Through three games of 2026, the Kansas City Chiefs have quietly rewired the pre-snap look of their offense — and, downstream of that, the job description of Xavier Worthy. It is a very small sample. But the mix shift is large enough to describe what the offense is asking each pass-catcher to do right now, which is not what their 2025 usage said they were.

The formation shift

Kansas City has run 191 offensive plays (pass + run) through three games — a 31–10 home win over Denver in Week 1, a 33–30 home win over Indianapolis in Week 2, and a 24–10 win at Miami in Week 3. Of those 191 plays, 77 (40.3%) have come from under center. The three prior seasons, that share was 17.0% in 2023, 21.6% in 2024, and 19.3% in 2025.

That's roughly a doubling. It is also worth stating plainly: the league is trending the same direction. Non-shotgun snaps leaguewide have gone from 28.0% (2023) to 29.5% (2024) to 34.4% (2025) to 41.3% so far in 2026. So Kansas City has moved from well below the league norm to right at it. The story isn't that the Chiefs invented anything; it's that a team that had spent three years as one of the more shotgun-heavy offenses has, in three weeks, joined the league median.

The play-type mix inside each formation looks about the same as always: shotgun 66.7% pass (114 plays), under center 33.8% pass (77 plays) in 2026 — essentially identical to 2024's 67.3%/34.0% split. So this isn't a run-heavy pivot dressed up as a formation change. It's the same balance, moved forward two yards.

What's happening on the pass plays

Through those 191 plays, here's how targets have distributed among Kansas City's top four pass-catchers:

Kelce, Rice, and Worthy are all being targeted, on average, within seven yards of the line of scrimmage. The one player working deep is Tyquan Thornton, and it isn't close: his 20.1-yard average depth of target is roughly three times any of the other three.

Worthy vs. Thornton, then and now

In 2025, Worthy was Kansas City's vertical outlet. On 73 targets he ran a 12.9 aDOT with +0.20 EPA/target. Thornton was the extreme deep specialist on a smaller diet — 37 targets, 26.9 aDOT, +0.71 EPA/target. Kelce (6.8 aDOT) and Rice (4.3 aDOT) worked underneath.

Three games into 2026, that map has shifted. Worthy's aDOT has fallen from 12.9 to 6.3. Thornton's has come down too — from 26.9 to 20.1 — but only because he's caught a couple of intermediate throws inside a still tiny sample; he remains, by a wide margin, the deepest-average target on the roster. Split by formation, the difference is even sharper: on the 11 shotgun targets Worthy has seen this year, his aDOT is 1.2 yards. On his four under-center targets, it's 20.5. Whatever the underlying design is, the shotgun snaps have him working at or behind the line, and the deep opportunities are going to Thornton (17.6 aDOT from shotgun on five targets, 26.5 from under center on two).

The efficiency numbers cut against Worthy so far — −0.47 EPA/target on 15 looks is bad — but at 15 targets across three games this is a description of role, not a verdict on production. One busted screen is a big fraction of that number.

The under-center pass menu, so far

Passing off of these under-center looks has actually worked. On 26 under-center pass attempts, Kansas City is averaging +0.47 EPA/play; on 76 shotgun passes, +0.16 EPA/play. The under-center throws also lean deeper by design — Kelce's under-center targets have averaged a 14.3 aDOT (three of them), Rice's 16.5 (two), Worthy's 20.5 (four), Thornton's 26.5 (two). The shotgun passing tree, by contrast, is where the short stuff lives, and it's where Worthy's average depth collapses.

Read that way, the WR22 pick isn't being demoted to underneath work; the offense has partitioned itself. Under-center dropbacks have become the vertical menu (small n, but consistent across all four receivers). Shotgun dropbacks are the quick-game menu, and Worthy is being used inside that quick game more than he was in 2025.

What this is and isn't

Three games is three games. On 15 targets a receiver's aDOT can move four yards on a single deep shot; on seven targets Thornton's can move ten. And Kansas City still hasn't played a divisional road game — the schedule includes trips to Denver (Week 8), Oakland already played, plus games at the Chargers and Seahawks that will stress the offense differently than Denver, Indianapolis, and Miami have.

What is real enough to describe now: Kansas City is snapping under center at roughly double its prior-three-year rate; that shift has coincided with a genuine partitioning of the passing tree by formation; and Worthy, whose 2025 role was to threaten deep, has spent his shotgun snaps averaging just over a yard downfield while Thornton has absorbed the vertical work.

If it holds, this is a new usage pattern for Worthy, not a demotion — his volume (15 targets in three games) is on pace with his 2025 rate. What's changed is the depth those targets travel. I'd want another month before calling it a permanent role, but the tape and the numbers agree on what it is right now.