Our Projection Went 12-4 in Week 4 and finally Took the Title Race Away From the Rams

By - Matt B

Friday, October 9, 2026

Week 4 was the model's best Sunday of the year. The re-run still has the Rams on top, with Detroit, Jacksonville and Baltimore within reach.

For three weeks the Rams were the one thing our model would not argue about. The record slid from 14.8 projected wins in July to 12.9, the title odds slid from 44 percent to 38.5, and the Rams were still the favorite by a margin no other team was within twenty points of. After Week 4 they are still the favorite. The margin is five points. Los Angeles went to Philadelphia and won 24-20, got to 2-2, and came out of the re-run at 29 percent to win the Super Bowl, with Detroit at 24, Jacksonville at 15.5 and Baltimore at 12.5. In July those three had 19 percent between them. They now have more than the Rams.

That is the story of this run. The Rams did not get worse. The field got better, and the model has spent four weeks watching Jacksonville's defense allow 53 points, Baltimore's quarterback play at a 120 rating, and Detroit's offense project for 506 points, and has moved those three up while the Rams held still. The sixty-four games that have been played are in the books. The remaining thirteen weeks and the postseason were played out two hundred times with every player's four weeks of real production seeded into his line and the injury list as it stands on Tuesday morning. What came back is below.

The scorecard

Our matchup projections picked a favorite in all sixteen games in Week 4 and got twelve of them right, the best week of the season by a distance after 7-9, 7-9 and 8-8. The four misses were Cleveland beating Pittsburgh at 16 percent, Carolina beating Detroit at 24, and two coin flips, Dallas in Houston and the Giants against Arizona, that the model had leaning the other way by a point or two. The season record is 34 of 64, which is still not something to put on a billboard, and the projections below have absorbed every result.

The teams that have made the model look worst are, as ever, the ones whose projections have moved most. Cleveland has won three games it was given 24, 26 and 16 percent in and is 3-1 with a seven-win projection that has nearly doubled since July. Minnesota is 4-0 against a model that favored it once in four weeks. Kansas City and San Francisco are 4-0 and 4-0 and have been pulled four and three and a half wins up from their summer numbers. The Chargers and Texans were favored in four of their first eight games between them, with a fifth a coin flip, and are 0-8.

The title race has four teams in it

The Rams are 2-2 and project for 13-4, unchanged from last week, with a 99 percent playoff chance. The Philadelphia win was the first result since Melbourne that matched the number. The defense is doing exactly what the summer sim said it would, projected to allow 280 points, second in the league, with Byron Young's four sacks pushing his line from ten and a half to 13. The offense is the question. Matthew Stafford has six touchdowns and six interceptions through four games, and the model has cut his touchdown projection from 34 to 30 in a week while lifting his interception line to 15. The division is the other question. San Francisco is 4-0 with the head-to-head banked, Seattle is 3-1, and the Rams are two games back of a team that has not lost.

Detroit is the strangest line on the sheet. The Lions lost 32-26 in Carolina as a 76 percent favorite, are 2-2 with a defense that has given up 127 points, and saw their title odds rise from 19 percent to 24. The record projection slid from 13.4 wins to 12.6, which is still a 13-4 display, and the title odds rose because the Rams' fell and Detroit absorbed most of the difference. Jared Goff has nine touchdowns and no interceptions and projects for 36 and four, which would be the best ratio in league history. The offense projects for 506 points, the most in football by 50. The defense is projected to allow 360, up from 273 in July, and two road losses to Buffalo and Carolina are where the extra 87 came from.

Jacksonville went to Cincinnati and won 22-17, is 3-1, and has the best defense in football by every number the model produces: 256 projected points allowed, 15.5 projected interceptions, Josh Hines-Allen at 15 sacks and Travon Walker at ten and a half. The Jaguars project for 12-5 with a 97 percent playoff chance and 15.5 percent title odds, up from 9.5 last week and four in July. Trevor Lawrence has eight touchdowns and two interceptions at a 111 rating. The market opened the Jaguars at nine wins. The model is three over it, which is now the largest lean on the board.

Baltimore beat Tennessee 24-18, lost Lamar Jackson to an ankle injury that is expected to cost him a week or two, and doubled its title odds to 12.5 percent anyway. The Ravens project for 11-5 with a 90 percent playoff chance. Jackson was 68 of 96 for 967 yards, six touchdowns and one interception when he went down, and the model lifted his rating projection nine points in a week to 109. Derrick Henry has seven rushing touchdowns. The defense projects to allow 297 points, fourth in the league, with the second-fewest yards. The model has priced the absence as short. If it runs long, every extra game costs half a win.

