Bears 27, Eagles 7 (Monday Night Football)

Broncos 30, Rams 26 (Sunday Night Football)

Falcons 35, Packers 14 (Thursday Night Football)

Ravens 34, Cowboys 31 (Brazil)

Commanders 33, Seahawks 31

Colts 19, Texans 17

Bills 24, Chargers 16

Steelers 30, Bengals 27

Lions 31, Jets 24

Browns 21, Panthers 18

Raiders 35, Saints 27

Jaguars 35, Patriots 6

Chiefs 24, Dolphins 10

49ers 36, Cardinals 30

Giants 12, Titans 7

Vikings 23, Buccaneers 16

Steelers at Browns (Week 4 Thursday Night Football Preview)

Yeah, did not have Case Keenum going off on Monday Night Football on my Week 3 bingo card. While we give the 38-year-old his flowers for completing 24 of 34 passes for 247 yards and two touchdowns (with no interceptions or sacks), maybe we should also spare a few for Ben Johnson.

After two weeks of struggling on early downs and living on late-down production, the Eagles offense flipped the script on Monday; Philadelphia was a paltry 2-for-11 on third/fourth down.

Welcome to the 2026 season, Luther Burden. The second-year player caught seven of his team-high 11 targets for 61 yards and a touchdown. However, volume remains a concern; per Pro Football Focus’ Nathan Jahnke, he played 44 of Chicago’s 72 snaps, trailing fellow receivers Kalif Raymond and Rome Odunze (who played 61 and 54 snaps, respectively).

Apparently Chicago favored Odunze and Raymond over Burden in two tight end sets on Monday, though Burden did run one more route overall than Odunze (27 to 26, while Raymond ran 31 routes on 35 Chicago dropbacks).

D’Andre Swift ran for 84 yards and had two catches but failed to score on all four of his goal-to-go carries.

Saquon Barkley had an efficient day on the ground, rushing for 82 yards on 15 carries (5.5 yards per rush) with a 53% success rate. He had 100% of the Eagles’ running back carries, but failed to score despite getting a pair of goal-to-go carries.

The Rams led 16-0 at halftime before the game completely turned on its head.

The Rams only lost the turnover battle, 2-1, but the Broncos’ net turnover EPA margin is lopsided due to how impactful Talanoa Hufanga’s fourth-quarter pick six was; that play alone was worth 9.3 expected points.

Los Angeles was 1-for-4 in the red zone, while Denver was 3-for-3.

Denver was 7-for-13 on late downs, while the Rams were 9-for-18. The Hufanga interception return is why the late-down margin is so far in Denver’s favor.¹

While Matthew Stafford threw for 390 yards and two touchdowns, he gave most of this value back via his two interceptions and four sacks.

Bo Nix was relatively inefficient but had a roughly average day in terms of EPA per dropback thanks in large part to sack avoidance (he was brought down once on 38 pass attempts).

His inefficient day came in spite of over 82% of his passes coming within nine yards of the line of scrimmage; the league average in 2025 was 67.4%.

Courtland Sutton and Jaylen Waddle were both targeted seven times, but managed just three and two catches, respectively (though the chart does not include Waddle’s two-point conversion reception, even that would only bring him up to 6.4 point-per-reception fantasy points). Two of Sutton’s targets came in the end zone. Only two wide receivers have underperformed their expected fantasy point production² by more than Sutton this season.

That being said, Sutton also isn’t even the biggest expected fantasy point underachiever on his own team; that would be J.K. Dobbins, who had 49 yards on 17 scoreless rushes against the Rams, including two in goal-to-go situations. For 2026 as a whole, the only running back, receiver or tight end to undershoot their fantasy expectation by more is Seattle’s Emanuel Wilson.

Davante Adams’ role in Puka Nacua’s absence continues to be outrageous, as he saw 13 targets with a 17.5 aDOT. He was held out of the end zone (despite receiving a target there), but still managed seven catches for 137 yards.

What a mess for the Packers, who would be 0-3 if not for a fourth-quarter comeback against the Jets in Week 2.

Atlanta put up 498 yards of total offense (7.5 yards per play) in Penix’s return, compared to 328 (5.2 yards per play) for Green Bay.

The Falcons failed to reach the red zone in both of their first two games this season but scored touchdowns in four of their five trips inside the 20 on Thursday.

The raw numbers look good for Jordan Love, who threw for 312 yards, but he needed 53 throws to get there.

Love was under constant pressure as the Packers offensive line looked overmatched.

I predicted that Love would go wild with deep throws in this one, given the air yards Atlanta allowed in 2025 and in its first two games of 2026. However, this turned out not to be the case, as the Packers passer ended up with a lower aDOT than both Aaron Rodgers and Bryce Young in their games against the Falcons. Perhaps the heavy pressure Love faced played a role here.

