Predraft Knowledge

Every draft season the same strategy claims circulate — wait on quarterback, hero running back, tight end is a lottery outside the top tier. We tested the ones that keep coming back against 18 years of real box scores and real historical draft position, and added a few findings of our own.

Every chart below is our own computation against real nflverse box scores (2008–2025, half-PPR) joined to real historical average draft position — nothing here is a survey or a projection.

A season only counts at 6+ games played, so a two-week fill-in doesn't get ranked as a real finish. Round = ADP ÷ 12, rounded up — the same math a real 12-team snake draft uses.

📊 ai/claude/data/adp_history_clean.csv · 38.6% of real season stat-lines matched to a real draft position

Where the picks actually pay off

What happens to a draft pick spent on a skill player, round by round.

Claim 1

The players who decide a season are almost always drafted early.

Where the top-5 PPG finishers at RB and WR were actually drafted, every year since 2008. 70.9% of top-5 running back seasons and 61.6% of top-5 receiver seasons came out of the first two rounds — a real season-ending finish, not a preseason ranking.

RBWR
Rounds 1–2 — RB: 70.9% · WR: 61.6%70.9%61.6%Rounds 1–2Rounds 3–4 — RB: 15.1% · WR: 22.1%15.1%22.1%Rounds 3–4Round 5+ — RB: 14% · WR: 16.3%14%16.3%Round 5+
Claim 2

The running back "dead zone" is real — the same rounds are safer at receiver.

Of every running back drafted in rounds 3 through 6, 51.9% finished the season outside the top 24 at the position — unplayable in most starting lineups most weeks. A receiver drafted in those same rounds missed a wider top-36 bar only 33.6% of the time. Same draft capital, a very different hit rate.

Running back
51.9%
Wide receiver
33.6%
Claim 3

There is close to no such thing as a late-round RB1.

Top-5 running back finish rate by the round the back was actually drafted in — of everyone drafted there, hit or bust, not just the winners. Across 622 real round-5-or-later running back seasons, the hit rate is 1.9%.

Round 1
32.8%
Round 2
25.9%
Round 3
8.9%
Round 4
10.7%
Round 5+
1.9%

How fast the value actually falls off

Scoring by finish rank and by draft round, pooled across 18 seasons.

Claim 4

Scoring is a cliff at the very top, then almost flat.

Average PPG by season-ending finish rank, pooled across all 18 years. There is a real, steep gap from the top of the position down to roughly the WR12/RB12 range — and then the curve goes nearly flat for the next two dozen ranks. Hover any point for the exact split.

RBWR
05101520#1#12#24#36
Claim 5

The middle rounds do not buy safety.

Median PPG by the round a player was actually drafted in, receiver and running back side by side. A real WR4 (round 4) medians 11.15 PPG; a real WR8 (round 8) medians 8.9 PPG — a gap of only 2.3 points per game across four full rounds of a real draft. Running back falls faster over the same stretch.

WRRB
R1 — WR: 15.2 · RB: 15.015.215.0R1R2 — WR: 13.6 · RB: 13.613.613.6R2R3 — WR: 12.1 · RB: 11.612.111.6R3R4 — WR: 11.2 · RB: 11.211.211.2R4R5 — WR: 10.9 · RB: 9.810.99.8R5R6 — WR: 10.2 · RB: 10.010.210.0R6R7 — WR: 9.8 · RB: 8.19.88.1R7R8 — WR: 8.9 · RB: 8.18.98.1R8R9 — WR: 8.3 · RB: 8.28.38.2R9R10 — WR: 7.0 · RB: 7.67.07.6R10R11 — WR: 8.3 · RB: 5.88.35.8R11R12 — WR: 7.2 · RB: 6.47.26.4R12
Claim 6

Rookie receivers usually need a full year before they pay off.

Top-24 finish rate for wide receivers, split by real NFL experience (nflverse roster data — draft year, not our own ADP window) rather than by draft slot. A receiver's rookie season hits top-24 only 6.3% of the time; the same player pool's second season nearly doubles that, to 11.9%.

Rookie season
6.3%
Second season
11.9%

Quarterback and tight end play by different rules

Two positions where waiting costs less than the board price suggests.

Claim 7

Waiting on quarterback costs less than the ADP gap implies.

Median full-season points by when the QB was actually drafted. A quarterback taken after pick 100 outscored the single first quarterback drafted that year in 9 of the last 18 seasons — essentially a coin flip on paying a first-round price for the position at all.

297
Drafted before pick 50
18.94 PPG
223
Drafted after pick 100
15.12 PPG

Every season 2008–2025 — green means a QB drafted after pick 100 outscored the first QB taken that year:

08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Claim 8

Tight end 7 through 18 is one tier sold at different prices.

Top-6 tight end finish rate, grouped by the order tight ends actually came off the board (TE1 = first TE drafted that year, etc.), not by overall round. TE7–11 and TE12–18 land within a few points of each other — 22.1% vs 14.7% — genuinely no separation between paying up inside that range and waiting.

TE1–6
50.5%
TE7–11
22.1%
TE12–18
14.7%
TE19+
10.5%
Claim 9

A rushing quarterback gives you a higher floor most weeks.

Real individual weeks, not season totals. Quarterbacks who ran the ball 4 or more times a game scored under 15 half-PPR points in only 38.3% of their own starts, against 56.6% for quarterbacks who rarely ran at all — a median of 17.4 vs 13.7 points a week. The rushing floor shows up start to start, not just in the season total.

56.6%
Pocket QB weeks under 15 pts
under 2 carries/gm · median 13.7 pts
38.3%
Rushing QB weeks under 15 pts
4+ carries/gm · median 17.4 pts

What no draft strategy fixes

How few players actually separate a season, and how little the league repeats itself year to year.

Claim 10

Only a handful of players separate a fantasy season.

Skill players scoring 20+ half-PPR points per game that season (min 8 games played, so a hot four-week stretch doesn't count). The count runs low single digits most years, with a real spike in 2018 and again around 2020–2021 — seasons where an unusually deep group of players separated from the pack.

2008: 0020082009: 2220092010: 3320102011: 8820112012: 6620122013: 3320132014: 7720142015: 5520152016: 8820162017: 4420172018: 131320182019: 8820192020: 141420202021: 121220212022: 4420222023: 5520232024: 7720242025: 552025
Claim 11

NFL offenses are far more volatile year to year than draft rankings assume.

Real team-level half-PPR points (every skill player combined), ranked 1–32 each season, tracked year to year. A top-8 offense is roughly a coin flip to repeat as a top-8 offense the following year — the rest either fall off some or fall hard.

50.7%
Top-8 offense repeats top-8 next year
20.6%
Top-8 offense falls to bottom half next year
30.1%
Any offense moves 10+ spots year over year
Honesty note. A few of these findings depend on a threshold we chose ourselves rather than one everyone would agree on — "bust" for a running back means outside the top 24, a "rushing quarterback" means 4+ carries a game, "elite" means 20+ PPG on at least 8 games played. Move any of those lines a little and the exact percentage moves with it, even where the direction of the finding does not. Every threshold is stated next to its number above rather than buried in a footnote, and the full query behind each one is in the repo for anyone who wants to test a different cutoff.
Full methodology and every query: predict/data/predraft_claims.R in this repo. Re-run it any time the underlying nflverse data or ADP history refreshes.