Josh Allen

Josh Allen (QB, BUF) — 2026

Verdict

HOLD, high confidence, at ADP 27.8 (FFC PPR mocks, 2026-07-07) — QB1 off the board, ~20 picks ahead of QB2 Joe Burrow (47.6). Allen is the archetypal dual-threat elite — 36.2 rush yds/gm, 28% of team red-zone carries, 17 inside-the-5 carries, elite CPOE, secure job, retained play-caller — the *only* profile worth an early 1QB pick per qb.md §10, and the market prices him as exactly that. An xTD-anchored projection that regresses his +5-TD rushing overperformance and applies the first age-30 haircut still lands him ~350 points, at or near the top of the position, but it removes the excess-value case at a full-round premium over Burrow/Lamar. Profile and price agree; no "market is wrong" thesis survives the xTD math in either direction, so the verdict is HOLD — take him happily if he slips toward pick 32+, and let someone else pay a round-2 price.

Projection & comps

Bottom-up, two components, 4pt pass TD / −1 INT / 0.04 pass yd / 0.1 rush yd / 6 rush TD / −2 fum (scoring assumed, see header). Baseline usage from team profile (data/team-profiles/BUF.md, 2026-07-07): ~64 plays/gm, ~55% pass rate, PROE ≈ −3%, Vegas win total 10.5 (positive script).

Passing (median): 16.5 games × 30.0 att/gm ≈ 495 att (2025: 28.8/gm, 2024: 30.2/gm; DJ Moore nudges neutral pass rate up marginally per team profile) × 7.6 YPA (2-yr 7.7–7.85, elite OL retained, slight regression) ≈ 3,760 yds. Pass TDs anchored to passing xTD rate ≈ 5.0% (2025 xTD 24.6 on 467 att; 2024 xTD 23.1 on 484 — scratch calc from nflverse pbp, 2026-07-07) → 25 pass TD. INTs from TWP, not last year's count: blended FTN INT-worthy rate 2.9% × ~545 dropbacks ≈ 16 worthy throws → 9 INT.

Rushing (median — projected separately; this is the floor): age-30 season (b. 1996-05-21, Sleeper 2026-07-07), first year of the qb.md §3 re-projection window, so haircut the 2-yr baseline (6.7 att/gm, 34.7 yds/gm) ~7–8% despite zero observed decline: 6.3 att/gm → ~104 carries, ~525 yds (31.8/gm). Rush TDs anchored to rushing xTD (8.8 in 2025, 7.3 in 2024) plus a +1 structural sneak adjustment — league-average field-position TD probabilities underrate QB sneaks, and Allen converted 8 TDs on 20 sneaks in 2025 with the package retained under Brady → 10 rush TD (regressed from 14 actual).

2025 actual in this scoring: 374.6 in 16 gms (23.4 PPG); 2024: 380.3 (23.8) — both computed from data/stats/<yr>/passing.csv + rushing.csv. The median sits below both actuals by design: −4 rush TDs of xTD regression (~24 pts) plus the age haircut. Even the floor is roughly a QB8–10 season; that floor is what the pick buys.

Games risk: medium — 7.0 carries/gm sits just under the ≥8 auto-nudge (qb.md §3), but he's 30, absorbs contact voluntarily, and played weeks 16+ and the playoffs on a broken fifth metatarsal (surgery late Jan 2026 — CBS Sports; fully cleared by April and a full OTA/minicamp participant — NFL.com, profootballrumors.com 2026-04, SI.com June 2026). Durability record itself is elite (16 gms each of last two seasons).

Comps (league scoring, computed from season lines): Josh Allen 2024 (~380/16 gms), Josh Allen 2022 (~415/16), Lamar Jackson 2023 (~348/16), Jalen Hurts 2024 (~330/15), Lamar Jackson 2024 (~442/17 — the ceiling-buster). No local data/projections/ to cross-check — noted as unavailable.