Love had a very ordinary aDOT, as his 7.4 mark was 17th among the week’s starting quarterbacks. 39 of his passes (74%) traveled fewer than 10 yards beyond the line of scrimmage.

When he did go deep, the results were poor, as Love was 1-for-7 on throws of 20 yards or more.

The Falcons have bumped last week’s effort from the Ravens for the top success rate in a single game in 2026 (63.8%, compared to Baltimore’s 60.3% mark against New Orleans).

Atlanta’s success rate also would have been the second-highest single-game mark last year;³ only the Rams’ rate of 64.2% in a Week 10 game against the 49ers was better.

Penix completed 18 of 25 passes for 256 yards (10.2 yards per attempt) with one touchdown, one interception and no sacks.

Atlanta was nearly as dominant on the ground, as its running backs combined for 244 yards on 39 carries. Bijan Robinson had 194 yards and a pair of scores on 29 rushes, while Brian Robinson had 50 and a score on the other 10.

I did predict a nice night for Christian Watson, and he came through here with seven catches for 96 yards and a touchdown on 10 targets.

Surprisingly to me, Watson managed a modest (by his standards) 9.1-yard aDOT; since the start of the 2025 season, he has averaged 15.1 air yards per target.

The deep role was Matthew Golden’s on Thursday, as his aDOT came in at 16.6; he caught five of 12 targets for 100 yards and a touchdown.

The Drake London splits with and without Penix made the rounds leading into the game and for good reason. In 15 games together, including Thursday, London has 100 catches on 158 targets for 1,480 yards and nine touchdowns, per StatMuse.

The much-ballyhooed poor field conditions did not stop this game from treating Brazilian fans to a back-and-forth shootout that was not decided until the last second.

I called Baltimore’s dramatic early downs/late downs split a sample size fluke in the preview for this game, and while I still think the Ravens will improve on third/fourth down, that sample got a bit bigger on Sunday.

The Ravens now lead the league in success rate on early downs (54.8%) and are second from the bottom on late downs (31.4%).

Derrick Henry ran for 89 yards and two touchdowns, but did so on 26 carries (3.4 yards per rush), while managing a mere 34.6% success rate. Who would have thought that Dallas would be the first team to slow him down this season?

The Ravens actually had six successful plays on their 10 non-Henry designed runs, split between Lamar Jackson, Justice Hill and Zay Flowers.

Ryan Flournoy caught two of his eight targets for 22 yards; two of these targets were in the end zone.

Per Jahnke, Zay Flowers only appeared for 21 of Baltimore’s 64 offensive plays, while running a route on 14 of 22 dropbacks. He made the most of them though, catching five of six targets for 84 yards.

Flowers’ per-route efficiency is off the charts so far, as he had 150 yards in less than two full quarters of action in Week 1 before leaving that game with an injury. He also missed Week 2, yet is still Baltimore’s leading receiver.

The Commanders pulled off the week’s biggest upset, but it was aided by three Seahawks turnovers worth nearly 19 expected points. Seattle gained 458 yards (6.8 per play) compared to 258 (4.0) for Washington, while the Seahawks also won the success rate battle, 48% to 32%.

The 33 points on the scoreboard belie the fact that this was actually a very poor offensive day for the Washington offense; the Commanders scored on a pick six and two of their other touchdowns came on short fields following the day’s other two turnovers.

Marcus Mariota did throw three touchdowns but also only averaged 5.3 net yards per attempt (the NFL average is 6.3).

That said, Mariota’s day might not have been as bad as the 33rd percentile dropback success rate implies, as Washington’s 42% mark wasn’t far off from what the Seahawks defense allowed last year.

While the two interceptions were costly, it was a nice return to action for Sam Darnold, even after adjusting for the opponent.

Seattle was hurt by a poor day from its running game and a 3-for-10 performance on late downs.

To its credit, the Washington defense has been way better against the run in 2026 than it was in 2025, improving from 29th to third in rushing success rate allowed. It still stinks against the pass (30th in dropback success rate allowed this season), setting it up to be one of the league’s biggest pass funnels — the Commanders have allowed the league’s second-highest pass rate over expectation (+7.2%).

Terry McLaurin still has a noticeable gap between his fantasy expectation (43.4 expected points⁴) and his fantasy reality (30.1 realized points) through three games, while only eight receivers have underperformed by more. Sunday’s 6-catch, 77-yard performance with a touchdown did help bring the two numbers closer together.

Jaxon Smith-Njigba really is just doing it again; after his 2025 breakout, he is the top-scoring PPR fantasy player (including quarterbacks) and is third in expected fantasy points among receivers, running backs and tight ends. He is commanding a 38% target share, the highest at his position,⁵ with a 9.7 aDOT and two end zone targets per game.

