> ## 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.

# Chiefs at Dolphins, Week 3: What Would Have to Break Each Way
- URL: https://www.statsdigest.com/kc-mia-week-3-2026-quantitative-preview/
- Published: 2026-09-24T22:15:04.000Z
- Updated: 2026-09-24T22:16:15.000Z
- Description: Kansas City's early EPA edge over Miami is mostly a byproduct of a soft schedule, not a talent gap.
- Author: Bill Seifert
- Tags: Chiefs, Dolphins, EPA, NFL Week 3, betting odds, schedule adjustment

1. The raw two-game EPA picture
2. The opponent adjustment (this is most of the story)
3. A measurable matchup: intermediate throws

---

## 1\. The raw two-game EPA picture

Through two games (KC has run 142 offensive plays and faced 109 defensive plays; MIA has 106 offensive and 111 defensive), the pooled EPA/play looks like this:

| Team | Off EPA/play | Off plays | Def EPA/play | Def plays |
| ---- | ------------ | --------- | ------------ | --------- |
| KC   | +0.200       | 142       | −0.056       | 109       |
| MIA  | −0.112       | 106       | +0.172       | 111       |

Taken at face value that's a chasm — KC has been roughly a third of an EPA/play better on both sides of the ball. But two things need to happen before you take that at face value: check against last year, and check who each team played.

The 2025 full-season baselines (a much bigger sample: \~1,000 plays per side) tell a much more compressed story:

| Team | 2025 Off EPA/play | 2025 Def EPA/play |
| ---- | ----------------- | ----------------- |
| KC   | +0.035            | −0.020            |
| MIA  | −0.017            | +0.088            |

Kansas City was a modestly above-average offense and a slightly above-average defense in 2025\. Miami was a slightly below-average offense and a below-average defense. The 2026 two-game numbers are much more extreme in the same direction — which is what schedule luck tends to do.

## 2\. The opponent adjustment

Here is where the two-game snapshot mostly evaporates.

**Kansas City's opponents so far (Denver, Indianapolis) have been below-average offenses and terrible defenses through two games:**

| KC opponent | Their off EPA/play (2026) | Their def EPA/play (2026) |
| ----------- | ------------------------- | ------------------------- |
| DEN         | −0.087                    | +0.066                    |
| IND         | −0.026                    | +0.269                    |

League average through Week 2 is 0.004 EPA/play (identical on both sides by construction, across 3,835 plays). So KC's +0.200 offense EPA/play was posted against defenses running +0.066 and +0.269 — the Colts' figure is comfortably the worst any 2026 defense has produced across 138 plays. And KC's −0.056 defensive EPA/play was posted against offenses running −0.087 and −0.026\. Both of KC's marks were produced against the softer side of the league.

**Miami's opponents (Las Vegas, San Francisco) have been the opposite:**

| MIA opponent | Their off EPA/play (2026) | Their def EPA/play (2026) |
| ------------ | ------------------------- | ------------------------- |
| LV           | −0.066                    | −0.250                    |
| SF           | +0.298                    | −0.133                    |

San Francisco has been the best offense in the league by EPA/play through two games (+0.298 across 111 plays). Las Vegas's defense has been the best (−0.250 across 112 plays). Miami's ugly −0.112 offensive number and +0.172 defensive number were both compiled against the strongest possible early-season sample. The 35-13 loss at San Francisco in Week 2 does most of the damage on both ends.

**Put more plainly: strip out the schedule and the gap between these teams' two-game EPA numbers is largely a schedule comparison, not a team comparison.** Their 2025 marks were much closer, and there is nothing in two games of 2026 that would let anyone say the gap has actually widened. It might have. There's no evidence yet either way that isn't confounded by who they played.

## 3\. A measurable matchup: intermediate throws

The larger, more trustworthy sample is 2025's full pass-defense-depth splits. Two numbers stand out for this game:

- **Miami's intermediate defense (10–19 air yards) in 2025 allowed 0.800 EPA/play across 90 targets** — dead last in the league (rank 32 of 32). League average at that depth was 0.386.
- Kansas City's intermediate defense allowed 0.391 EPA/play across 85 targets — right at league average (rank 19).

So on the sample that's actually large enough to mean something, Miami's clearest weakness is the intermediate zone.

Does Kansas City attack that zone? In 2025 they did — 96 intermediate attempts (average air yards 14.1) at +0.471 EPA/play, essentially matching what the league did to Miami. Through two games in 2026, KC has thrown 28 "mid" attempts (roughly 5–14 air yards, avg 7.75) at +0.200 EPA/play, plus 8 deep shots. It's a small window, but the shape of the offense hasn't obviously changed.

Miami's own passing game showed the same intermediate strength in 2025 — 103 attempts at +0.430 EPA/play. Kansas City's intermediate defense in 2025 was league-average, not a wall. This is where the passing game is likely to be decided in both directions.

**Caveat that matters:** injury and personnel data for Week 3 is not populated in this dataset, and coverage matchups are decided by who's actually on the field. I don't know who's healthy for either team.

## What would have to be true

For Kansas City to cover a number this large:

- The 2026 offense/defense gap needs to be real, not a schedule mirage. That would require KC to reproduce +0.200 EPA/play or close to it against a defense whose 2025 baseline (+0.088 EPA/play allowed) was much stiffer than Denver's or Indianapolis's have been.
- KC needs to keep exploiting the intermediate zone where Miami was worst in the league in 2025.
- Miami's offense, which was competent against tough defenses in 2025, needs to look like the version that showed up in San Francisco.

For Miami to keep it close:

- The offense has to bounce back to its 2025 form (−0.017 EPA/play against a full-season slate) rather than the version that got shut out of the game in Week 2 against the league's best defense so far.
- The defense's 2025 intermediate hole has to be schemed around by whoever is calling coverages.
- The 2026 defensive EPA/play (+0.172 through 111 plays) needs to be read as what it likely is — a San Francisco tax — rather than a true talent estimate.

At this sample I'd stop short of saying which one is more likely. The honest read on two weeks of games is that Kansas City looks better and has played a much softer schedule, and there is not enough data yet to separate those two facts.