Usage profile (opportunity core, qb.md §2)

All 2025/2024 REG figures from nflverse pbp scratch computation (2026-07-07) unless noted; season lines from data/stats/<yr>/.

Efficiency (qb.md §5): EPA/dropback +0.161 2025 (good) / +0.305 2024 (elite). CPOE +3.5 2025 / +0.8 2024 (nflfastR pbp; NGS CPOE model: +1.55 / +0.78 — both elite-to-good, QB-owned, sticky). aDOT 7.30 (NGS 7.14) 2025, down from 8.33 — deep-ball rate 11.3% (from 15.5%) as Brady's offense got quicker/flatter; ceiling now leans more on rushing TDs than deep passing. PA rate 23.6% (up from 18.1%). The two 2025 blemishes: FTN INT-worthy rate 3.43% (mid band, up from an elite 2.40%; 10 INTs vs 19 worthy throws = ledger roughly fair, no luck to buy or fade) and pressure-to-sack 26.8% (40 sacks) vs 9.3% (14 sacks) in 2024 — a concern-band spike (participation.csv join, pressure rate ~27% both years). 2024 was historically elite sack avoidance; 2025 is the payback plus interior-OL leakage. True level is between; the open LG battle (team profile) keeps interior pressure a live watch item.

Archetype: Dual-threat elite (§10) — the only early-pick-worthy 1QB profile. Pattern watch: §11 rushing-QB aging — age 30 is the re-verdict window; the decline hits the premium, not the job.

Context (from data/team-profiles/BUF.md, built 2026-07-07)

• Play-caller continuity through a HC change: McDermott fired post-divisional-round; OC Joe Brady promoted to HC 2026-01-27 and keeps calling plays (Buffalo News, 2026-02-07). He has called Allen's plays since Nov 2023 — the designed-run and sneak packages carry forward, and Brady publicly declined to add run limits ("We're not there yet" — SI, July 2026).
• O-line: top-5 unit — PBWR 4th, RBWR 1st (ESPN 2025 win rates), Sharp #3 for 2026; 4/5 starters return. The crack: an open LG battle (Anderson/Corbett/rookies) → early-season interior-pressure risk that feeds the sack problem.
• Pass-catchers: DJ Moore acquired for the #60 pick (2026-03) — first true perimeter claim of the Brady era, reunion with his CAR 2020 play-caller — plus Shakir, Kincaid (5th-yr option exercised), Coleman, Cook. Only ~12% of targets vacated; continuity check passes → 2025 efficiency carryover is believable (qb.md §4).
• Script/volume: win total 10.5 (BetMGM, 2026-05-20) → positive script, ~64 plays/gm at ~55% pass → ~30 att/gm for Allen. This offense wins on efficiency (+0.140 EPA/play, best-in-class 2025), not dropback volume — the passing ceiling is structurally capped, which is precisely why the rushing floor carries the profile.
• Job security: absolute — 6-yr/$330M (Mar 2025), backup is journeyman Kyle Allen (tier C). Benching/bridge risk zero.
• Health: broken right fifth metatarsal (Week 16, 2025), played through it, surgery late Jan 2026, fully cleared by April; full OTA + minicamp participant, no reported limitations (NFL.com; profootballrumors.com; SI.com injury roundup, June 2026).

Tripwires (re-run this eval if any fire)

• Right-foot setback — any camp/preseason report the fifth-metatarsal issue recurs or limits him → games risk jumps, floor drops.
• Reduced run package reporting — Brady or beat reports describing fewer designed runs/sneaks or a "protect Josh" install (he is currently on record against limits, SI July 2026) → rush TD anchor collapses toward 6–7, verdict → FADE.
• ADP movement — slides past ~32 (into the round-3/4 turn) → TARGET; rises inside ~20 → FADE.
• LG battle resolves badly + preseason interior pressure → 2025's sack spike persists; trim median ~10 pts.
• League settings confirmed different from assumptions — 6pt pass TD or superflex materially raises his relative value (re-verdict likely TARGET); half-PPR doesn't touch QB scoring here.