The young-QB premium isn't a premium. It's a discount.

What the consensus says

Every dynasty startup piece running this July agrees on one thing before they disagree on everything else: pay for QB age. FantasyPros' 2026 startup values piece frames the scarcity math directly, superflex leagues need roughly twice as many startable QBs as exist, so a young one is a multi-year moat. RotoBaller's July rankings and Footballguys' dynasty tiers both slot young arms ahead of statistically comparable veterans on age alone. Fantasy Six Pack's superflex guide puts it plainest: investing in a QB during his "rookie window" is the best long-term dynasty value there is. Dynasty League Football and FootballScout365 both build their startup boards the same way. Not one of them checks whether that premium survives once you control for how much the guy has actually thrown the ball.

The claim, in plain English

The whole test comes down to one question: does a quarterback entering next season at 25 or younger outproduce an equally-busy older quarterback? "Equally busy" is doing all the work in that sentence, so the test only compares QBs inside the same Y0 pass-attempt band (the exact cutoffs live in the study script linked below). The control variable, Y0 pass attempts, is the entire game here. A QB who threw 550 times is obviously going to outscore one who threw 220. Any real age effect has to survive inside bands of similar workload, or it isn't an age effect, it's a volume effect wearing an age costume.

How I beat on it

Y0-to-Y+1 pairs from player_stats_season.csv, QB position, 2015-2024 Y0 seasons (Y+1 reaching 2025), 6+ Y0 games, 150+ Y0 attempts, scored in Dynasdeez Half-PPR. n=271 pairs. Full two-gate framework: Session 97 spec. One honest note on the setup: the literal "young" cutoff our own known-error benchmark flags is age 23 and under, but that produces only 8 to 12 qualifying QB pairs across ten NFL seasons, not enough to trust. I widened the cutoff to 25 and under, still squarely inside the "rookie window" the 2026 pieces are describing (Drake Maye is 24, Caleb Williams is 24.8), to get a sample that clears 10-per-side in multiple volume bands.

What the data actually said

First swing: aggregate. QBs entering Y+1 at 25 or younger averaged 0.84 more points a game than everyone older, unconditional on volume. That's the kind of gap that could just be a bunch of high-volume young starters skewing the pool.

Because aggregates lie, band it by Y0 attempts next.

(Two thinner bands, 200-300 and 150-200, didn't clear 10 pairs on the young side and were excluded rather than trusted.)

Every single band that clears the sample bar points the same direction, and it's the opposite of what a hunt for "overrated youth" needs to find. Pooled across those three bands, the weighted gap is +1.05 points a game, in favor of the younger, equally-busy quarterback. I built this test trying to prove dynasty managers were overpaying for age. It wouldn't cooperate.

What the engine already figured out

SignalTuned's QB age curve already discounts age 22 and 23 by 20 to 23 percent off peak value before any of this ran, so I went in expecting to find the discount wasn't steep enough. I found the opposite. At matched volume, the players the curve discounts hardest are outproducing their un-discounted older peers. One tension worth naming directly: our own outcome-tracking benchmark flags the under-24 crowd, across every position combined, as over-rated by a wide margin. Running the identical test isolated to just quarterbacks flips that sign entirely. Whatever is driving the cross-position miss, RBs and WRs most likely given how young-skill-position hype usually runs, it isn't happening at QB. No config change follows from a test that failed to find what it was built to find.

The 2026 names

Real board ranks, pulled fresh off our 10-team superflex dynasty board.

Drake Maye, QB2, age 24. The single biggest name this test defends. Hold with confidence, his age isn't a hype tax on top of his production, it's a real, measured lift.

Caleb Williams, QB4, age 24.8. Same story. The DJ Moore trade has some of the market souring on his outlook, this test has nothing to say about weapon changes, only about age-at-equal-volume, and on that axis he's exactly where he should be.

Jaxson Dart, QB5, age 23.3. The clearest test of the thesis. He doesn't need a 500-attempt season to "prove" the premium, the 300-400 attempts band shows the biggest lift of the three (+1.40). Buy or hold before the market demands a bigger sample he doesn't need to give.

Anthony Richardson, QB45, age 24.3. The caveat made concrete. He clears this test's age cohort easily, and the board still has him outside the top 40. Youth is not a blank check, this study measures a same-volume comparison, and Richardson's issue has been volume and availability, not age. Fade the "he's only 24" argument specifically, hold your actual opinion of him on everything else.

What to do about it

Stop discounting a young, high-attempt-volume QB just because he hasn't been in the league long. The market thinks it's paying a premium for upside. The receipts say it's closer to a discount, the same-volume comparison favors youth by a full point a game or more, in every band with the sample to trust it. One honest caveat: this only covers QBs who'd already thrown 150-plus passes in a season, so it says nothing about a rookie who hasn't taken meaningful reps yet, that's a different, harder question this test doesn't answer.

Receipts

• The cohort test: scripts/h79_young_qb_age_premium.py
• Hypothesis record: data/hypothesis_backlog.json (H-79)
• Framework: Session 97 two-gate spec
• Board snapshot: Dynasdeez Nutz-config 10-team superflex dynasty rankings, generated locally 2026-07-24

← Back to all posts