Two independently built models measure this: the play-by-play model used throughout, and a player-tracking model used in the cross-check section. They occasionally disagree, and this report notes it wherever they do.
PROE (Pass Rate Over Expected): how much more or less a team passed than the game situation called for. Positive means more pass-heavy than expected, negative means more run-heavy.
Dropbacks: pass attempts, sacks and scrambles combined, the full sample of a quarterback's snaps as a passer.
1. The weekend in one table: scoreboard vs. play-by-play
Gap is the winner's actual margin of victory minus the winner's net EPA in that game. A positive gap means the scoreboard flattered the winner, and the losing team was closer than the final score says. A negative gap means the play-by-play says the winner was even better than the final margin shows.
Four teams won this weekend while losing the efficiency battle outright. Washington, Indianapolis, Denver and Pittsburgh all posted negative net EPA in wins. Their own play-by-play says they were the worse team on the field.
| Carolina at Cleveland | 18-21 | Browns | 3 | +12.1 | -9.1 | Cleveland dominated the tape and still only won by 3 |
| Arizona at San Francisco | 30-36 | 49ers | 6 | +9.2 | -3.2 | Deserved a bigger margin |
| Las Vegas at New Orleans | 35-27 | Raiders | 8 | +10.3 | -2.3 | New Orleans' 4 turnovers kept this closer than the play-by-play |
| NY Jets at Detroit | 24-31 | Lions | 7 | +7.8 | -0.8 | Roughly matched |
| Kansas City at Miami | 24-10 | Chiefs | 14 | +12.3 | +1.7 | Roughly matched |
| 16-24 | Bills | 8 | +5.9 | +2.1 | Roughly matched; Buffalo survived 5 of its own turnovers | |
| Baltimore at Dallas (Brazil) | 34-31 | Ravens | 3 | +0.3 | +2.7 | Essentially a coin flip on the tape too |
| Tennessee at NY Giants | 7-12 | Giants | 5 | +1.7 | +3.3 | Roughly matched |
| Seattle at Washington | 31-33 | Commanders | 2 | -2.1 | +4.1 | Washington won a game it lost on the play-by-play |
| Houston at Indianapolis | 17-19 | Colts | 2 | -3.4 | +5.4 | Indianapolis won a game it lost on the play-by-play |
| 26-30 | Broncos | 4 | -1.6 | +5.6 | Denver won a game it lost on the play-by-play | |
| Philadelphia at Chicago | 7-27 | Bears | 20 | +14.4 | +5.6 | Dominant, and the scoreboard basically agrees |
| Minnesota at Tampa Bay | 23-16 | Vikings | 7 | +1.3 | +5.7 | Minnesota, again, won by more than its offense earned |
| Cincinnati at Pittsburgh | 27-30 | Steelers | 3 | -4.5 | +7.5 | Pittsburgh won a game it lost on the play-by-play |
| New England at Jacksonville | 6-35 | Jaguars | 29 | +17.3 | +11.7 | Dominant either way, but 3 New England turnovers padded an already-lopsided score |
- Cleveland beat Carolina
- San Francisco beat Arizona
- Las Vegas beat New Orleans
Detroit beat NY Jets- Kansas City beat Miami
Buffalo beat LA Chargers- Baltimore beat Dallas
- NY Giants beat Tennessee
- Washington beat Seattle
- Indianapolis beat Houston
- Chicago beat Philadelphia
Denver beat LA Rams- Minnesota beat Tampa Bay
- Pittsburgh beat Cincinnati
- Jacksonville beat New England
Gap = winner's margin minus winner's net EPA. Bars to the left: won by less than the tape says. Bars to the right: the scoreboard flattered the winner.
2. Game by game, all 15 games
EPA and success rate leave out garbage time. Plays, yards, third downs and turnovers are whole-game totals.
Carolina at Cleveland: 18-21
| +2.7 | -9.4 | +12.1 | 37.9% | 62 | 263 | 5-15 | 1 | |
| -9.4 | +2.7 | -12.1 | 35.1% | 74 | 360 | 4-15 | 1 |
Carolina out-gained Cleveland by nearly 100 yards and still lost, and still posted the worse net EPA. Neither offense was sharp. This was Cleveland's defense doing the damage.
