The easy story about George Pickens' first year in Dallas writes itself: the deep-ball artist finally landed a quarterback who could actually deliver the ball to him. On the volume it's fair enough — 138 targets, 93 catches, 1,429 yards, 9 touchdowns, a +0.643 EPA per target that ranked third among every receiver in the league with 100+ targets in 2025, trailing only Puka Nacua (+0.697) and Stefon Diggs (+0.690).

The problem is that when you actually go look at where the production came from, the deep-ball narrative falls apart. Pickens got fewer deep targets in Dallas than he did in any of his three seasons in Pittsburgh. His leap happened on the short and intermediate throws that were the one thing he couldn't do as a Steeler.

Table of Contents

1. The best receiver season Dak Prescott has ever produced
2. The same-QB Pickens-vs-Lamb split
3. The role inversion — short and mid became the engine
4. What to make of it

The best receiver season Dak Prescott has ever produced

Prescott's 2025 was a good one on its own — 631 dropbacks, +0.167 EPA per play, +2.20 CPOE. Inside it, Pickens' line is the best individual-target season any receiver has ever posted with Prescott throwing the ball. On 137 Prescott targets specifically, Pickens ran +0.637 EPA/target with a +6.96 CPOE for 92 catches and 1,405 yards.

That +0.637 clears every prior benchmark on Prescott's career resume at the 80-target minimum:

Nothing about Prescott's 2023 season with CeeDee Lamb — 180 targets, 133 catches, 1,740 yards — reads as anything less than a season-long masterclass, and it's still second here on a per-target basis. Pickens topped it.

The same-QB Pickens-vs-Lamb split

The cleanest way to see how much of a heater this was is to hold the quarterback fixed. Prescott's three biggest 2025 target buckets:

Lamb was very good — a +0.246 EPA/target season on 116 looks is a fine year for a No. 1 receiver. Pickens was 0.4 EPA/target better than him, off nearly identical average depth (11.24 vs. 12.06 air yards). Same quarterback, same offense, same defensive attention, similar route depth — Pickens simply generated more expected points every time the ball came his way.

The role inversion — short and mid became the engine

Now for the part that broke my priors. Bucket Pickens' targets by air-yards depth (short: <10, mid: 10–19, deep: 20+) across all four of his seasons:

Deep targets (20+ air yards)

His deep-target volume in Dallas was the lowest of his career, and his per-target deep efficiency was roughly in line with, not above, what he'd already done in Pittsburgh. The deep game isn't what changed.

What changed is everything underneath.

Short targets (<10 air yards)

Mid targets (10–19 air yards)

Three years in Pittsburgh: Pickens' short-target EPA sat between −0.10 and 0.00. Every screen and hitch that came his way was, on average, either a wash or an actively negative play. In Dallas that number jumped to +0.441 on nearly 80 attempts. The middle tier — dig routes, deeper crossers, seam-type stuff — went from good-but-not-elite (+0.34 to +0.65) to +0.977 EPA per target on 36 looks, with a completion percentage jumping from the 44–65% range in Pittsburgh to 69.4%. That's the part of the field where he'd been an inefficient volume receiver for three years, and it became his best zone.

Put another way: in 2024, Pickens' deep-target EPA (+0.802 on 30) was almost nine times his short-target EPA (−0.095 on 57). In 2025, his short (+0.441), mid (+0.977), and deep (+0.802) tiers all cleared +0.4 EPA per target. The distribution is flat now instead of top-heavy.

What to make of it

If you're grading the trade on why it worked, "Pickens finally got a QB who could throw him the deep ball" isn't really the story the data tells. His deep volume actually dropped from 30 to 22 targets, and his per-target deep efficiency was roughly where it had been in 2023 and 2024. What Dallas unlocked — or what Pickens unlocked in Dallas — was the underneath game: a short-throw floor that had been dragging his overall EPA down for three straight years, plus a mid-range tier that flipped from a modestly positive part of his profile into the most productive per-target zone on the offense.

You can read that a few different ways. I'd lean toward the boring-but-plausible one: better play design, better spacing next to Lamb and Jake Ferguson (101 targets, +7.89 CPOE from Prescott), and a quarterback whose short-and-intermediate accuracy was already a career strength. What the split doesn't look like to me is a stylistic transformation of Pickens as a route runner. It looks like the same player being asked to run a different mix of routes in a better situation, and connecting on the ones he used to bleed value on.

The obvious follow-up question — whether this repeats in 2026 with the same personnel, or whether it was closer to a shooting-star efficiency year for a receiver who's still averaging 22 deep looks — is one the data can't answer yet. But the 2025 tape is what it is. Dallas didn't get the deep-ball guy. They got a full-field one.