Every August we treat the draft board like a set of instructions. Player 41 is better than Player 42, so we take Player 41, and we feel briefly competent about it.
The trouble is that somewhere in the middle rounds the board quietly stops being an order and starts being a guess, and almost nobody notices the moment it happens. So we measured it. Auto Draft Years joined FantasyPros PPR consensus ADP from 2016 to 2025 to end-of-season scoring across 1,602 drafted player-seasons inside the top 180.
The board is not useless. It is just far less precise than those confident little numbers make it look — and the precision runs out earlier than most drafts assume.
How precise is a ranking, really?
The honest way to score a draft board is to ask how many spots a pick missed its actual finish at its own position. A player drafted as the RB3 who finishes RB8 missed by five. That is a number a drafter can feel.
Inside the first 12 overall picks the median miss is 5 spots. That is tight. Take a back third off the board and he lands, on a typical outcome, roughly where you expected. By picks 37-48 the miss has grown to 10. By the end of the draft it is 16, which means the player you drafted as your WR40 finishes closer to WR24 or WR56 than to WR40.
The obvious objection, and why it fails
Someone always raises the same objection, and it is a good one: deeper positions are simply bigger, so a 16-place miss at WR40 is proportionally the same as a five-place miss at WR3. If that were true, the whole finding would be an accident of counting.
It is not true. Normalised for how deep each position is actually drafted, the error still climbs from 3.1% of the positional field early to 12.5% late. The board gets blurrier in real terms, not just in absolute places.
| Draft cost | Picks | Median miss | % of field | Top-12 at position | Median PPG |
|---|---|---|---|---|---|
| Picks 1-12 | 120 | 5 spots | 3.1% | 61.7% | 17.8 |
| Picks 13-24 | 120 | 6 spots | 4.5% | 55% | 16.1 |
| Picks 25-36 | 120 | 7 spots | 4.7% | 39.2% | 14.6 |
| Picks 37-48 | 120 | 10 spots | 6% | 29.2% | 13.8 |
| Picks 49-72 | 240 | 11 spots | 7.8% | 27.5% | 12.2 |
| Picks 73-96 | 235 | 13 spots | 9.2% | 18.7% | 11 |
| Picks 97-120 | 203 | 13 spots | 9.3% | 21.2% | 10.2 |
| Picks 121-150 | 224 | 14 spots | 11.1% | 12.9% | 9.5 |
| Picks 151-180 | 220 | 16 spots | 12.5% | 11.4% | 8.3 |
Severe-bust and hit rates use the same definitions as the 2026 Guide to the Markets: top-12 is a final positional finish on PPR total points.
One wrinkle worth staring at
The decline is not perfectly smooth. Picks 97-120 actually produced a higher top-12 rate (21.2%) than picks 73-96 (18.7%), which is the kind of detail a tidier chart would have quietly smoothed away.
That bump is not a strategy. With a couple of hundred picks per band it is well inside the noise, and it is exactly the point: once you are past the early rounds, adjacent bands are close enough that the ordering between them stops carrying information. The trend is real. The step-by-step ranking inside it is not.
What this means for your draft
Early, the board is close to an order. Late, it is a range wearing the costume of an order. The practical line sits around pick 36.
Before pick 36: small ADP gaps carry real information. Reaching past a player usually costs you something measurable, so let the board break your ties.
After pick 36: a five- or ten-spot difference is inside the noise. Draft the player whose case you can state in one sentence — a target share, a vacated role, goal-line work — because the board has stopped making the case for you.
This is also why refreshing consensus rankings every day in late August is mostly wasted effort for middle-round picks. You are chasing movement smaller than the error bar. Compare a current board like FantasyPros PPR ADP against last year's finishes on Pro Football Reference and the same shape shows up every season: orderly at the top, a coin flip with a slight lean in the middle.
Where I would be careful
Two caveats. Median error is not maximum error — early picks still miss spectacularly, and the next piece in this series is about exactly that. And this is ten years of a changing league: passing rates, committee backfields and rookie usage have all moved, so the precise numbers will drift.
The shape almost certainly will not. Draft boards have always gotten blurrier the deeper you go, and there is no particular reason to expect 2026 to break the streak.
The bottom line
Trust the top of your board. Interrogate the middle. Once you pass roughly pick 36, the ranking is describing a neighbourhood rather than an address — so let your own read break the tie instead of deferring to a number that is quietly guessing.
Sources: FantasyPros — PPR Overall ADP; Pro Football Reference — fantasy finishes; Auto Draft Years — 2026 Guide to the Fantasy Football Markets. Historical sample: 1,602 drafted player-seasons inside the top 180, 2016–2025. For informational and entertainment purposes only.