Glossary

A rank sits beside the number it ranks, as of the same week. Teams that tie share a rank, and the next one is skipped (1, 2, 2, 4).

Three kinds of rank run another way: strength of schedule (#1 is the hardest schedule), a defense’s rank shown beside a player (#1 is the toughest defense for him to face), and a team’s tendencies (#1 does it the most).

Shown on

A game page · The Teams board · A team page

See also

Rank colorsStrength of schedule (SOS)Opponent defense rankPass rate

A team’s own rank is green near #1 and red near #32, in even steps. A defense’s rank beside a player is read from his side, so the colors flip: the softest defense (#32) is green and the toughest (#1) red — on the Props and Players boards, the Matchup ranks and the player page.

A rank that isn’t a judgment of quality — schedule strength, a tendency — stays gray.

The colored bars are something else: an edge, where warm means the offense is favored and cool the defense (see Matchup edge).

Shown on

The Teams board · The Players board · A player page · The Props board

See also

League rankMatchup edge

So a finished game still shows what the numbers said beforehand. What the game itself produced is labeled as the game’s: the result, the box score, the “This game” column and bar.

DVOA and pressure come from weekly imports and can trail a week; the page’s “Entering numbers” line says so when they do.

Shown on

A game page

See also

DVOAPressure rate

Shown on

The Teams board

See also

EPA per playPressure rate

A rate over several games is the totals divided — all the yards over all the attempts — not an average of each game’s rate, so a busy game weighs more than a quiet one.

L3 is a player’s last three games played: a week he missed is skipped, not counted as a zero.

Shown on

A game page · A team page · The Players board · A player page

A player’s prop numbers divide only by his games that count for that prop, so they can differ from the Players board’s per-game numbers.

Shown on

A game page · The Players board

See also

Games that count

It compares the season-to-date number with where it stood three weeks back (never before week 1), so it shows a team trending — not its last three games alone. A green ▲ is places gained, a red ▼ places lost.

Shown on

The Teams board

Every situation — down, distance, field position, time left — carries a number of points a team can expect to score from it. A play’s EPA is that number after the play minus before it: a 12-yard gain on 3rd and 10 is worth well over a point; a sack on the same down costs about one.

Counted on every pass and run, sacks and scrambles included; kicks, kneel-downs, spikes and two-point tries are left out. A pass or run wiped out by a penalty still counts, carrying the penalty’s result. Regular season only.

Pass EPA is per dropback (sacks and scrambles are dropbacks); rush EPA is per designed run. A team’s overall number pools both, so it leans toward whichever the team does more.

the EPA of its plays added up ÷ the number of plays

Which way

Higher is better for an offense. For a defense it is what it allows, so lower is better.

From

Play-by-play

Shown on

A game page · The Teams board · A team page

See also

Success rateMatchup edgeNet EPA

Counted over the same plays as EPA per play. It isn’t the yards-needed version some sites use (a share of the distance to gain on each down), so it won’t match theirs.

plays with EPA above zero ÷ plays

Which way

Higher is better for an offense, lower for a defense.

From

Play-by-play

Shown on

A game page · The Teams board · A team page

See also

EPA per play

A game has four: each team’s passing and running against the other’s defense. The edge is how far the offense is above an average offense, plus how far what the defense allows is above what an average defense allows — a strong defense allows less than average, so it pulls the edge down.

The averages are the 32 teams’ means as of the same week as the two teams’ numbers: the averages of the pass and rush EPA columns on the Teams board’s Offense and Defense views.

An offense at +0.17 per dropback, with the league’s offenses averaging +0.07, is 0.10 above average. A defense allowing +0.04, with defenses averaging +0.07, allows 0.03 less than average. Together: an edge of +0.07, the offense’s way.

A game page spells each edge out under its bar: “League average −0.11 a rush. CHI’s offense is 0.03 better; NE’s defense allows 0.04 more.” The two pieces add up to the edge as printed — each is rounded so they do, which can move one by 0.01.

Measured from an average matchup, passing and rushing edges read on one scale. (Until October 2026 the edge was the offense’s number plus the defense’s, which put every passing edge near +0.15 and every rushing edge near −0.22 before either team was counted.)

The bar is full at ±0.30 and stays gray inside ±0.08, where the difference is small. Warm favors the offense, cool the defense.

