NFL Analytics Glossary

The percentage of a team's offensive plays a player is on the field for. Snap share is the floor of opportunity — a player who isn't on the field can't produce.

Special-teams snaps are excluded. Pre-snap motion still counts as a snap.

The percentage of a team's pass-route opportunities in which a receiver ran a route. Route participation isolates passing-game involvement from rushing snaps.

It needs charted routes, which are published after the season. In a season we do not hold yet, the rate is refused by name rather than estimated (see Charted after the season).

The percentage of his team's targets — passes thrown to a receiver — that went to him. Target share is the single most predictive volume metric for fantasy receivers.

The full-season share divides by every target his team threw; the while-active share counts only the weeks he played, so a missed game does not dilute it. Read it beside route participation to see how often he was out on a route.

A running back's percentage of team rushing attempts while on the field. Carry share separates true workhorses from committee backs.

Targets inside the opponent's 20-yard line. A red zone target is worth more than a target between the 20s because it ends in a touchdown far more often.

The sum of a player's targets plus carries inside the 10-yard line. HVT correlates more strongly with fantasy production than raw touch totals.

The change in a team's expected points from before to after a single play. Positive EPA means the play increased the team's scoring expectation; negative EPA means it decreased it.

EPA is the standard play-level efficiency measure. Aggregate it per play, per target or per route to compare players fairly.

Average Expected Points Added across every pass and run an offense ran. In 2025 the fifth-best offense ran +0.098 EPA per play and the fifth-worst −0.115.

The percentage of plays that produced positive EPA. Success rate captures consistency in a way that yardage totals do not.

Yards a receiver gains after making the reception. YAC isolates a receiver's ability to create yardage beyond what the throw produced.

It prints as a season total, or as an average per reception.

How many receiving yards a receiver produces for every route he runs, whether or not the ball comes his way. YPRR is the most predictive efficiency stat for receiver talent because it normalizes for opportunity.

receiving yards ÷ routes run

From 2026 the routes are estimated (see Estimated routes) and the rate carries the estimated mark; under 10 estimated routes it prints a dash.

The distance in yards between a receiver and the nearest defender at the moment a pass arrives — the catch or the incompletion. It is measured on every target by player tracking.

In 2025, 120 receivers qualified: the middle of them was 3.02 yards, the average 3.03, and the top tenth 3.75 or more.

The average distance past the line of scrimmage at which a receiver is targeted. aDOT identifies role: how far downfield his targets are thrown.

air yards ÷ targets

In 2025, among the 78 wide receivers with 50 or more targets, the middle was 11.3 yards, the deepest tenth 13.8 or more and the shortest tenth 7.6 or less; tight ends sat lower, a middle of 6.2 across 29.

The percentage of a team's total air yards directed at one receiver. AYS captures the share of downfield opportunity, not just total target volume.

The percentage of routes a receiver ran from a given pre-snap position (outside, slot, in-line, backfield). Alignment rates flag scheme changes and emerging role shifts.

Alignment is charted after the season, so the current season's rates wait for it.

The set of route concepts a receiver runs and their frequency. A diversified route tree signals scheme versatility; a concentrated tree signals specialization.

Route types are charted after the season (see Charted after the season).

Fantasy points allowed by a defense to a specific position. DvP is the most-cited matchup tool in fantasy football, but raw DvP is heavily distorted by opponent quality.

The matchup letter on this site is the rank of that allowance per game (see the fantasy points allowed grade).

Defense vs Position with opponent strength removed. Schedule-adjusted DvP isolates the defense's true difficulty for a given position.

The percentage of opposing dropbacks on which a defense generated pressure. Pressure rate is more stable week-to-week than sack rate.

Pressure on each play is charted after the season; until then a season's rate is not printed.

Average yards a runner gains before first defender contact. YBC isolates the offensive line and front-seven matchup from the runner's own talent.

The percentage of opposing plays that went for 15+ pass yards or 10+ rush yards. Explosive rate captures the volatility of a defense.

