> ## Content Index
> Fetch the complete content index at: https://www.statsdigest.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Week 6's Play-by-Play Numbers Told a Different Story Than the Final Scores
- URL: https://www.statsdigest.com/week-6-2026-what-actually-happened/
- Published: 2026-10-22T11:49:00.000Z
- Updated: 2026-10-22T11:48:59.000Z
- Description: Week 6's EPA per play exposed which final scores told the truth—and which ones badly understated the gap.
- Author: Bill Seifert
- Tags: EPA, NFL analytics, Week 6, Panthers, Chiefs, Texans

Week 6 of the 2026 season produced 14 games, and as always, a handful of final scores that don't match what the plays underneath them said. A single game is still a thin sample — 50 to 75 offensive snaps per team, nowhere near enough to rank anyone — but when the down-by-down pattern and the final margin line up, that's worth paying attention to. When they don't, the scoreboard can be actively misleading. Here's what the play-by-play from five of Week 6's games actually showed.

## Carolina's win over Philadelphia was a beating on every down

Carolina went into Philadelphia and won 25-14 — a result the common read treats as a mild upset. The snap-by-snap numbers say it wasn't close.

Across 73 offensive plays, Carolina averaged +0.119 EPA per play with a 43.8% success rate. Philadelphia, across 61 plays, averaged -0.171 EPA per play and succeeded on just 37.7% of snaps. That gap held on every down: Carolina was +0.167 EPA/play on 31 first-down snaps to Philadelphia's -0.292 on 23; +0.172 on 23 second-down snaps to Philadelphia's -0.011 on 22; and even on third down, where Carolina's number went negative (-0.142 on 17 plays), Philadelphia was still worse (-0.086 on 15).

That's 134 combined offensive plays pointing the same direction. It doesn't prove Carolina is a good team now — one week's sample still sits inside a lot of season-long noise, and none of this accounts for who Philadelphia's defense actually was on this particular Sunday. But a down-by-down sweep across that many snaps is a different, heavier kind of evidence than a single fourth-quarter possession deciding a close game.

## Kansas City's win over the Chargers was exactly as lopsided as the final score

The Chiefs beat the Chargers 22-6 at home, and the box score undersells it only slightly. Kansas City's offense produced +0.081 EPA per play on 55 snaps with a 52.7% success rate — a solid, unspectacular day. Los Angeles's offense was the actual story: -0.311 EPA per play across 63 snaps, a 27.0% success rate, and a passing game that managed -0.407 EPA on 41 dropbacks. The Chargers took four sacks, threw an interception, and lost two fumbles. Kansas City turned the ball over once, on an interception of its own, and that's close to the entire gap between "blowout" and "annihilation."

## Houston's win at Jacksonville was the week's most efficient football

The final score reads Jacksonville 31, Houston 34 — Houston won on the road in the week's best offensive showcase. Houston's offense ran at +0.225 EPA per play across 59 snaps; Jacksonville's wasn't far behind at +0.174 across 67\. Both offenses scored four touchdowns. Houston had five gains of 20-plus yards, Jacksonville six. This wasn't one offense beating up a bad defense — it was two offenses sustaining a well-above-average rate against each other for a full game, with Houston's handful of extra plays of efficiency enough to win by three.

## Atlanta's rushing day looked better in yards than in EPA

Chicago beat Atlanta 31-23, built mostly on the passing game: the Bears were +0.266 EPA per play on 40 dropbacks, while the Falcons were -0.246 on their 40\. On the ground, Atlanta actually averaged a strong 5.25 yards per carry on just 12 attempts — but those same 12 carries still graded out at -0.289 EPA per play. Chicago ran it far more (32 carries for 3.09 yards each, -0.248 EPA/play), landing at roughly the same per-play grade despite the worse per-carry average. It's a reminder that yards-per-carry and EPA can diverge sharply in a small sample — 12 carries isn't enough to call Atlanta's ground game either a strength or a weakness off this game alone.

## San Francisco won on the ground; Washington's passing game, not its running game, was the strong unit

San Francisco beat Washington 23-17 at home behind a run-heavy approach that worked: the 49ers ran for +0.483 EPA per play on 29 carries, well above their -0.050 EPA/play mark on 31 dropbacks, for a +0.208 EPA/play day overall across 60 total plays. On the other side, the more interesting number belongs to Washington: its passing offense actually out-graded its running game, +0.069 EPA per play on 42 dropbacks against -0.203 on 16 carries. The loss wasn't a case of Washington abandoning a good ground game too early — the ground game was the weak link all day, while the quarterback play held up fine by comparison.

## What to watch next week

None of this is opponent-adjusted, and none of it should be mistaken for a full-season verdict off one Sunday. But a few of these patterns are worth carrying into Week 7 as working hypotheses rather than settled scores — and this is opinion, not a forecast:

- If Carolina's first- and second-down efficiency shows up again next week against a competent defense, that's a real signal, not just one good Sunday against Philadelphia.
- If the Chargers' passing game posts another outing in -0.400 EPA territory, a bad week starts to look like a pattern worth worrying about.
- Houston and Jacksonville both played efficient enough football that either should be taken seriously going forward, regardless of who "won" this particular game by three points.

Reading only the final scores off Carolina-Philadelphia, Jacksonville-Houston, or the other three games here would have missed most of what actually happened on the field this week.

*All figures are from 2026 Week 6 play-by-play for the five games discussed above; per-team offensive snap counts in this sample ranged from 52 to 73\. None of the EPA rates here are opponent-adjusted, and a single game's sample is not a basis for ranking any team — it's a data point to weigh against what the next several weeks show.*