(offense EPA per play − league average) + (defense EPA per play allowed − league average), in the same phase

Which way

Neither — positive favors the offense, negative the defense.

From

Play-by-play

Shown on

The slate · A game page

See also

EPA per playBiggest edgeExpected

It names the side the edge favors: “SEA pass +0.20” is Seattle’s passing offense, “SEA rush D −0.33” is Seattle’s run defense — a negative edge belongs to the defense.

Edges are measured from an average matchup, so the biggest is the most lopsided of the four. It is more often a passing matchup: teams differ more in how they pass and defend the pass than in how they run.

When none of the four reaches 0.05 either way, the slate card and the game page read “Even matchup” instead.

Shown on

The slate · A game page

See also

Matchup edgeWho has the edge

Its bar sits under the game’s own EPA per play (“This game”), on the same scale, so the two read against each other: above it, the offense did better than the matchup said; below, worse.

The average and the edge are as of the week before the game, like every entering number. One game is a few dozen plays a phase, so a gap either way says little on its own.

league average offense EPA per play + matchup edge, in the same phase

Which way

Neither — it is what the game is read against.

From

Play-by-play

Shown on

A game page

See also

Matchup edgeEntering, and this gameEPA per play

Pass and Rush compare the two teams’ matchup edges in that phase — each team’s offense against the other’s defense — and name the team whose edge is larger. Within 0.05 EPA per play of each other they are even, the same threshold as an even matchup.

Total DVOA names the team rated higher, and calls it even within 3 points.

It compares numbers already on the card. It isn’t a prediction of who wins.

From

Play-by-play · FTN

Shown on

The slate

See also

Matchup edgeDVOABiggest edge

Halving keeps it on the same scale as the offense and defense numbers.

(offense EPA per play − defense EPA per play allowed) ÷ 2

Which way

Higher is better.

From

Play-by-play

Shown on

A game page · The Teams board · A team page

See also

EPA per play

A team’s passing yards don’t take off the yards lost to sacks, so they add up to its passers’ yards. A drive’s yards in the drive log do take them off.

From

Play-by-play

Shown on

A game page · The Teams board

DVOA (Defense-adjusted Value Over Average) is FTN’s own measure, imported each week rather than worked out here. It compares every play with the league average in the same situation and adjusts for the strength of the opponent; 0% is average.

Weighted DVOA counts recent games more than early ones.

On the slate card, “Total DVOA” is each team’s, entering the week, with the team rated higher named beside it.

Which way

Higher is better for total and offense. For defense, lower — below zero — is better.

From

FTN

Shown on

The slate · A game page · The Teams board · A team page

See also

WESEntering, and this game

Net WES is 40% the DVOA rank, 35% the net EPA rank, and 25% split evenly across four success-rate ranks (passing and rushing, offense and defense). Offensive and defensive WES use that side’s DVOA, EPA and two success-rate ranks with the same 40 / 35 / 25 weights.

The result is ranked again, 1 to 32. It needs DVOA, so it is missing until the season’s first DVOA import. It is the recipe from the group’s spreadsheet, fed with this site’s EPA and success rates.

0.40 × DVOA rank + 0.35 × EPA rank + 0.25 × the average success-rate rank

Which way

Lower is better.

From

FTN · Play-by-play

Shown on

A game page · The Teams board · A team page

See also

DVOAEPA per playSuccess rate

Before a game, the gap between the two ratings gives each team’s chance of winning, with 50 points added to the home team (none at a neutral site). Afterward each team moves 20 points times its result minus that chance — a win is 1, a tie ½ — scaled up for a bigger margin, and less so when a favorite wins big, so strong teams don’t run away with it.

Every team started at 1500 in 2021, and ratings have carried forward since, moving a third of the way back toward 1500 before each season. It’s FiveThirtyEight’s NFL method. Unlike EPA, it counts the playoffs.

Which way

Higher is better.

From

Scores and schedule

Shown on

A game page · The Teams board · A team page

See also

Strength of schedule (SOS)

It’s the average ELO of the opponents a team has played, each at its rating going into that game. Before week 1 it is projected from the whole season’s schedule instead.

Which way

Neither — a description. #1 is the hardest schedule.

From

Scores and schedule

Shown on

A game page · The Teams board · A team page

See also

ELOLeague rank

Pressure is a charter’s judgment, not something the league records, so it comes from FTN’s charting, imported one week at a time. Over several games it is all the pressures over all the dropbacks.

