James Cook III

Where He's Expected to Land

Nova projects 247 points at 16.1 PPR per game, a likely finish anywhere from RB8 to RB38, which is the widest way possible of saying "back-end RB1 with real slippage risk." The stat line underneath: 1,188 rushing yards, 9.9 rushing touchdowns, 31.5 catches for 280 yards, 2.1 receiving scores.

His DNA grade is B, baseline volatility, neither metronome nor lottery ticket. P(Tier-2 or better) sits at 60%; P(Tier-1) at 37%. In plain terms: more likely than not he's a weekly starter you never think about, and a bit better than a coin flip's-worth short of elite.

The Last Three Seasons

Pro Bowl nods in 2023 and 2024, then a 2025 line that led the NFL: 1,621 rushing yards on 309 carries at 5.2 per attempt, with 14 total touchdowns (per FantasyData and FantasyPros' Derek Brown, July 6, 2026). The trajectory is clean, and so is the durability, 50 of a possible 51 regular-season games the last three years.

What the table hides: the workload jumped 102 carries year over year, and the receiving role went the other way, 40 targets in 2025, down from a career-high 54 in 2023. His 2025 usage profile reads 58% of carries but just an 8% target share and a 0.127 WOPR. He's a rusher who catches, not a dual threat.

What Changed for Him This Offseason

Almost nothing, and after the 2025 contract saga, that's the story. His four-year, $48 million extension is signed, so this spring's absence from voluntary OTAs was noise, not leverage. "There is no such thing as holding out when it's voluntary," WGR 550 Bills reporter Sal Capaccio posted on X on May 27. Cook spent the time with family in Miami, showed up for mandatory minicamp in June, and told reporters his personal focus is ball security after six fumbles in 2025 (per ESPN's Alaina Getzenberg, June 17). He turns 27 in September having missed one game in three years, the age curve is knocking, but it hasn't entered.

What Changed Around Him

Everything else did.

Coaching. Buffalo fired Sean McDermott in mid-January and promoted offensive coordinator Joe Brady to head coach on January 27 (CBS Sports). Brady's units ranked fourth in scoring and rushing in 2025, so continuity is the base case, but Brady now answers to nobody on identity. NFL Films' Greg Cosell put it flatly on the Ross Tucker Podcast in June: "Now that Joe Brady is the head coach, he dictates how he wants the offense and the game to be played." Buffalo's run-first identity is what drove that rushing rank; if Brady nudges the pass rate toward league-normal, carries come out of Cook's pile first.

Teammates. The passing game got heavier: Buffalo traded with Chicago for WR DJ Moore this offseason (per ESPN). The run infrastructure got lighter: left guard David Edwards and swing tackle Ryan Van Demark left in free agency, and fullback Reggie Gilliam, a core lead blocker, signed with New England (SI's Alex Brasky, May 4). Free-agent addition Austin Corbett holds the early edge over Alec Anderson at left guard per Getzenberg's minicamp report. Ray Davis lurks for the same change-of-pace carries he's had for two years.

What They're Saying

• Getzenberg (ESPN, June 17): Cook says he's prepared for a workload similar to last year's 309 carries, no camp talk of a touch cap.
• Brasky (SI Bills On SI, June 22): the sharpest usage signal of the offseason is a possible passing-game bump for Cook under Brady, with the target count (40) having room to run back toward his 2023 high of 54.
• Brasky again (May 4): the regression case, O-line churn, no fullback, more mouths in the pass game, with a projection of roughly 280 carries and 1,380 yards.

Nova's projection already splits these takes down the middle: 1,188 rushing yards prices in the line churn and some rush-rate normalization, while 31.5 receptions concedes only a partial rebound in targets. The model isn't waiting for camp to make the regression call, it's made it.

The Buzz

The crowd is running a referendum, not a consensus. FantasyPros' Derek Brown crowed on July 6 that Cook "crushed all of the haters" as 2025's RB6 in points per game, while SI's fantasy desk framed him against De'Von Achane as "the safer investment" at the first/second-round turn (Matt Brandon and Shawn Childs, April 30). The regression crowd cites the trenches; the believers cite two straight RB6 seasons. At ADP 11.4, the believers are currently setting the price, and Nova, at 15, is quietly siding with them on rank while siding with the skeptics on scoring.

Swing Factors

• Rush-rate drift (−20 to −30 pts): every 2% of pass rate Brady adds costs Cook roughly a carry a game off 2025's fourth-ranked rushing volume.
• Target rebound (+15 to +25 pts): 54 targets instead of 40 adds a catch a game in full PPR, the cheapest path to beating Nova's median.
• Left guard resolution (±15 pts): Corbett stabilizing Edwards' spot protects the 5.2 YPC; an Anderson camp loss plus no fullback erodes it.
• Touchdown variance (±20 pts): 9.9 projected rushing TDs off 59 red-zone opportunities; his 2024 (16) and 2023 (2) rushing-TD seasons mark how wide this swings.
• Ball security (−15 pts and trust): six fumbles in 2025; a repeat invites the Ray Davis conversation nobody in Buffalo currently wants.

If Everything Goes Right / If Everything Goes Wrong

If everything goes right: Brady keeps the run-first identity he built, Corbett holds the left guard job, and the DJ Moore trade lightens boxes instead of stealing snaps, Cook catches 50 balls for the first time since 2023 while keeping 280-plus carries. That's the top of the simulation: a ceiling near 500 points, and Nova prices the tier-1 season at 37%.

If everything goes wrong: the pass rate climbs, the rebuilt interior leaks, the fumbles resurface, and a first missed stretch of his career meets a rising Ray Davis. The distribution bottoms out at 71 points, and Nova puts the full miss, a season outside Tier 4 entirely, at 40%. The honest read of the RB8–RB38 likely range is that the back half of it doesn't even require an injury, just an offense that stops leaning run-first.

Nova's Edge

The edge here isn't the rank, Nova 15 versus ADP 11.4 is agreement within rounding. The edge is the tier label. The market sees back-to-back RB6 finishes and prices continuation; Nova slots him Tier 4 with a 247-point median, an 18% haircut off his 2025 scoring, because the model weighs the O-line churn and rush-rate risk that a finish-based price ignores. Same slot, different distribution: you're buying a fair price on a player whose median owner will be mildly disappointed and whose 37% tier-1 branch bails the pick out.

The Bottom Line

Take Cook at the first/second-round turn (picks 12–18) and feel fine; taking him in the top 10 pays for the 2025 outcome, not the 2026 distribution. He's stronger in standard and half-PPR than full PPR, an 8% target share doesn't cash reception bonuses. Roster Ray Davis late as the handcuff to a 309-carry back with no proven insurance behind him. The gradeable claim: Cook finishes inside the top 12 at running back but below 280 PPR points, the 60% tier-2 branch, not the 37% repeat.