NY Jets at Detroit: 24-31
| +11.8 | +4.0 | +7.8 | 59.7% | 62 | 383 | 5-10 | 0 | |
| +4.0 | +11.8 | -7.8 | 45.9% | 63 | 337 | 6-12 | 1 |
Detroit's 59.7% success rate is the best offensive efficiency mark of the weekend. Geno Smith still graded positively for the Jets in the loss.
Houston at Indianapolis: 17-19
| -8.1 | -4.7 | -3.4 | 40.8% | 71 | 300 | 7-15 | 3 | |
| -4.7 | -8.1 | +3.4 | 41.2% | 51 | 223 | 4-11 | 0 |
Houston won the per-play battle on 20 fewer plays, forced three Indianapolis turnovers, gave the ball away zero times itself, and still lost. Indianapolis is the rare team that wins after giving the ball away three times in the same game.
New England at Jacksonville: 6-35
| +9.7 | -7.6 | +17.3 | 52.2% | 59 | 315 | 8-12 | 1 | |
| -7.6 | +9.7 | -17.3 | 35.9% | 60 | 317 | 4-12 | 3 |
New England's three turnovers turned a competitive yardage game (317 to 315) into a 29-point blowout on the scoreboard. Jacksonville's +17.3 net EPA is the best number of the entire weekend, but the 29-point margin still overstates it by almost 12 points once garbage time is stripped out.
Kansas City at Miami: 24-10
| -4.2 | +8.1 | -12.3 | 29.5% | 67 | 329 | 8-16 | 2 | |
| +8.1 | -4.2 | +12.3 | 54.3% | 49 | 334 | 3-7 | 1 |
Kansas City managed 334 total yards on just 49 plays. Miami needed 67 plays to manage only 329. De'Von Achane left in the first quarter with a non-contact knee injury and didn't return.
Tennessee at NY Giants: 7-12
| -6.6 | -8.4 | +1.7 | 38.6% | 57 | 239 | 4-13 | 0 | |
| -8.4 | -6.6 | -1.7 | 40.4% | 57 | 262 | 5-13 | 2 |
The lowest-scoring game of the weekend at 19 combined points. Nobody topped 41% success on either side.
Cincinnati at Pittsburgh: 27-30
| +11.2 | +15.7 | -4.5 | 48.2% | 56 | 413 | 7-12 | 2 | |
| +15.7 | +11.2 | +4.5 | 55.4% | 56 | 352 | 6-12 | 1 |
Cincinnati out-graded Pittsburgh on offense, on both success rate and EPA, in Cincinnati's best offensive showing of the season, and lost anyway. Aaron Rodgers' own game was genuinely good. The gap is Pittsburgh's defense, which gave up its opponent's best offensive number of the year while Rodgers still outscored it.
Seattle at Washington: 31-33
| -5.0 | -3.0 | -2.1 | 31.7% | 63 | 260 | 3-13 | 0 | |
| -3.0 | -5.0 | +2.1 | 50.0% | 64 | 437 | 3-9 | 3 |
Seattle out-gained Washington 437 to 260, the second-most yards by a losing team this weekend behind the Rams' 482, and posted a 50% success rate. Washington won its first game of the season without Jayden Daniels (elbow, out), on three Seattle turnovers.
Arizona at San Francisco: 30-36
| +20.9 | +11.6 | +9.2 | 57.4% | 49 | 438 | 5-8 | 0 | |
| +11.6 | +20.9 | -9.2 | 51.9% | 80 | 365 | 9-18 | 1 |
San Francisco's +20.9 offensive EPA is the best single-team offensive number of the weekend, on 31 fewer plays than Arizona ran.
Minnesota at Tampa Bay: 23-16
| -15.4 | -14.1 | -1.3 | 26.7% | 64 | 243 | 2-17 | 2 | |
| -14.1 | -15.4 | +1.3 | 21.6% | 55 | 230 | 2-15 | 1 |
The single worst offensive game of the weekend on either side of the ball. Minnesota's offense stayed "on schedule" on barely one in five plays, the worst offensive success rate by any team this week. The Vikings are 3-0 on defense and special teams alone. This is the third straight week Minnesota has won with a negative-or-flat offensive EPA number.
