NFL RESEARCH • DRAFT VALUE & CAREER OUTCOMES
THE CORE FINDING
Draft position strongly predicts expected NFL value—but it does not determine individual NFL success.
At the aggregate level, NFL value can be modeled as expected value by draft pick plus a large error term. That error includes injury, scheme, coaching, development, opportunity, team quality, positional fit, competition, work ethic, adaptation, and randomness. The draft is highly predictive of expected value; it is much less predictive of realized individual value.
CAREER wAV
R² 0.90
SEASONS STARTED
R² 0.90
GAMES PLAYED
R² 0.82
The 2025 peer-reviewed study analyzed players drafted from 1993 through 2012, averaged outcomes by individual draft slot, and modeled the relationship between selection number and NFL career performance. Its R² values are powerful—but they describe average value by draft slot, not the exact career of an individual player.
AVERAGE wAV BY SLOT
R² = 0.90
AVERAGE STARTS BY SLOT
R² = 0.90
AVERAGE GAMES BY SLOT
R² = 0.82
Across the study population, the average draft pick produced approximately 18 career wAV, 65 games played, and 2.3 seasons started. These are population benchmarks—not universal career cutoffs.
RANGE
EXPECTED VALUE
SIGNAL
STRATEGY
Picks 1–10
Extremely high
Very high
Transformational talent
Picks 11–20
High
High
Starters and premium positions
Picks 21–32
High
High / moderate
Value and positional fit
Round 2
Moderate-high
Moderate-high
Starter hunting
Round 3
Moderate
Moderate / high variance
Traits and development
Rounds 4–5
Low-moderate
Lower
Find inefficiencies
Rounds 6–7
Low
Low / very high variance
Upside and specialization
Draft position is not equally informative at every position. The historical rates below come from the supplied draft research and use its own starter, Pro Bowl, All-Pro, or composite success definitions. They are directional benchmarks, not universal cutoffs.
QUARTERBACK
R1 63% / R2 27%
RUNNING BACK
TOP 10 54% / LATE R1 40%
WIDE RECEIVER
TOP 10 49% / R2 47%
OFFENSIVE LINE
R1 STARTER RATE 87%
DEFENSIVE BACK
TOP 10 69% / R2 63%
EDGE RUSHER
TOP 10 55% / PICKS 21–64 13%
A player’s raw career value is only half the evaluation. Draft Surplus Value compares actual production with the historical expectation of the selection: DSV = actual value − expected value. A Pick 20 player who produces 65 wAV against an expected 40 creates +25 DSV. A Pick 5 player who produces 30 wAV against an expected 50 creates −20 DSV.
PICK 20 EXAMPLE
65 wAV − 40 expected = +25 DSV
PICK 5 EXAMPLE
30 wAV − 50 expected = −20 DSV
NFL teams are buying player value minus contract cost. Draft efficiency can therefore be understood as production divided by contract cost, or surplus value created by a controlled rookie deal. This is why late-round starters are so valuable—and why a productive first-round player can still be a disappointing investment if the draft capital and contract expectation were much higher.
NFL player value is heavily right-skewed. Most players generate modest career value, while a small number create enormous careers. Late-round success is therefore driven disproportionately by outliers: the mean can look surprisingly high even when the median remains low. Teams should track mean, median, standard deviation, starter rate, Pro Bowl rate, All-Pro rate, and total value together.
PREDICTION HAS LIMITS
One combine-based study predicted whether a drafted player recorded an NFL snap with approximately 83% accuracy, but its best model explained only about 17% of the variation in later snap counts. Better inputs help—but uncertainty remains enormous.