The Colts overcame three Daniel Jones turnovers to win the battle of 0-2 AFC South teams, thanks in large part to late-down execution and four Spencer Shrader field goals (including the eventual game-winner with 16 seconds left).

The Colts were 9-for-17 on third/fourth down, while Houston was 4-for-12.

Shrader’s four field goals added 4.5 expected points, as they came from 39, 58, 37 and 53 yards out.

The Colts’ offensive performance against the stout Texans defense marked a reversal from what we had seen in Weeks 1 and 2; Indianapolis came into the game ranked eighth in early-down success rate but 25th in late downs.

Houston’s passing attack was okay in terms of efficiency, but the solid success rate also came with negative plays and a lack of explosiveness.

C.J. Stroud was sacked on three of his 30 pass attempts (a 10% rate compared to the league average of 6.4%) while the Texans only managed two pass plays that gained at least 20 yards (a 39th percentile big play rate).

The Texans running game continues to be a disaster.

Josh Downs took full advantage of Alec Pierce’s absence, leading Indianapolis in targets with 11. He had a solid fantasy afternoon, but the quantity and quality of these targets suggests it could have been even better.

Tyler Warren had a nice afternoon, but continues to mostly see targets around the line of scrimmage; his average depth of target (aDOT) on Sunday was 2.8, while it is just 1.7 for the season. If this continues, his ceiling is significantly capped, especially in leagues that don’t have full point-per-reception scoring.

Jonathan Taylor could only turn his 23 carries into 68 yards. The tough matchup didn’t help, as the Colts’ 37% success rate on designed runs was only slightly worse than the mark Houston allowed in 2025.

This was a really strange one; the yardage (350 to 348 in Buffalo’s favor) and success rates (both teams had marks of 46%) were both virtually even and the Chargers won the turnover battle 5-1. This would usually suggest a comfortable Chargers win, but the Bills clawed back with an edge on the other high-leverage plays.

Buffalo also had an edge on special teams, netting more than three expected points apiece on field goals⁶ and kickoffs.

It was an atypically tough day for the Bills passing game but Buffalo was able to move the ball at will on the ground. James Cook ran for 154 yards on 24 carries (6.4 yards per carry) with a 63% success rate and a touchdown (though he did lose a fumble).

Josh Allen added a pair of rushing touchdowns.

Allen and Cook split Buffalo’s goal-to-go rushes with two apiece.

DJ Moore earned a 39% target share, getting 10 passes thrown his way while the next-most-targeted Bill had four.

Omarion Hampton failed to score on three goal-to-go carries and might be leaving the door open for Keaton Mitchell to take on more work.

Hampton had 15 carries, while Mitchell had eight and Kimani Vidal had one. Mitchell was arguably more effective than Hampton, as he had a 75% success rate compared to Hampton’s (still solid) 47%. Mitchell also had three targets and a touchdown through the air (Hampton was only targeted once).

Per Jahnke, the snaps were: 29 for Hampton, 26 for Mitchell and 17 for Vidal; the routes run were: 14 for Mitchell, 13 for Vidal and six for Hampton.

The Steelers gained 5.5 expected points on a muffed punt (the “epa_turnovers” in the chart above is only for plays from scrimmage).

After a couple of solid games to start the year, the Bengals defense on Sunday bore a much closer resemblance to their awful unit from 2025.

Like Week 1 (but unlike Week 2), Aaron Rodgers attempted to stretch the defense with deep balls. His 11.9 aDOT was the second-highest among starting quarterbacks in Week 3⁷ and he attempted eight passes at least 20 yards down the field.⁸

Rodgers only completed two of those passes and actually had a lot more success on passes within nine yards of the line of scrimmage. Pittsburgh generated nearly 11 EPA on these short passes with a 71% success rate, compared with 0.61 EPA and a 31% success rate on his passes at least 10 yards down the field.

On the rare occasion Jaylen Warren had a backfield all to himself, he took full advantage of the juicy opportunity afforded to him by Rico Dowdle’s absence. He had 17 of Pittsburgh’s 19 running back carries and was targeted four times.

Tee Higgins made the most of his opportunities, catching six of seven targets for 90 yards and a touchdown. His expected stats? 3.9 receptions, 60.5 yards, 0.3 touchdowns.

Ja’Marr Chase had 12 targets, and was one of three Bengals with more than three targets on 37 Joe Burrow throws (joining Higgins and Dohnte Meyers, who had five targets).

The Lions’ 59.4% success rate was the third-highest mark for any offense this season and the 12th-best since the start of the 2025 season.