Baltimore at Dallas (Brazil): 34-31
| +7.5 | +7.8 | -0.3 | 47.2% | 72 | 417 | 5-13 | 1 | |
| +7.8 | +7.5 | +0.3 | 50.0% | 58 | 378 | 3-9 | 0 |
Essentially a coin-flip game by every measure. Lamar Jackson's individual number, +11.93 EPA on 20 dropbacks, was the better of the two quarterbacks' in this one.
Las Vegas at New Orleans: 35-27
| -1.2 | +9.1 | -10.3 | 55.1% | 72 | 381 | 9-15 | 4 | |
| +9.1 | -1.2 | +10.3 | 42.4% | 63 | 342 | 5-12 | 1 |
New Orleans out-graded Las Vegas on success rate and out-gained them, and turned the ball over four times. That's the whole story of this one.
LA Rams at Denver: 26-30
| -0.4 | +1.1 | -1.6 | 38.7% | 62 | 261 | 7-13 | 1 | |
| +1.1 | -0.4 | +1.6 | 41.2% | 80 | 482 | 8-17 | 2 |
The Rams put up 482 yards on 80 plays, the most of any team this week, without Puka Nacua, who was inactive (hip). Los Angeles graded better on both success rate and net EPA and still lost, its second such result in three weeks.
LA Chargers at Buffalo: 16-24
| -1.2 | -7.1 | +5.9 | 49.2% | 62 | 351 | 6-12 | 5 | |
| -7.1 | -1.2 | -5.9 | 45.8% | 64 | 348 | 3-15 | 1 |
Six combined turnovers in this one, and five of them were Buffalo's own: two interceptions and three lost fumbles. Buffalo still won, because the Chargers couldn't turn five takeaways into enough points, converting just 3 of 15 third downs. Buffalo moves to 3-0. The Chargers are 0-3 and look like the most disappointing team in the league through three weeks.
Philadelphia at Chicago (Monday night): 7-27
| +3.2 | -11.2 | +14.4 | 46.8% | 67 | 377 | 5-13 | 0 | |
| -11.2 | +3.2 | -14.4 | 41.5% | 47 | 248 | 2-9 | 3 |
Case Keenum, starting for the injured Caleb Williams, threw zero interceptions and graded positively, while Chicago's defense turned Jalen Hurts's night into one of the worst offensive performances by any team this week. Philadelphia was missing Dallas Goedert, Marquise Brown and a starting tackle. Even accounting for that, the tape says Chicago was exactly as dominant as the 20-point final margin suggests.
3. Looked better than they were
Washington, Indianapolis and Denver all won this weekend with negative net EPA, the clearest "looked better than they were" signal there is. All three wins count. None of the three teams' own tape supports "we were the better team this weekend."
Pittsburgh won 30-27 against Cincinnati's best offensive day of the season. Aaron Rodgers' individual game, not the defense, is the real reason it won.
Jacksonville's 35-6 rout of New England was dominant, but the scoreboard overstates it by about 12 points once garbage time is removed. New England's three turnovers did a lot of the heavy lifting.
Minnesota, for the third straight week, won on defense and special teams while its own offense graded out poorly. The Vikings are 3-0 without a single positive-EPA offensive week.
Quarterbacks who won despite a negative grade
| Kyler Murray | W 23-16 | -12.90 (31) | -5.20 | |
| Daniel Jones | W 19-17 | -6.87 (40) | -0.76 | |
| Jameis Winston | W 12-7 | -6.11 (25) | -4.80 | |
| Josh Allen | W 24-16 | -5.38 (30) | -0.88 | |
| Bryce Young (for contrast; lost) | L 18-21 | -5.84 (50) | -2.69 |
Josh Allen's place on this list stands out on its own: Buffalo is 3-0, and its quarterback posted a negative grade in a win while throwing two interceptions himself.
4. Looked worse than they were
Houston lost by 2 despite forcing three Indianapolis turnovers and committing zero of its own, on 20 fewer plays. That's as clean a "won the turnover battle, lost the game" result as exists in Week 3.
Cincinnati lost by 3 with the better offensive grade on the field. Joe Burrow's +14.65 EPA game is one of the best of the week, in a loss.
Seattle and the LA Rams both out-gained and out-graded the team that beat them and still lost.