A game FTN hasn’t charted yet — Monday night’s, sometimes, for a day or two — reads “not charted yet” and is left out of the rate; its sacks and hits still count. Another charter’s pressure rate, such as Pro Football Reference’s, runs several points lower and won’t match.

pressures ÷ dropbacks

Which way

Lower is better for an offense, higher for a defense.

From

FTN

Shown on

A game page · The Teams board · A team page

See also

Sack rateQB hit rate

FTN’s sacks over FTN’s dropbacks, so it sits on the same base as the pressure rate beside it. On the Teams board’s Pressure view, “Sacks” and “Sacked” are these rates, not counts.

sacks ÷ dropbacks

Which way

Lower is better for an offense, higher for a defense.

From

FTN

Shown on

A game page · The Teams board · A team page

QB hits ÷ dropbacks

Which way

Lower is better for an offense, higher for a defense.

From

FTN

Shown on

The Teams board

These come from FTN’s own EPA model, not the EPA on the rest of the site, so don’t set them beside it or an edge. They are team numbers, blending all of a team’s quarterbacks.

Which way

Higher is better.

From

FTN

Shown on

The Teams board

See also

EPA per play

dropbacks ÷ plays (sacks and scrambles are dropbacks)

From

Play-by-play

Shown on

A team page

Each play carries a model’s chance of a pass given the down, distance, field position, score and time left (nflverse’s). PROE is the average of “passed or not” minus that chance: +5 means passing five points more often than a typical team in the same spots.

From

Play-by-play

Shown on

A team page

See also

Pass rate

air yards ÷ pass attempts with air yards recorded

From

Play-by-play

Shown on

A team page

deep attempts ÷ attempts with a recorded length

From

Play-by-play

Shown on

A team page

From

Play-by-play

Shown on

A team page

scrambles ÷ dropbacks

From

Play-by-play

Shown on

A team page

Shares of the plays where the play-by-play records a direction; each set adds up to 100%. Targets by position split the team’s targets among wide receivers, tight ends and running backs (fullbacks count with the backs).

From

Play-by-play

Shown on

A team page

On the Players board Δ is shown bright when it is 10 yards or more either way. On the leader tables an arrow beside L3 marks a last-three average at least 15% above (▲) or below (▼) his season average.

last-three yards per game − season yards per game

From

Play-by-play

Shown on

A team page · The Players board

See also

Season, Last 3, Last 5, Home, Away

Two-point tries are left out on both sides. Over several games it is the totals divided, not an average of each game’s share.

his targets ÷ his team’s targets, in the games he played

From

Play-by-play

Shown on

A game page · The Players board · A player page

Air yards are how far a pass travels past the line of scrimmage, caught or not. The team’s total counts every pass with air yards recorded, including ones not thrown to anyone in particular.

his air yards ÷ his team’s air yards, in the games he played

From

Play-by-play

Shown on

The Players board · A player page

his carries ÷ his team’s carries, in the games he played (two-point tries left out)

From

Play-by-play

Shown on

A game page · The Players board · A player page

On the Players board that is every game he played; on the Props board, only his games that count for the prop.

From

Play-by-play

Shown on

The Players board · The Props board

See also

Games that count

On the Players board and the leader tables, G is every game he played this season. On the Props board it is only the games where he had a real role in that stat.

Shown on

The Players board

See also

Games that count

The 32 defenses ranked by EPA allowed — per dropback for pass D, per run for rush D. Beside an upcoming game it is the defense as it stands now; in a game log, as it stood going into that game.

It is colored from the player’s side: the softest defense green, the toughest red — the reverse of a team’s own rank, because a top defense is bad news for the player.

From

Play-by-play

Shown on

The Players board · A player page · The Props board

See also

EPA per playLeague rankRank colors

The favorite’s number is negative, the underdog’s positive; PK or Pick’em means neither is favored. Before kickoff it is the current line; once the game is final, the closing line and whether it was covered.

The lines come from nflverse’s schedule file, which doesn’t name the sportsbook, so one can sit half a point off a particular book’s.

From

Game lines

Shown on

The slate · A game page · A team page

See also

Record against the spread (ATS)

From

Game lines

Shown on

The slate · A game page · A team page

See also

Over/under record

From

Game lines

Shown on

A game page

See also

Prices (American odds)

A team covers when its winning margin plus its own spread is above zero; exactly zero is a push, shown only when there is one. A game without a line isn’t counted. On a game page it is the record going into that game.

