Advanced Stats for Start/Sit Decisions: Air Yards, aDOT, and More

Air yards, average depth of target, target share, racially adjusted stats — the vocabulary of advanced football analytics has expanded fast enough that even experienced fantasy managers sometimes nod along while quietly wondering what any of it actually means for Sunday's lineup. This page breaks down the most decision-relevant advanced metrics: what they measure, why they're predictive, where they mislead, and how to weight them against each other when the deadline is closing in.

• Definition and Scope
• Core Mechanics or Structure
• Causal Relationships or Drivers
• Classification Boundaries
• Tradeoffs and Tensions
• Common Misconceptions
• Checklist or Steps
• Reference Table or Matrix

Definition and Scope

Air yards measures the total distance a pass travels through the air from the line of scrimmage to the point of the catch — or the point of incompletion. It doesn't care whether the ball was caught or dropped. A 25-yard target that falls incomplete still generates 25 air yards for the intended receiver.

aDOT (average depth of target) is exactly what it sounds like: air yards divided by total targets for a given receiver over a defined sample. A wide receiver with a 15.2 aDOT is consistently being targeted deep. A running back with a 1.8 aDOT is working the flat almost exclusively.

Racially adjusted or scheme-adjusted metrics are less common but growing in use. More immediately relevant to most fantasy managers are target share (a receiver's targets as a percentage of total team pass attempts), air yard share (a receiver's air yards as a percentage of total team air yards), WOPR (weighted opportunity rating, a composite of target share and air yard share developed by Josh Hermsmeyer and published through the open-source FBref database), and ADOT-adjusted expected points added (EPA).

These aren't esoteric toys for quants. Air yards and aDOT appeared in mainstream fantasy analysis after the publication of the Sports Info Solutions Football Analytics Manual and the public release of Next Gen Stats by the NFL, which gives every interested party access to route participation, target distance, and separation metrics at the player level.

The scope of this page covers NFL receiver and skill-position metrics that are publicly available and directly applicable to start/sit decision-making.

Core Mechanics or Structure

How air yards are compiled: Every pass play generates a target distance logged from the line of scrimmage. NFL Next Gen Stats and the open-data platform nflfastR (an R and Python package maintained by Sebastian Carl and Ben Baldwin) record this on a play-by-play basis. nflfastR's play-by-play dataset, which is publicly accessible on GitHub, includes air_yards as a column on every pass attempt.

WOPR construction: Josh Hermsmeyer's WOPR formula weights target share at 1.5 and air yard share at 0.7, then divides by a normalizing factor to produce a 0-to-1 scale. A WOPR above 0.50 is considered strong for a receiver; league leaders in a given week typically post WOPR values in the 0.55–0.70 range. The formula weights target share more heavily because raw target volume is a stronger predictor of short-term fantasy output than air yard concentration alone.

Expected fantasy points (xFP) from air yards: Several models, including those described in the Rotoviz Air Yards analysis framework, estimate a receiver's expected fantasy production by taking their share of team air yards and translating it against expected pass attempts and completion probability by depth zone. A receiver capturing 35% of their team's air yards when the team throws 38 times generates a materially different opportunity than the same share on a team throwing 22 times.

Snap count and route participation serve as the denominator check. A high air yard share from 40% route participation is a red flag; the same air yard share from 90% route participation signals a legitimate usage role.

Causal Relationships or Drivers

Air yards and aDOT don't just describe what happened — they point toward why fantasy production is likely to persist or regress.

Touchdowns and air yards: End zone targets require vertical distance. Receivers with higher aDOT values are, almost by definition, getting thrown to in areas where scores are possible. Research published by the Pro Football Reference / Sports Reference database team has consistently shown that red zone target share and deep target frequency are among the strongest predictors of receiver touchdown rate.

Yards after catch (YAC) dependency: A receiver with a low aDOT who posts strong fantasy numbers is leaning heavily on YAC — broken tackles, open field running after short catches. YAC is less stable week-to-week than air yards, because it depends on defensive alignment, traffic patterns, and individual game flow in ways that scheme doesn't directly control.

Quarterback influence: aDOT is partly a quarterback trait. Some QBs (Patrick Mahomes, Josh Allen) consistently attempt passes at 9+ yards average depth; others (Tua Tagovailoa running the Air Raid, particularly in 2022–2023) operate high-volume, short-distribution offenses where individual receiver aDOT values compress. Interpreting aDOT without QB context produces distorted conclusions.