Cleveland won by only 3 despite the best net EPA of any winner other than Jacksonville, Chicago and Kansas City. No winner this weekend was undersold by the scoreboard more.
San Francisco and Las Vegas both deserved bigger margins than they got.
Quarterbacks who lost or barely won despite a strong grade
| Joe Burrow | L 27-30 | +14.65 (38) | +8.89 | |
| Jared Goff | W 31-24 | +12.14 (34) | +6.94 | |
| Lamar Jackson | W 34-31 | +11.93 (20) | +9.19 | |
| Kirk Cousins | W 35-27 | +11.15 (37) | +8.56 | |
| Geno Smith | L 24-31 | +8.95 (42) | +16.00 | |
| Jacoby Brissett | L 30-36 | +9.26 (50) | +6.79 |
Geno Smith posted the second-best play-by-play CPOE of any quarterback in Week 3, behind only Patrick Mahomes, in a losing effort, and the hand-charted data puts none of his five sacks on him. Jacoby Brissett had 50 non-garbage dropbacks, a heavy workload, and still graded positively in a loss.
5. Best and worst quarterback games of the weekend
Best five
| 1 | Brock Purdy | +18.28 (26) | -7.41 | |
| 2 | Joe Burrow | +14.65 (38) | +8.89 | |
| 3 | Patrick Mahomes | +12.92 (24) | +18.68 | |
| 4 | Jared Goff | +12.14 (34) | +6.94 | |
| 5 | Lamar Jackson | +11.93 (20) | +9.19 |
Worst five
| 1 | Kyler Murray | -12.90 (31) | -5.20 | |
| 2 | Jalen Hurts | -10.93 (23) | +2.74 | |
| 3 | Baker Mayfield | -10.88 (40) | -14.82 | |
| 4 | Drake Maye | -9.25 (19) | -4.06 | |
| 5 | Justin Herbert | -8.88 (31) | -3.45 |
Purdy's game carries an asterisk: the best EPA of the week, but a negative CPOE, meaning yards after the catch and explosive plays did more of the work than accurate throwing. The tracking-based accuracy model has Purdy near neutral (+0.26), disagreeing with the play-by-play model's -7.41.
Mahomes is the more trustworthy huge game of the two: elite on both EPA and CPOE, with no gap between the two measures. His +18.68 CPOE is the highest in the league this week by the play-by-play model. The tracking model has him second (+10.20), behind Geno Smith's +21.11, so the two independent accuracy models rank Mahomes and Smith in opposite order.
6. Tracking-data cross-check: quarterback accuracy
A second, independently built look at the passers, from player-tracking data: how long each quarterback held the ball, how often he threw into tight coverage (aggressiveness), and the tracking model's own CPOE.
Top five by tracking CPOE
| Geno Smith | 2.89s | 27.0% | +21.1 | |
| Patrick Mahomes | 2.61s | 8.3% | +10.2 | |
| C.J. Stroud | 3.22s | 25.9% | +8.3 | |
| Trevor Lawrence | 2.71s | 13.8% | +7.7 | |
| Jared Goff | 2.89s | 18.8% | +7.6 |
Bottom five by tracking CPOE
| Baker Mayfield | 3.08s | 8.8% | -16.1 | |
| Deshaun Watson | 3.17s | 10.0% | -15.9 | |
| Bo Nix | 3.03s | 20.6% | -10.0 | |
| Jordan Love | 2.54s | 17.0% | -8.4 | |
| Jameis Winston | 2.55s | 13.6% | -6.2 |
Geno Smith and Jared Goff sit near the top on both the tracking model and the play-by-play numbers, both in losing efforts. That's genuine agreement between two separately built systems. C.J. Stroud's place is worth flagging: the play-by-play model has Houston's offense grading roughly break-even, while the tracking model specifically likes his accuracy. Case Keenum (+1.07) and Jalen Hurts (-4.58) both sit in the unremarkable middle of the tracking list. Monday night's blowout was decided by more than accuracy.
7. Charted-data cross-check: sacks, drops and pass rush
From hand-charted data covering all 16 Week 3 games, Thursday night included.