From

Game lines · Scores and schedule

Shown on

A game page · A team page

See also

Spread

It describes the games, so both teams in a game get the same result.

From

Game lines · Scores and schedule

Shown on

A game page · A team page

See also

Total

Rest counts calendar days back to the team’s last game on the schedule; more than 10 means a bye sat in between.

From

Scores and schedule

Shown on

A game page

FanDuel’s lines, captured during the week and again shortly before kickoff. “Line” is the newest one captured. When it has moved since it was first seen, the Props board puts the first line and an arrow for the way it went before it — “55.5 ▲ 56.5” — and under it on a phone; a player’s page says “was 55.5” beside it.

A price shows with a line only where the book has priced the two sides differently — “U -128”, the side it favors — and then the line is not a coin flip.

From

Prop lines

Shown on

A game page · A player page · The Props board

See also

Prices (American odds)Closing line

A prop’s price shows only where the book has priced the two sides differently: the side it favors and its price, “U -128” beside the line on the Props board and a player’s page (both prices in its hover text, and on a game page in the line’s). A line priced alike on both sides — −114 each way, as nearly every line is — says nothing about the player, so it shows no price.

A shaded line is not a coin flip: at −128 the under has to win 56.1% of the time to break even, so his median against that line reads differently. Shaded lines are mostly the small ones — a back’s 2.5 receiving yards, a quarterback’s 0.5 rushing.

From

Prop lines

Shown on

A player page · The Props board

See also

Break-evenProp line

−110 needs 52.4%; +120 needs 45.5%. The two sides of a line add up to more than 100%: the extra is the sportsbook’s margin.

for −110: 110 ÷ (110 + 100); for +120: 100 ÷ (120 + 100)

Shown on

The Props board

See also

Prices (American odds)

His median, averages, high, low and “went over” count use only these, so a relief appearance doesn’t drag them down. That’s why they can differ from the Players board.

Shown on

The Props board

See also

Median and its gap to the lineAvg and L3Went over

A line is set near the point where the over and the under are about equally likely, which is a median, so his median is the like-for-like number to set beside it. Yardage is lopsided: one big game pulls an average above most of his games, but not a median. A receiver with 40, 45, 50, 55 and 150 yards averages 68 — well over a 52.5 line — yet went over it twice in five games; his median is 50.

With an even number of games it is halfway between the middle two, and on a handful of games it can move a lot from one week to the next.

These are his numbers, not a forecast: the line is set with all of them in hand, and a wide gap is not a stronger signal than a narrow one.

Med ± = median of his games that count − line

From

Play-by-play · Prop lines

Shown on

The Props board

See also

Games that countAvg and L3Went over

One big game lifts an average, so on the Props board the median leads and the averages sit with the rest of his season; L3 is dimmed when he has three or fewer games, since it is then the same games as Avg.

From

Play-by-play

Shown on

The Props board

See also

Games that countMedian and its gap to the line

Today’s number tried against every game that counts, not the line each of those games actually closed at — that is the closing hit rate. A game exactly on the line is a push, counted apart.

It describes his season against today’s number; it is not a forecast — the line is set with those same games in hand.

From

Play-by-play · Prop lines

Shown on

A player page · The Props board

See also

Closing hit ratePush

Captured 10 to 40 minutes before each game. A prop taken off the board before the final hour — a player ruled out, say — has no closing line. The history starts with week 5 of 2026, the first week captured.

From

Prop lines

Shown on

A player page

See also

Closing hit rateHit margin

Only games with a captured closing line count, and pushes are left out of both sides. On the Props board it reads “3 of 4”, or “3/4” on a phone.

A game he played without a target, a carry or a pass counts as 0 yards — an under, the way the book settles it — when Pro Football Reference’s snap counts show him on the field for an offensive snap. His game log has no row for that game; in the props history, the 0 says why when you hover over it. A game he sat out, or played on special teams alone, is not graded.