Geno Smith and the New York passing game put in an admirable effort trying to keep up with the Lions, but the Jets got next to nothing from their running game. They managed just a 24% success rate and 46 yards on 17 designed runs (2.7 yards per carry).

This had less to do with Breece Hall leaving the game with an injury and more to do with general ineffectiveness. Hall had a 23% success rate and 32 yards on 13 carries (2.5 yards per carry), while backup Braelon Allen was only successful on one of his four carries.

Jahmyr Gibbs is another player you can run out of superlatives talking about, especially from a fantasy football perspective. He is in the perfect marriage of talent, role and offensive environment. No player in the league can top his 86% designed rush share, while he has also received over 19% of Detroit’s targets. Add in the goal line role⁹, high-powered offense, indoor home stadium and skill to make the most of all of the above, and it’s not hard to see why he was the consensus top pick in fantasy drafts this summer.

Cleveland was 7-for-18 on late downs; Carolina was just 4-for-18.

KC Concepcion and Harold Fannin both had nine targets on 31 Browns throws but otherwise had vastly different days.

Fannin caught seven passes for 51 yards and two touchdowns; Concepcion managed just two scoreless catches for nine yards.

Concepcion’s poor catch rate looks even worse given his aDOT of just 4. On the positive side, he led the Browns in routes run (36 on 38 dropbacks) and was on the field for 60 of the team’s 70 snaps, per Jahnke.

Quinshon Judkins had 18 carries compared to just one for Raheim Sanders, but the two backs had a similar number of snaps: 36 for the former, 30 for the latter (again according to Jahnke). Sanders was targeted five times on 17 routes, while Judkins had two targets on 12 routes.

The Raiders won the turnover battle 4-1. There were seven fumbles (four for New Orleans and three for Las Vegas) in the game and Las Vegas recovered five of them (three on defense, two on offense). Two of the Saints’ lost fumbles came on Tyler Shough sacks, while the second-year passer also threw an interception.

The three turnovers were obviously costly but Shough and the Saints passing game were otherwise efficient through the air.

More than half of New Orleans’ 40 targets went either to Chris Olave (13 targets) or Juwan Johnson (8).

Ashton Jeanty continues to have plenty of opportunities, as he had nearly 80% of the Raiders’ running back carries while also getting three targets. The problem has been efficiency, as he only ran for 56 yards on 19 carries (2.9 yards per carry) while posting a 21% success rate on the ground.

For the season, he is sixth in the league in designed rush share (77%) but is only averaging 3.3 yards per carry. This is in spite of what appears to be adequate run blocking, as Las Vegas is 15th in ESPN’s run block win rate.

That said, per NFL Next Gen Stats, only David Montgomery has faced a higher rate of eight or more defenders in the box. Jeanty is 26th out of 43 qualified running backs in terms of rushing yards over the NGS expected yards model.

It was an amazing game from Brock Purdy that looks even better once you consider the 49ers had almost none of their top wide receivers available after Mike Evans left the game with an injury. Ricky Pearsall, Christian Kirk, De’Zhaun Stribling and Demarcus Robinson were already all out.

Evans did have a nice day before leaving with an injury, but the chart above is crediting him with the touchdown Deebo Samuel scored, since Evans caught the pass on that play before lateralling to Samuel.

Samuel therefore has the odd 0-catch, 80-yard, 1-touchdown receiving line (he otherwise was not targeted, but did get two rush attempts).

After two weeks of a somewhat equitable playing time split with Kaelon Black, Christian McCaffrey got the bulk of the backfield work on Sunday. McCaffrey’s advantage in snaps, carries and routes run were 45-8, 15-4 and 23-2, respectively, per Jahnke’s data.

Michael Wilson had 17 targets (34% share), including four in the end zone. He turned these into 11 catches for 89 yards and a touchdown.

Jeremiyah Love had 21 carries to Tyler Allgeier’s two; the routes run out of the backfield were more evenly split (22 for Love, 20 for Allgeier), per Jahnke. The two backs had a goal-to-go carry apiece, while Jacoby Brissett had two himself, including the Cardinals’ lone rushing touchdown on a quarterback sneak.

Cam Skattebo got 20 carries but only ran for 60 yards, failing to score on four goal-to-go carries. He did manage to catch three of his four targets for 40 yards.

Wan’Dale Robinson revenge game!

Minnesota had a big special teams day: the Vikings returned a punt for a touchdown, while kicker Will Reichard was 3-for-3 on field goal attempts (with makes from 54, 56 and 43 yards out).

The two worst offensive performances of the 2026 season by success rate were in this game. Tampa Bay can’t use the Baker Mayfield injury as an excuse either, since he only managed a 28% success rate on his 40 dropbacks before exiting.