Sack blame: of 62 charted sacks league-wide this week, only 9 were charted as the quarterback's own fault. Tampa Bay took 6 sacks with zero charged to Baker Mayfield, and the Jets took 5 with zero charged to Geno Smith. Rough weeks for both offensive lines. Las Vegas is the outlier: 2 of Kirk Cousins's 3 sacks were charted as his own fault, the highest quarterback-fault share of any team with multiple sacks.
Drop rate (share of catchable targets dropped): Pittsburgh 12.5%, NY Giants 11.8%, Tampa Bay 11.5%, Green Bay 11.4%, Washington 11.1%. Each lost about one catchable ball in nine, or more, to drops. Zero drops were charted for Atlanta, Kansas City, New England, Las Vegas, San Francisco and Tennessee.
Blitz rate (share of pass plays with an extra rusher): Minnesota 36.4%, the highest in the league, then Washington 31.4% and Denver 23.8%. Chicago and Philadelphia both blitzed 0% of the time Monday night. This shows tendency, not effectiveness, on a three-game sample.
8. Coaching and play-calling
Team level only.
Pass rate over expected, this weekend
| Chicago | +13.4 | NY Giants | -19.4 |
| Seattle | +10.4 | Baltimore | -14.2 |
| Kansas City | +7.2 | Tennessee | -12.1 |
| Pittsburgh | +6.2 | Buffalo | -11.9 |
| +5.9 | Miami | -7.8 |
Chicago leaned pass-heavy with a backup quarterback making his first start of the season, and it worked (+14.4 net EPA, a 20-point win). Worth watching as a philosophy note.
Pass rate over expected, season to date (Weeks 1-3)
| Pittsburgh | +5.4 | NY Giants | -11.4 |
| San Francisco | +4.9 | Atlanta | -10.9 |
| Cincinnati | +4.8 | Tennessee | -10.5 |
| Kansas City | +4.3 | Miami | -8.5 |
| Carolina | +4.2 | Philadelphia | -7.3 |
Fourth-down aggressiveness, this weekend: Carolina, Chicago, Cleveland and Miami each went for it three times, the most of any team. Carolina and Miami combined to convert zero of six attempts. Chicago and Cleveland both went 2-for-3. Six teams (Denver, Jacksonville, Las Vegas, the Giants, Pittsburgh and San Francisco) didn't attempt a fourth down at all.
9. Schedule and injury context
International site: Baltimore at Dallas was played in São Paulo, Brazil, the league's third-ever regular-season game there. Both teams face a short recovery window into Week 4.
No short-week teams this weekend, and no byes yet this season.
In-game injuries
- De'Von Achane (Miami, knee): left in the first quarter after four touches with a non-contact injury. The team described itself as "not very optimistic" afterward.
- Puka Nacua (LA Rams, hip): downgraded to out before the game and did not play. The Rams still posted the week's high of 482 yards without him.
- Breece Hall (NY Jets, thigh): exited the Detroit game. Severity not yet reported.
- Jayden Daniels (Washington, elbow): officially out. Marcus Mariota started Washington's first win of the season.
- Caleb Williams (Chicago, hamstring): officially out for Monday night. Case Keenum started and won.
Notable pregame injuries
Dallas Goedert, Marquise Brown and starting tackle Fred Johnson were all out for Philadelphia's Monday-night loss. Zay Flowers (Baltimore, hamstring, questionable) played in the Brazil game. Jaylen Warren and Rico Dowdle (Pittsburgh) were both questionable or out ahead of the win over Cincinnati. Nico Collins (Houston, hamstring) was out for the loss to Indianapolis.
10. Full 32-team stack ranking through Week 3
Every team now has three full games. The adjusted rating corrects each team's raw rating for how tough its three opponents have been.