The difference between a quarterback's actual completion rate and the expected completion rate given throw difficulty. Positive CPOE means the QB is hitting throws other passers would miss.

Average seconds from snap to release on a passing attempt. A quick release points to fast reads or a quick-game plan; a long hold to slow reads or to a line giving him time.

In 2025, 42 passers qualified: the middle was 2.82 seconds, the quickest tenth 2.67 or less and the slowest tenth 3.01 or more. None averaged under 2.5.

Average air yards per pass attempt. AY/A flags whether a QB throws downfield or relies on YAC-heavy short attempts.

Net passing yards adjusted for sacks, touchdowns, and interceptions, divided by attempts plus sacks. ANY/A is the most complete single-number QB efficiency stat.

(pass yds + 20 × TD − 45 × INT − sack yds) ÷ (attempts + sacks)

The percentage of pressured dropbacks that result in a sack. A low rate signals QB pocket awareness; a high rate signals a stuck passer.

A scoring format that awards 1 fantasy point per reception. PPR boosts pass-catching backs and slot receivers relative to standard scoring.

A scoring format that awards 0.5 fantasy points per reception. Half-PPR sits between standard and full PPR.

A league format where teams can start a second quarterback in a flex spot. Superflex raises every starting quarterback's value, because two of them score in one lineup.

A drafting framework that values a player by their projected points above the replacement-level player at the same position. VBD is the foundation of every modern positional ranking.

The average pick number at which a player is drafted across many drafts. ADP is the market's price tag; our value set against it is the mispricing signal.

The average of expert analyst rankings for a player. ECR is the analyst-side counterpart to crowd-sourced ADP.

A group of players whose values sit close together, separated from the next group by a gap. Tiers compress small ranking differences into actionable groupings.

A data-quality flag indicating that a metric was computed on fewer than its floor's observations. A LOW_SAMPLE read should be treated as directional, not definitive.

Floors are derived, not picked (see Derived floors). Estimated routes need 10 before a per-route rate prints.

Estimated routes: dropbacks while on the field, adjusted by position, not charted routes. Every per-route rate from 2026 on divides by this count and carries the estimated mark.

Each game's estimate is his share of the offense's snaps times its dropbacks, tilted by position and role, at least two routes for every target and never more than the team's dropbacks. Under 10 estimated routes the count still prints, marked, and the per-route rate is a dash.

Needs 10 estimated routes.

The matchup letter: all 32 clubs sorted by the fantasy points they have allowed per game to the position, in your scoring. The club that allowed the most is an A+ (the softest matchup) and the club that allowed the fewest is an F.

Letters follow rank, ties sharing a midpoint: A+ is the top 5%, A the next 10%, B+ the next 15%, B the next 15%, C the next 20%, D the next 20%, F the bottom 15%. Before any game this season, a club is ranked on last season's allowance; with no prior season on file either, every club reads as the league average, the middle letter (C). A letter is withheld only for a club the table holds no read for. Yards per carry and every efficiency rate are never part of it.

Neutral script: win probability between 20% and 80%, with more than two minutes left in the half.

Neutral rates (neutral pass rate over expected, neutral EPA per play, neutral pace) count only these snaps, so a team protecting a big lead or chasing a deficit is not read as its play-caller's habit.

Per-play charting is published after the season: who was on the field for each play, coverage shells, man and zone coverage, personnel groupings, route types, receiver-versus-corner matchups and offensive-line on-field splits. A season we do not hold yet is refused by name with the newest held season as the door, never shown as last season's numbers under this season's heading.

What the labs print in that case: "Coverage shells are charted after the season. 2025 is the newest season we hold."

The sample a player needs before a number is ranked or painted as a percentile. It is measured, never picked: each player's events are split into two random halves at a ladder of sample sizes, and the floor is the size at which the corrected half-to-half rank reliability first reaches 0.50.

Measured over every 2018–2025 regular season. Under its floor a number still prints, marked "too noisy to rank at this sample", and is never ranked. A share of his team’s own total, so it is exact rather than an estimate and needs no sample floor.