Kansas City is 4-0 after beating Las Vegas 30-27, still projects for 12-5, and saw its title odds trimmed from ten percent to four. That is not the model doubting the Chiefs. It is the model pricing the three teams above, and a pass rush that projects for 32.6 sacks, last in the league, against a defense that gave up 27 to the Raiders. Kenneth Walker III has 537 rushing yards, the most in football, and projects for 1,692. Rashee Rice left the Raiders game with a hamstring and is out for several weeks. The Chiefs are the model's AFC West champion in nearly every simulated year. They are no longer its most likely AFC champion.

Atlanta scored 45 and the model put the Falcons back where it had them in July

The Falcons scored 16 points in two games with Cooper Rush at quarterback. They have scored 80 in two games since Michael Penix Jr. returned, the second of them a 45-24 win in New Orleans on Monday night in which Atlanta ran for 205 yards and five touchdowns. When we re-ran the season the Falcons came out at 11-6 with an 85 percent playoff chance and a five percent title shot, up from 10-7, 67 percent and two percent a week ago, and back to the eleven wins the summer sim projected before anyone knew Rush would start two games.

Bijan Robinson ran for 145 yards and two touchdowns against the Saints, has 494 yards and four scores for the season, and projects for 1,956 rushing yards and 19 touchdowns, the highest rushing projection in the league by 250 yards. Brian Robinson scored three times on Monday and projects for ten. The offense projects for 439 points, third in football, up from 393 and tenth. The defense is the hedge at 353 projected points allowed, 15th, and under nine interceptions. Atlanta is the model's pick to win the NFC South by a wide margin, and the Baltimore game at home this week is the one that tests whether Penix's two games were a real sample.

The NFC East flipped in a week

A week ago Philadelphia was the model's most likely NFC East champion at 10.6 wins. The Eagles then lost 24-20 at home to the Rams without DeVonta Smith, Dallas Goedert and a healthy Saquon Barkley, and the re-run has them at 9-8 with a 46 percent playoff chance, down from 83 percent, and half a percent title odds. The offense projects for 325 points, 23rd in the league, down from 388 in July. The defense is still fifth. The gap between those two numbers is the injury report, and the model has priced it.

Dallas took the position. The Cowboys won 34-30 in Houston and project for 10-7 with a 69.5 percent playoff chance, up from 8.8 wins and 37.5 percent, the largest upward move in the conference along with the Giants. CeeDee Lamb has 498 receiving yards, the most in the league, and projects for 115 catches and 1,565. Dak Prescott has eight touchdowns and one interception. The defense is still projected 21st, but that is up from 24th, and Rashan Gary's sack projection jumped from eight to twelve. The model now has Dallas winning the division more often than Philadelphia, and the Week 7 game in Philadelphia is where it expects that to be decided.

The Giants are 3-1, lead the division, and have climbed from 4.7 wins in July to 8-9 with a 16.5 percent playoff chance after beating Arizona 36-24 with Jameis Winston throwing three touchdowns and Deonte Banks returning an interception for the score that sealed it. The defense has allowed 79 points in four games. J.J. McCarthy is the backup for now. The model has him at zero attempts, and the day he starts the projection gets rerun from scratch. Washington is 1-3 and projects for four and a half, with Jayden Daniels expected back from his dislocated elbow on Sunday and Marcus Mariota now the one with the knee injury.

Where the arguments moved

Buffalo lost 29-26 at home to New England and fell from 9.6 wins to 8-9, with the playoff odds cut in half to 33 percent. The defense projects to allow 416 points, 28th, which is the number the July preview was built on when it had the Bills at under five wins. The model now has New England, at 9-8 with the head-to-head banked, as the AFC East favorite, and the Jets at 9-8 despite a 1-3 record because the defense has allowed 84 points and projects third in the league. The Rams at 94 percent are next for the Bills.

Cleveland beat Pittsburgh as a 16 percent underdog, is 3-1, and climbed from 5.7 wins to 7-10. The Browns are tied for the AFC North lead and project to finish last in it, because Baltimore, Cincinnati and Pittsburgh project for eleven, ten and nine and a half and the Browns play all three again. Deshaun Watson is at a 101 rating. The offense still projects for 266 points, 30th. The model has moved three wins on Cleveland since July and is still asking the Browns to lose nine of their last 13.