From

Play-by-play · Prop lines

Shown on

A player page · The Props board

See also

Closing lineWent over

On the Props board and in the props history summary it is the average over his games with a closing line — whether his overs were squeakers or blowouts.

actual yards − closing line

Shown on

A player page · The Props board

See also

Closing line

Shown on

A player page

The second rank is FTN’s DVOA against passes to that group of receivers, season to date and imported each week — so a defense that is solid against the pass overall but weak against tight ends shows it. Hovering on it gives the DVOA and the passes and yards it allows that group a game.

A tight end reads against tight ends and a running back against running backs. FTN doesn’t say which receivers it counts as a team’s WR1 and WR2, so the app decides: his team’s wide receivers with a line this week, ranked by targets per game this season — the first is WR1, the second WR2, the rest WR3 (FTN’s group for every receiver past the top two). A receiver out that week has no line and takes no place. On a phone the group rank is under his name.

The first rank is the app’s own pass defense rank, from EPA; the second is FTN’s, a different model — the two can disagree, which is the point of showing both. Both are colored from his side: green a soft defense, red a tough one.

From

FTN · Play-by-play

Shown on

The Props board

See also

Opponent defense rankDVOA

The feed is read every hour from morning to evening, Eastern time — so the teams’ afternoon reports are in within about an hour — and again in the hour before each kickoff, after the inactives are announced. A player is filed under his team’s next game, so a report is always about one game.

Closed, the card is each team’s count by status; open, it lists the players, each with his injury and how much he plays, and a tap gives ESPN’s latest news on him and when it was written.

Once a game has kicked off its report stops changing: what it shows is what was known at kickoff. A player cleared before then played, and isn’t listed.

From

Injury reports

Shown on

A game page

See also

Injury statusSnap share, and depth players

Teams report injuries on practice days and give game statuses on the last report — usually Friday for a Sunday game, Saturday for a Monday game and Wednesday for a Thursday game. Before then ESPN’s label is a placeholder (early in the week it lists most injured players as questionable), so the card says “no game status yet” and gives the injury and the news instead.

Out counts whenever it is given — a team can rule a player out early, never in. So do IR (injured reserve: out for at least four games), PUP (physically unable to perform), Suspended, and Inactive — left off the game-day roster, announced 90 minutes before kickoff.

“IR, may return” and “PUP, may return”: designated to come back from injured reserve or PUP. ESPN keeps him on the list even when his team lists him questionable, and only an activation puts him in the game — by 4 PM ET the day before it. Until then he has no game status yet, and his news says what his team has said. Still on the list after that, he is out (IR or PUP); activated, he is cleared.

Cleared: on the report earlier in the week and no longer injured — he is expected to play. A player on IR or PUP, or suspended, is cleared only when ESPN reports him activated or reinstated: ESPN also drops such players from its feed without activating them, and that alone clears no one.

In a game page’s passing, receiving and rushing tables, a player on the report has his status beside his name: O (out, or inactive), D, Q, IR, PUP or SUS, in the card’s color for it, and a gray cross while he has no game status yet — a player designated to return included, until 4 PM ET the day before the game. The injury is in the hover text. A cleared player has no tag. It is the same report as the card’s, so once the game has kicked off it is his status at kickoff — a Q in a played game’s box score is a player who was questionable and played.

From

Injury reports

Shown on

A game page

See also

Injury report

His snaps over his team’s snaps in the same games, summed — not an average of each game’s share. His side is whichever of offense and defense he played more.

At 50% or more he is a regular; from 20% he plays part-time; both are listed. Under 20%, a kicker, punter or special-teamer, or no snaps at all this season (on injured reserve since the summer, say) is a depth player: counted, and named when the count is opened.

A receiver’s share of his team’s targets, or a back’s share of its carries, follows when he has one. Before week 1 is played there are no snaps to go on, so every injured player is listed.

his offense (or defense) snaps ÷ his team’s offense (or defense) snaps, over the games he played

From

Injury reports

Shown on

A game page

See also

Target shareRush share

A score is drawn at the snap of the play it came on, so it can read a few seconds earlier than a broadcast’s scoring summary. A touchdown and its extra point are one step.

From

Play-by-play

Shown on

A game page

The EPA cell is colored like an edge: warm when the offense won those plays.

From

Play-by-play

Shown on

A game page

See also

EPA per play

Plays and yards count passes and runs only — no penalties, kicks or kneel-downs — so they can differ from a broadcast’s drive chart, and yards take off sacks. EPA is the drive’s total, not per play.

From

Play-by-play

Shown on

A game page

See also

EPA per play