Week 5 ran from Thursday, October 8, when Tampa Bay won 24-16 in Dallas, to Monday night, when Buffalo won 37-34 against the Rams. Kansas City and Carolina had byes, so it was 15 games. Two teams are 5-0, San Francisco and Minnesota, and Kansas City is 4-0. Three are 0-5: Miami, the Chargers and Tennessee. The Giants and Washington tied 27-27. Ten teams are 3-2, the most crowded record in the league.
Five games is where standings start to feel permanent. They are not, and the useful question is whether a record is telling the truth about a team: close wins can carry it ahead, close losses behind. So this issue builds a ruler for five games from the same ten seasons as the last two, 2016-2025: every team's first five games against the rest of its regular season, 320 team-seasons in all.
Here is the ruler. Each number is a correlation between something a team did over its first five games and how often it won afterwards; 1.00 would mean a perfect guide, zero no guide at all.
- Record, as a share of games won: 0.41.
- Point differential, per game: 0.43.
- Net expected points added (EPA) per play, which is how many points each offensive snap added to the expected score minus how many each defensive snap gave up: 0.42.
- Record in one-score games, decided by eight points or fewer, against the record in one-score games after: 0.09.
The first three are a tie: at five games, the standings, the scoreboard and the play-by-play carry about the same information. The fourth is close to nothing. Winning close games early says almost nothing about winning them later.
Two more lines from the same ten seasons update the ladder from Issue 9. Five games of offense predict the rest of a team's offensive season at 0.56; five games of defense predict its defense at 0.32. Offense keeps firming as games pile up (0.50 at three, 0.54 at four). Defense barely moves (0.29 at three and at four).
Because the three main measures are equally good, the teams worth reading are the ones where they disagree.
Ten teams are 3-2. Net EPA splits them into two different futures.
From 2016 to 2025, 96 teams started 3-2, and 50 reached the playoffs. That is a coin flip, and it hides two groups. The 3-2 teams that ranked in the top ten by net EPA per play after five games made the playoffs 25 times in 39, 64%, and won 60% of their remaining games. The 3-2 teams that ranked 17th or worse made it 12 times in 33, 36%, and won half their remaining games. Same record, nearly double the playoff rate.
Point differential draws the same line. The 3-2 teams that had been outscored over five games reached the playoffs 7 times in 22 and won 43% of the rest; those with a positive differential made it 42 times in 72.
This season's ten 3-2 teams sort cleanly. Cincinnati ranks seventh by net EPA and Seattle tenth, the two in the top ten. New England (11th), Detroit (14th) and Green Bay (15th) sit in the middle. Five are in the bottom half: Chicago 17th, the Raiders 18th, Denver 20th, Pittsburgh 21st and Cleveland 24th.
Cleveland is the clearest case of a record running ahead of a team. The Browns have been outscored by 19 points, the 25th-ranked differential, and are 3-1 in one-score games. Their offense ranks 26th by EPA per play. Denver (-14) and Green Bay (-8) are also 3-2 while being outscored, though Green Bay's 31-7 win over Chicago had the best net EPA of Week 5. The Giants, 3-1-1, are a cousin: +3 in point differential, 22nd by net EPA.
The Raiders are the case last week's issue flagged. Its verdict was "the record, yes; the defense, wait," and the defense did not wait long. It ranked fourth by EPA allowed per play after four games, then gave up 45 points in New England, where Drake Maye went 26 of 34 for 374 yards and four touchdowns and ran for a fifth. The defense ranked 30th for the week and now ranks 19th over five games. The ruler is cold: of the 50 defenses that ranked top five after five games, 13 stayed top five and 16 finished in the bottom half.
The disagreement runs the other way too. The Rams are 2-3 with the fifth-best net EPA in the league and a point differential of -1, and lost 37-34 to Buffalo on Monday. Over ten seasons, 12 teams were 2-3 with a top-ten net EPA; 5 reached the playoffs and they won 54% of their remaining games. The 50 teams at 2-3 in the bottom half made it 11 times.
Minnesota's 5-0 is real. No unbeaten team in ten seasons has carried an offense this far down.
Minnesota beat New Orleans 23-13 on Sunday and is 5-0 because of its defense. It ranks first at -0.21 EPA allowed per play (on defense, a negative number is good), it is the only defense in the top ten in all five weeks (10th, 9th, 4th, 4th and 4th) and it has allowed 64 points, the fewest in the league. Through Week 4 it sent five or more pass rushers on 72.5% of opponents' charted dropbacks, first in the league.
The ruler starts in Minnesota's favor. Fifteen teams started 5-0 from 2016 to 2025, and 14 reached the playoffs. The miss was the 2016 Vikings. Every one of the 15 ranked in the top ten by net EPA after five games, and Minnesota ranks sixth, so by the play-by-play this 5-0 looks like the others.
The offense does not. Minnesota ranks 28th by EPA per play. The lowest-ranked offense on any of the 15 unbeaten teams was 17th, and that team was the 2024 Vikings, who went 9-3 the rest of the way. Widen it to every team with four or more wins after five games and the lowest offense ranked 26th. Three teams were there (the 2019 Bills, 2020 Bears and 2022 Cowboys), and all three made the playoffs. Minnesota is below all of them.
