2026 ADP Values: Players Going Too Late in Drafts

What counts as a steal

Every value on this page is one number: the gap between where the market drafts a player and where our model ranks him. Say the market prices a player at pick 88 and our board has him 18th — that's a +70 gap. We flag anyone going 12 or more spots late inside the top 120. Beyond pick 120, ADP gets too noisy to price gaps honestly. Those two cutoffs are v1 judgment calls, not backtested constants — if we retune them, this page will say so.

The rank side of that math matters. Our board orders players by value over a positional baseline, not raw projected points — a quarterback has to clear the position's replacement level before he outranks a wide receiver. So when a player shows up here, scarcity is already in the price. The gap is pure disagreement between our projection and the room. How the projection gets made, including the season it lost, is on the methodology page.

Here is the current list, recomputed against fresh ADP daily:

The quarterback discount

Run this page in July 2026 and one pattern jumps out: quarterbacks stack the top of the gap list. That isn't a bug in the table. The market waits on the position because streaming and late-round quarterbacks have worked for a decade — but our rankings already charge quarterbacks for that replaceability, and the gaps survive anyway.

Bo Nix is the current poster case: a top-40 player on our board, drafted like a bench stash.

Notice the range before you draft the narrative. Quarterback projections run wide — rushing volume swings outcomes — and even bands this wide caught fewer outcomes than designed in backtesting, so if anything the true spread is wider still. The coverage number is on the methodology page. The play isn't "Nix is a lock." The play is: when the median outcome sits several rounds above the price, you take the discount and let someone else pay retail at the position.

One format caveat before you queue anyone here: in 2QB and superflex rooms the quarterback discount disappears. The market prices the position early there, and this section of the list goes quiet.

The widest gap gets the hardest look

When our model and the market disagree by five-plus rounds, someone is wrong — and it isn't automatically the market. Wan'Dale Robinson is the test case this summer: our rank says top-20 pick; the market, as of this run, says round eight.

Two honest facts belong next to that card. First, our model builds from usage shares, and it is re-basing Robinson's target volume onto a new team — that projection carries real role uncertainty, and the wide band says so. Second, wide receiver is our weakest position: the backtest shows our WR ranks finishing 0.006 Spearman behind a prior-season baseline, the one position we lost. We publish that, and it changes how you should play this pick.

So size the bet to the spread. A gap this wide on a wide receiver is a discounted lottery ticket, not a lock — the band shows the bust risk you're buying. Pay a mid-round price for the swing, never an early-round one. If the range narrows as camp news lands, the table will show the market closing the gap.

How to use this list without ruining it

A steal you reach for stops being a steal. The gap prices are relative to where the room drafts — take Nix four rounds early and you've spent the discount you came for.

Three rules. First, check the table the morning of your draft: ADP moves daily through August, and the gaps that survive to draft day are the real ones. Second, queue every player on this list one round before his ADP, not at our rank — if he falls to the queue spot, the pick makes itself. Third, when your league mates read the same page and a value evaporates, take the board's next-best price on the full rankings instead; the CSV export sorts by the same gap math.