Pittsburgh paid for that loss, sliding from 10.3 wins to 9-7 with the title odds gone from three percent to half a percent. Aaron Rodgers has been sacked 13 times, the most in the league. New Orleans gave up 45 and fell from 7.6 wins to 6-11, with the defense now projected to allow 442 points, second-worst in football, while Tyler Shough projects for 5,106 passing yards and Chris Olave for 1,926 receiving, the highest in the league. Green Bay won in Tampa by three and recovered a full win to 8-9, still without Micah Parsons on the physically unable to perform list and Josh Jacobs on the exempt list, and now without Jayden Reed, who is reportedly facing season-ending neck surgery.

Minnesota is 4-0 and projects for 9-8, which is the model telling you it has watched the Vikings win four games by a combined 35 points with an offense that projects 21st and does not expect the margin to hold. The pass rush is the third-best in football by projection, Dallas Turner has six sacks and projects for 17, and the Vikings have allowed 51 points, the fewest in the league. The division is still Detroit's in most simulated years, by four projected wins, and the Week 8 game in Detroit is the one the model gives Minnesota eight percent in.

Houston lost its fourth straight by four points and is 0-4 with the best pass rush in football, Will Anderson Jr. at seven sacks and a projection for 23, and a quarterback with no interceptions. The model has the Texans at 8-9, down from eleven in July, and they are the second-largest underperformer of our matchup projections behind the Chargers, who were favored at 80 percent twice and are 0-4 with Justin Herbert at six interceptions. Los Angeles projects for 6-11, three wins below the summer number.

San Francisco is still 4-0 and still not the division favorite

The 49ers beat Denver 24-14 without Nick Bosa and Brock Purdy has now dropped back 113 times without being sacked, with eleven touchdowns and one interception at a 126 rating. The re-run holds San Francisco at 10-7 with a 68 percent playoff chance, which is where it sat a week ago and three and a half wins above July. The projection has stopped climbing because the model has priced most of what the roster is showing. The division has not followed. The Rams still project for three more wins than the 49ers and still win the NFC West in most simulated years, two games back and with the head-to-head lost. Seattle at home this week, where the model gives San Francisco 36 percent, is the first of the two road games that decide whether that holds.

The quarterback list

Eight teams have a quarterback story this week and the model has priced every one of them. Lamar Jackson's ankle is a week or two. Caleb Williams' hamstring is week to week and Tyson Bagent, who beat the Jets, starts in Green Bay. Baker Mayfield's thumb is at least three weeks and Jalon Daniels lost his first start by three. Jayden Daniels is on track to return for Washington while Mariota heals a sprained MCL. Jaxson Dart is out for the year and McCarthy is waiting behind Winston. Sam Darnold has started three of Seattle's four games with Drew Lock at a 127 rating in the other. Kyler Murray has started three of Minnesota's four with Carson Wentz in the other. Penix has started two for Atlanta and the Falcons have scored 80 in them.

What did not move

Seattle held at ten wins. Chicago climbed a fraction to ten with a backup quarterback and a defense that has allowed 65 points. Carolina held at nine after beating Detroit, with Bryce Young at a 100 rating and a 103 projected rating that is fourth among starters. Las Vegas is at nine with Kirk Cousins at eleven touchdown passes, the most in the league. Denver slid to nine with an offense that has scored 74 points. Indianapolis held at eight with Jonathan Taylor at six rushing touchdowns. Arizona and Washington are at five, Tampa Bay and Tennessee at three, and Miami, with one touchdown pass in four games, is at two.

Go argue with it again

The method has not changed. The model prices the players on the field, what they have actually done, who is actually available, and the schedule ahead. A week ago that produced a league with one title favorite and a lot of daylight behind it. Today it produces a league with four teams inside 17 points of each other at the top and a Rams team that has to win its division before it can be the favorite in anything. The difference is four weeks of Jacksonville's defense, four weeks of Lamar Jackson, and one Sunday in which the model's favorites went 12-4 and the three teams closest to the Rams all won.

Every one of the 32 team previews has been rewritten for this run: the projected record against the July and September ones, the four-game body of work, how our matchup projections did on those four games, the player lines that moved most since last week, and what the next month looks like by our numbers. Start with the full 2026 league projection, standings and odds, then go find out whether a month was enough to change your mind about your team.