Players I want less exposure to at their current price — a fade is not "never draft him"; there is a cheaper price where the same player becomes a buy. The model recommends players the Board ranks behind the market; the final list is mine.
WR35 and WR34 the last two years, the second on 122 targets that became just 777 yards, 6.4 per throw. He's 34, he signed in Indianapolis on August 19, was arrested on preliminary DWI charges eleven days later, and Pierce and Downs already hold the top of the target tree. WR59 pays for the name. My board pays for the last two seasons: WR72.
A thousand-yard season on 84 targets is a magic trick: twelve yards per target, the kind of efficiency that doesn't repeat, on the kind of target volume that does. The market is drafting the yards at WR25. My board prices the targets, and they say WR35. I'm not out on the player — I'm out at this price.
WR50, WR15, WR53 over the last three years, and the WR56 price pays for the middle one. Last season he caught 50 of 106 targets for 602 yards in a Cleveland offense that fed nobody, and the Browns answered in April by drafting two receivers, picks 24 and 39, who already sit right behind him on the depth chart. My board says the 2024 season is the outlier: WR64.
This one costs me something, because he's finished WR13 and WR11 back to back. But a round-5 price pays him AT that ceiling in his age-31 season, while the underlying volume drifts the wrong way: 135 targets down to 124, 1,081 yards down to 1,017. When the best realistic outcome is the price you paid, there's no upside left in the pick — I'd rather be a round late on a 31-year-old than a round early.
The special part went first: 727 receiving yards on 99 targets last year — barely ten a catch for a player whose whole game was yards after it — and the rushing package is down to 17 carries. Now he's 30, back in San Francisco on a one-year deal, and the market still wants a round-8 pick for the name. My board has him fifteen receiver spots later, and I'm not paying for the memory.
These are my picks. Many started as the model's calls — its backtest graded about 59% against a coin-flip floor of about 51% across the 2018–2025 drafts — and some are mine alone. All of them are mine to answer for: every pick is logged the day it publishes and graded when the season ends.
THE 2026 FILM ROOM · MY PICKS, MODEL-ASSISTED · EVERY PICK LOGGED ON PUBLISH AND GRADED AT SEASON’S END
Two players can share an ADP and be nothing alike. I measured how far a draft slot actually moves, by position and by round, across eight seasons.
Jon Stone · AUG 18, 2026I priced every tight end pick against the guy you could have had for free. Only one range of the draft came back worth the pick.
Jon Stone · AUG 18, 2026Everyone knows a 400-touch season breaks a running back. I measured it against seventeen years of data, and the thing everyone knows isn't there. Something else is.
Jon Stone · AUG 18, 2026