| 1 | +12.0 | +13.7 | +0.404 | +0.073 | 60.5% | 49.7% | |
| 2 | +10.7 | +11.9 | +0.151 | -0.151 | 50.4% | 40.6% | |
| 3 | +8.2 | +8.4 | +0.082 | -0.082 | 43.5% | 40.1% | |
| 4 | +7.9 | +9.1 | +0.022 | -0.139 | 45.1% | 35.5% | |
| 5 | +7.4 | +10.4 | +0.214 | +0.025 | 49.4% | 46.8% | |
| 6 | +7.0 | +12.2 | +0.188 | -0.026 | 50.6% | 38.5% | |
| 7 | +6.1 | +9.7 | +0.214 | +0.022 | 53.6% | 44.5% | |
| 8 | +6.0 | +10.9 | +0.102 | -0.098 | 43.6% | 45.5% | |
| 9 | +5.8 | +9.2 | +0.118 | -0.058 | 46.1% | 43.8% | |
| 10 | +5.5 | +4.4 | +0.056 | -0.030 | 44.7% | 42.8% | |
| 11 | +3.1 | +4.2 | -0.131 | -0.179 | 30.6% | 31.9% | |
| 12 | +2.7 | +5.5 | +0.059 | -0.042 | 46.1% | 46.2% | |
| 13 | +1.7 | -3.0 | -0.027 | +0.027 | 41.8% | 41.5% | |
| 14 | +0.5 | -1.4 | +0.072 | +0.121 | 47.0% | 47.6% | |
| 15 | +0.3 | -1.6 | +0.092 | +0.104 | 51.1% | 47.0% | |
| 16 | +0.1 | -0.2 | +0.051 | +0.061 | 40.7% | 44.5% | |
| 17 | -0.6 | -2.7 | -0.095 | -0.040 | 42.5% | 42.0% | |
| 18 | -1.1 | -1.8 | -0.090 | -0.035 | 38.7% | 41.0% | |
| 19 | -1.2 | -4.3 | +0.002 | +0.072 | 46.5% | 49.7% | |
| 20 | -2.1 | -2.5 | +0.155 | +0.188 | 49.7% | 52.4% | |
| 21 | -3.0 | -3.2 | -0.001 | +0.061 | 39.3% | 46.8% | |
| 22 | -4.4 | -5.2 | -0.054 | +0.036 | 42.1% | 38.8% | |
| 23 | -4.8 | -4.0 | -0.015 | +0.070 | 43.3% | 45.2% | |
| 24 | -5.3 | -7.3 | -0.204 | -0.086 | 36.6% | 37.4% | |
| 25 | -5.6 | -5.3 | -0.029 | +0.067 | 39.0% | 44.0% | |
| 26 | -7.0 | -7.9 | -0.160 | +0.002 | 36.3% | 44.9% | |
| 27 | -7.1 | -4.4 | -0.150 | -0.055 | 47.5% | 38.0% | |
| 28 | -7.5 | -13.6 | -0.038 | +0.290 | 37.2% | 55.3% | |
| 29 | -7.7 | -8.1 | -0.155 | -0.009 | 38.2% | 47.7% | |
| 30 | -7.8 | -7.1 | -0.081 | +0.058 | 41.7% | 44.6% | |
| 31 | -8.1 | -11.5 | -0.159 | +0.049 | 42.0% | 46.3% | |
| 32 | -11.7 | -14.4 | -0.068 | +0.188 | 42.7% | 48.5% |
San Francisco is #1 on both the raw and adjusted number, the best team in the league right now by any cut of this data. Chicago's Monday-night win vaults the Bears to third on a small but real sample. Kansas City and Las Vegas both slide several spots once schedule strength is factored in. Both have padded raw numbers against one of the softer three-game slates in the league. Denver moves the other direction: a mediocre raw number, but the second-toughest schedule in the league so far, behind only Miami's. Treat the order as directional. The middle of the table, from Minnesota down to Carolina, is bunched within a field goal.
- San Francisco
- Jacksonville
- Chicago
- Seattle
- Buffalo
- Kansas City
- Baltimore
- Las Vegas
- Cincinnati
LA Rams
NY Giants- Tampa Bay
- Philadelphia
- Green Bay
- Atlanta
- Miami
- Pittsburgh
- Tennessee
LA Chargers- Indianapolis
Adjusted rating: points per game better or worse than average, corrected for opponent strength.
11. What these numbers can't tell you
Three games is a small sample. Every team has three games, and the opponent adjustment runs on that same three-game sample. Early instability is expected and will settle as the season goes on.
Defensive EPA is opponent-adjusted in the stack ranking only, nowhere else in this report.
Turnovers and special teams are box-score context, not valued in EPA, and several results this week leaned heavily on turnovers that don't reliably repeat week to week.
Play-calling is shown at team level only, with no coordinator-level attribution.
This is not a betting input. This report is for understanding team and player value, not for building a betting card.