That matters because defense predicts itself at 0.32 over five games and offense at 0.56. Minnesota's record is built on the side of the ball that repeats less. The defenses that ranked first after five games are a fair guide: of the ten, five stayed top five the rest of the way and nine stayed top ten. That is a high floor, not a guarantee.
The close games are the other caveat. Minnesota is 3-0 in one-score games, and one-score records carry at 0.09. Even the 18 teams that won all of at least three close games in their first five won 67% of their later ones, not all of them.
Sunday in New Orleans was a fair sample. The defense intercepted two passes and ranked fourth for the week; the offense ranked 20th, though Kyler Murray went 24 of 30 for 229 yards and a touchdown to Justin Jefferson, who caught 6 of 7 targets for 79 yards. Minnesota also ranked fourth in defense across 2025, so last season agrees on the half that is carrying it.
San Francisco lost the matchup last week's issue pointed at and won the game anyway.
Issue 9 said Week 5 would put San Francisco's first-ranked offense against Seattle's second-ranked defense. Seattle's defense won that matchup. The 49ers' offense ranked 21st for the week at -0.12 EPA per play, its worst of the season, and Brock Purdy went 19 of 29 for 189 yards, two touchdowns and an interception.
San Francisco won 32-29 because its own defense ranked first in the league for the week, at -0.43 EPA allowed per play. Sam Darnold threw two interceptions, both to Renardo Green, and was sacked five times. The play that swung it came with 18 seconds left in the first half. A 49ers punt had just pinned Seattle deep. On first-and-10 from the Seattle 4, George Holani ran at left guard and was stopped at the goal line by Gracen Halton, who forced the fumble and recovered it in the end zone. A one-point lead became 21-13 at the half, on a snap the offense never touched.
Seattle led 26-21 in the third quarter. Purdy's 7-yard touchdown to Brandin Cooks and a two-point throw to Mike Evans put San Francisco back in front, and after a Seattle field goal tied it, Eddy Pineiro's 44-yard kick with two minutes left won it.
The offense is now second over five games, a thousandth of a point per play behind Buffalo, and San Francisco ranks first in net EPA. The ruler is generous to offenses this high: of the 20 that ranked first or second after five games, 14 finished the rest of the season top five and 19 top ten. Seattle's defense, tenth for the week, is still second over five games.
Houston won the 0-4 game. A 1-4 team that plays like Houston has almost no precedent.
Last week's issue noted that no 0-4 team had reached the playoffs in ten seasons, and flagged Houston and Tennessee meeting at 0-4. Houston won 26-3 in Nashville. The defense ranked third for the week, held Tennessee to a field goal and sacked Cam Ward four times; Ward went 23 of 36 for 170 yards with an interception. C.J. Stroud went 26 of 34 for 261 yards and a touchdown and ran for another, and Ka'imi Fairbairn kicked four field goals.
Now Houston is 1-4, and the ruler has two things to say that pull in opposite directions. The record: 53 teams started 1-4 from 2016 to 2025, and 4 reached the playoffs. The play: Houston ranks 12th by net EPA per play, with an offense 13th and a defense ninth, after that defense ranked 21st a week ago. It has been outscored by 2 points in five games and lost all three of its one-score games, by 5 to Buffalo, by 2 to Indianapolis and by 4 to Dallas.
That combination is close to unknown. Of the 67 teams that were 1-4 or 0-5 after five games in the last ten seasons, one ranked in the top half by net EPA: the 2016 Chargers, 11th, who went 4-7 the rest of the way. And the four 1-4 teams that did reach the playoffs ranked 22nd, 28th, 30th and 32nd by net EPA at the time, so none of them looked like Houston either. Houston plays like a middle team with a bottom-of-the-league record, and ten seasons have not produced enough teams like that to say how it ends.
Tampa Bay's 1-4 is the close-game kind: all five of its games were one-score games, and it ranks 19th by net EPA.
The three 0-5 teams have none of Houston's case. Fourteen teams started 0-5 in the ten seasons and none reached the playoffs. Miami ranks 31st by net EPA, Tennessee 32nd and the Chargers 28th, with the league's worst offense by EPA per play. Tennessee has scored 58 points in five games, the fewest in the league, and Miami 70, the second fewest.
Atlanta's run-first plan bent in Week 5. It is still the third-most run-leaning offense in football.
Issue 9 said a fifth game would show whether Atlanta's run game was a plan or a streak. The answer was some of each. Atlanta lost 30-23 in overtime to Baltimore after leading 23-15 with 2:35 left; Tyler Huntley threw a 7-yard touchdown with 44 seconds to go, Derrick Henry ran in the two-point try, and Huntley scored from the 1 in overtime.
The plan moved first. Atlanta's pass rate over expected (PROE), how much more or less a team throws than down, distance and score predict, was -1.7 for the week, close to neutral and only the 12th-most run-leaning of the 30 teams that played. In its first four games it had been -17.1, -11.2, -5.3 and -7.7. Michael Penix Jr. went 19 of 32 for 240 yards and a touchdown and found Drake London for 118 yards on 7 catches. Atlanta's 38 pass plays, sacks included, averaged +0.16 EPA.
