Most of these measures are read the wrong way round, by assuming that more is better. None is a score out of ten. Each describes how a job is done, and the useful part is usually what the number does not say.
This page used to sit at the back of every issue. It lives here now, so it can grow with the Data Desk rather than lose a definition every time the section gains a page. Every Data Desk page in the magazine points to it, so you can come back whenever a figure looks wrong.
Standings and league leaders
The standings are the only page in the section that needs no explaining, and the leaders page is the one most often misread. Every leader board carries a qualifying floor, a minimum number of carries or attempts or targets, and that floor rises as the season goes on. It is there to stop a back with four carries leading the league in yards a carry. Early in the year it means a name you expect to see is sometimes missing, and the note on the page says which floor is in force that week.
Success rate
A play succeeds if it gains 40 per cent of the yards needed on first down, 60 on second, and all of them on third and fourth. There is no partial credit, which is the point: yards a play rewards one sixty-yard run in a drive of failures, while this asks how often a team stayed on schedule.
It comes before third down because four fifths of its plays are first and second, and those decide what third down looks like. Where the two disagree, and they often do, believe this one: it rests on about a hundred and eighty plays a team where third down has barely forty. The strip at its foot is the kicking game, which no success-rate measure counts. It is league-wide rather than by team because six attempts apiece cannot tell thirty-two kickers apart.
Third down, by how far
Everyone prints a third-down percentage and almost nobody prints the thing that makes it mean anything, which is how far the team had to go. Converting 42 per cent from third and five is ordinary. Converting 42 per cent from third and nine is exceptional. The single figure treats them as the same, so we split it into four distance bands and let you read across the row.
Read the row, not the cell. A team strong from one and two but poor from seven and ten can lean on a back for a yard and cannot win a passing down. The reverse, strong when long and weak when short, usually means a line that cannot move anyone off the ball, and it is more common than you would think. Seven to ten is the band that decides how watchable a team is, because that is where an offense either has a plan or punts.
The attempts figure beside each percentage is the honesty check. A hundred per cent of one attempt is not information, so treat any cell with a small attempts number as an empty one until the season fills it in. Our conversion is also stricter than a broadcast graphic: the third-down play itself has to reach the line, so a first down given by penalty does not count, and neither does converting on the fourth down that follows. That puts us a point or so below the official number, and we would rather say why than quietly match it.
The deeper numbers
Six measures, all from tracking data rather than the box score.
Quickest release is snap to throw, in seconds. Fast can mean a quarterback who sees it early behind a system built to get the ball out, which is a compliment, or a line that cannot hold a pocket for three seconds, which is not. Read it against sack rate: quick and rarely sacked is a design, quick and sacked anyway is a problem. Slowest is not automatically worst, because holding the ball is how deep throws happen.
Into a loaded box is the share of carries run at eight or more defenders. It is not really about the back. It is the defense saying what it thinks of the passing game. A high number is a compliment to the runner and an insult to the quarterback, and when it falls sharply across a season something has changed about the passing game.
Yards over expected is what a back gains beyond what the blocking in front of him was worth. It is the closest we print to the runner without his line. Read it beside the loaded box figure: high on both is a back carrying his offense, high on the box and low on this is a back being asked to do something nobody could.
Throwing into traffic is the share of passes put into tight coverage, and it measures appetite rather than quality. A quarterback at the top of it is taking on throws most will not, which is either nerve or recklessness, and the column alone cannot tell you which.
Beating the throw is the one that can. It is completion rate measured against what each throw was worth, given its distance and the coverage. Positive means he completes passes he had no business completing. High on traffic and high here is a quarterback whose gambling is paying. High on traffic and negative here is one whose gambling is not, which early in the season is the more useful thing to know.
Getting open is the yards between a receiver and the nearest defender at the catch. The man at the top may simply be running the shortest routes, because separation is easy five yards from the line and hard twenty-five yards down it. Read it with yards a catch: wide open and short is a system working as designed, wide open and deep is somebody else’s problem.
Red zone intelligence
A number that looks like the familiar one and is not. Everywhere else, red zone efficiency is the share of trips that ended in a touchdown. The top of our page is the share of plays that scored, by how far out the team was, which is why the figures look low beside a broadcast graphic. The trips version is lower down the same page.
The red zone is still the twenty. The trips block at the foot of the page counts every possession that reached the opponent’s twenty. The blocks above it are narrower on purpose: inside the ten is where the call is a real choice, and where a yard changes the answer.
The p figure is the chance of seeing a gap this big if there were no real difference. Small means it is probably real. Where it is large, as it usually is at one and two yards out, the page says level rather than invent a winner.
Team tendencies
A different shape on purpose. The other pages are lists of the best eight at something. This is all thirty-two teams, one to a row, ranked by the share of first downs won through the air. The bar is the point: red for the air, pale grey for the ground, and the sliver at the end for the ones handed over by penalty. You can find your team and see how it has chosen to play without reading a figure, and the sort order does the arguing. The top of the table throws and the bottom runs.
Use it as a lens rather than as a league table. Tendency is not quality: a team can be last for throwing because it runs well or because it cannot throw at all, and the columns beside the bar tell you which. Fourth down is made over attempted, so small numbers mean a team has not gone for it rather than that it is bad at it.
Every block carries a minimum volume that rises through the season, so a back with four carries cannot lead a rate. Early in the year, treat all of it as a sketch.
The Data Desk runs every week in The Grid. This guide ran in the magazine until Issue 11 and moved here, to the Reading Room, on 25 October 2026.