Team pace and pass volume: A receiver with a 28% target share on a team that averages 28 pass attempts per game has roughly 7.8 targets per game. The same share on a team averaging 40 pass attempts produces 11.2 targets. Pass volume — traceable through Vegas implied team totals, covered in depth at Vegas Lines and Game Totals — multiplies the impact of every usage metric.

Classification Boundaries

Not all advanced metrics belong in the same decision bucket.

Volume metrics (target share, air yard share, snap count, route participation) predict opportunity. They say nothing about efficiency.

Efficiency metrics (yards per route run, EPA per target, passer rating when targeted) reflect execution quality. They can drift based on coverage scheme, defensive personnel, and weather.

Predictive stability differs sharply by metric type. According to analysis in the nflfastR documentation, play-by-play EPA stabilizes as a reliable player measure only after roughly 50 targets for receivers — meaning single-week EPA numbers are nearly noise. Target share begins stabilizing meaningfully after 4–6 games. Air yard share stabilizes faster than raw air yards because it's already normalized to team context.

aDOT classification by range:
- Under 5 yards: Primarily check-down and screen targets; most common for running backs and slot receivers in short-area schemes
- 5–10 yards: Intermediate zone; typical for possession receivers and tight ends
- 10–15 yards: Vertical intermediate; wide receivers with size or route variety
- Above 15 yards: Deep threat classification; touchdown upside, high variance

Tradeoffs and Tensions

The most useful tension in advanced stats is between opportunity and efficiency — and it shows up most clearly in PPR formats. A receiver with 12 targets (elite opportunity) at 5.1 aDOT (minimal air yard upside) may outscore a receiver with 6 targets at 16.2 aDOT in standard PPR simply because he catches 9 of them. Volume wins PPR. Air yards win in big-play formats.

For more on how scoring format alters the value of these metrics, the PPR vs Standard Scoring Impact breakdown addresses the positional consequences directly.

Regression traps: A receiver who posts two massive games on 40+ air yards each and suddenly has a high-aDOT reputation may simply have benefited from specific game scripts — a team playing catch-up, forcing vertical throws. Checking rolling 4-week air yards against full-season aDOT exposes these flukes.

Scheme rigidity vs. injury-driven opportunity: When a receiver's usage spike comes from a teammate's injury rather than a deliberate scheme change, the target share is real but fragile. The Injury Report and Start/Sit page covers the specific dynamics of usage changes driven by roster attrition.

The false precision problem: WOPR and similar composite scores compress multiple variables into a single number that feels authoritative. A WOPR of 0.54 vs. 0.51 is not a meaningful difference — it sits inside the normal week-to-week variation from sample size alone.

Common Misconceptions

"High air yards means high fantasy points." Air yards are opportunity, not production. A receiver who runs 10 deep routes and is targeted once has extremely high air yards per target but low volume. Calvin Ridley's 2020 season is a named example frequently cited in analytics circles: he led the NFL in receiving touchdowns with 90 receiving yards per game, posting a top-5 aDOT, while teammate Julio Jones had more raw yards. Both metrics were necessary context; neither alone was sufficient.

"aDOT doesn't apply to tight ends." Tight ends show consistent aDOT variation — Travis Kelce has historically operated between 8 and 11 yards aDOT, while George Kittle frequently clusters in the 7–9 range, reflecting their respective roles as seam-route threats vs. after-the-catch weapons. The full tight end application is explored in TE Start/Sit Strategy.

"Target share is a season-long number." Target share shifts week to week with game script, defensive coverage deployment, and injury. A receiver who held 28% target share across a season's first eight weeks may be getting 18% in weeks when the team runs a ground-heavy script against a weak front seven.

"Air yards from the nflfastR dataset and Next Gen Stats are identical." They are not. Next Gen Stats measures from the point of the snap; nflfastR measures from the line of scrimmage. The difference is typically 0–2 yards but matters when comparing across sources.

Checklist or Steps

Signal-evaluation sequence for receiver start/sit using advanced stats:

1. Layer Vegas implied team total to scale opportunity — a high WOPR receiver on a team with an 18.5 implied total has less volume context than the same receiver with a 27.5 implied total (current game totals tracked at Vegas Lines and Game Totals)

Reference Table or Matrix

Advanced Stat Quick Reference for Start/Sit Decisions

Stabilization points based on nflfastR documentation and the reproducible analysis framework described in the Sports Info Solutions Football Analytics Manual.