The run game did not hold up. Atlanta ran 26 times at -0.29 EPA per rush, after +0.31 in Week 3 and +0.37 in Week 4. Bijan Robinson had 17 carries for 86 yards and Brian Robinson 7 for 16.
Over five games, the plan is intact. Atlanta's PROE is -8.5, third lowest in the league, behind only the Chargers and Miami. PROE carries from the first five games to the rest of the season at 0.51, and of the 30 teams that ranked in the three lowest after five games over the last ten seasons, 21 stayed in the ten lowest the rest of the way. The run lean is the part to believe.
The efficiency is the 0.56 question again, and five games of it now say average. Atlanta's offense ranks 19th by EPA per play, its net EPA 16th, and it is 2-3 with a point differential of -3 and an 0-2 record in one-score games. A middle team that has lost its close games.
Real or Noise
The verdicts, calibrated on five games. These grade the story, not the team.
- "Minnesota is the best team in football": the record is real; the offense is the risk. 5-0 teams made the playoffs 14 times in 15, and Minnesota's sixth-ranked net EPA fits the group. No team with four or more wins after five games in ten seasons had an offense ranked lower than 26th; Minnesota's is 28th, on a defense-first profile that repeats at 0.32.
- "San Francisco showed it's beatable in Seattle": noise. The offense had its worst week, ranked 21st, and the team won anyway behind the week's best defense. Of 20 offenses ranked first or second after five games, 19 finished the rest of the year top ten.
- "Cleveland is a playoff team at 3-2": noise for now. Outscored by 19, 24th by net EPA, 3-1 in one-score games. The 3-2 teams in the bottom half by net EPA reached the playoffs 12 times in 33.
- "Houston is back": the play, leaning real; the record, still a long way down. 12th by net EPA and outscored by 2 at 1-4. Four of 53 teams at 1-4 made the playoffs, and only one 1-4 or 0-5 team in ten seasons ranked in the top half by net EPA.
- "Atlanta's run game was a streak": the plan is real; the production is noise both ways. -0.29 EPA per rush in Week 5 after +0.37 in Week 4, but a five-game PROE of -8.5, third lowest, on a measure that carries at 0.51.
The Receipts
Verified numbers worth keeping in your pocket.
- After five games, record, point differential and net EPA per play predict the rest of a season's win rate at 0.41, 0.43 and 0.42. A one-score-game record predicts later one-score games at 0.09. Ten seasons, 2016-2025.
- 3-2 teams ranked top ten by net EPA reached the playoffs 25 times in 39; 3-2 teams ranked 17th or worse, 12 times in 33. This season has ten 3-2 teams: two in the top ten, five in the bottom half.
- Teams that started 5-0 reached the playoffs 14 times in 15; teams that started 0-5, none of 14. Minnesota and San Francisco are 5-0; Miami, the Chargers and Tennessee are 0-5.
Read more
- Tampa Bay's Week 5 win in Dallas and Dallas's side of it
- Minnesota, Kyler Murray's Week 5 and Justin Jefferson's
- San Francisco, Brock Purdy in Seattle and Christian McCaffrey
- Houston and C.J. Stroud in Tennessee
- Atlanta, Bijan Robinson's Week 5 and Drake London's
- Drake Maye's four touchdowns against the Raiders
- The boards behind this issue: team EPA, pass rate over expected and stat stickiness
- Last week's issue: Which first-month signals are real?
Follow the players you care about on Muffed, across all your leagues, and every Wednesday their weeks and the Read land in one email.
Next week: Cleveland's 3-2 meets Baltimore, the Raiders' defense faces Buffalo's top-ranked offense, Houston visits 4-1 Jacksonville, and Minnesota has its bye.
Methodology: Week 5 of 2026 is the 15 games from Thursday, October 8 through Monday, October 12; Kansas City and Carolina had byes, so their season figures cover four games. Team offense and defense are EPA per play on regular-season plays whose type is a pass or a run (scrambles included, penalty no-plays excluded), offense as the possession team and defense as EPA allowed; net EPA is offense minus defense. Ranks are among all 32 teams over Weeks 1-5 together unless a single week is named, in which case they are among the 30 teams that played. PROE is the mean of nflverse's pass_oe on regular-season plays that carry it. Records and point differentials are from nflverse's schedule file, with ties counted as half a win; a one-score game is decided by eight points or fewer. The five-game ruler is 2016-2025: each team's first five regular-season games against its sixth game onward (320 team-seasons), as Pearson correlations, with rest-of-season win rate as the outcome for record, point differential and net EPA, and each measure against itself for offense, defense and PROE. The one-score correlation uses the 276 team-seasons with at least two one-score games in each window. Ranks in the ruler are within each season, for the first five games and the rest separately; playoff rates use each team's record after five games and whether it played a postseason game that season. The Minnesota blitz figure is the share of charted dropbacks on which FTN counted five or more pass rushers, from the charting week snapshot through Week 4. Player lines are nflverse's weekly player stats. Data via nflverse; FTN Data via nflverse.