The NFL pass interference rule: the penalty with no number

A flag lands on a deep incompletion down the right sideline. Nobody has caught anything, nobody has gained a yard, and the ball is about to move forty-one yards with a fresh set of downs attached. That is the NFL pass interference rule working exactly as written, and it is the reason one judgement made by one official standing twenty yards away can be worth more than an entire drive.

Every other significant foul in the sport has a number stapled to it. Offensive holding is ten yards. Roughing the passer is fifteen. Illegal contact is five. Defensive pass interference is the spot of the foul, which means its value is not set by the rulebook at all. It is set by how far the receiver happened to have run when a defender's arm crossed his body.

Nothing else in American football works like that, and almost every argument about the penalty traces back to that one design decision rather than to the definition of the foul itself. People think they are arguing about whether there was contact. They are usually arguing about whether the contact was worth forty yards.

The NFL pass interference rule has no number attached to it

Start with enforcement, because enforcement is where the strangeness lives.

Defensive pass interference gives the offence an automatic first down at the spot where the foul occurred. There is no maximum. A foul committed eight yards downfield moves the ball eight yards. A foul committed fifty-two yards downfield moves the ball fifty-two. If the foul happens inside the end zone, the ball is placed on the one-yard line, or halfway to the goal line if the previous spot was already inside the two.

Offensive pass interference is ten yards from the previous spot, and the down is replayed.

That asymmetry is not sloppiness. The two fouls are doing different jobs, and once you see which job each one is doing the whole structure becomes legible.

Most penalties in the sport are punitive. They take something from the offending team to discourage the act, and the amount is fixed because the act is what is being punished. Defensive pass interference is not punitive in that sense. It is compensatory. It is trying to restore the offence to the position it would have occupied if the pass had been completed, which is why the amount has to vary: a receiver deprived of a catch at forty yards lost more than a receiver deprived of a catch at nine.

Compensation for a loss cannot be a fixed sum, because the loss is not a fixed size.

The offensive version has no equivalent logic. When a receiver shoves a defender to create separation, nobody has been deprived of yardage, because the down did not happen. So the penalty reverts to the ordinary punitive model, ten yards and play it again.

• 5Yards past the line where contact is legal
• 1Yards downfield before interference can exist
• 10Yards for offensive pass interference
• 1Yard line used for a defensive foul in the end zone

All four are set by the professional rulebook. The distance a defensive interference foul actually costs is not among them, because it is the spot of the foul.

There is a quiet assumption buried in the compensatory model, and it is the source of most of the honest criticism of the rule. Restoring the offence to where it would have been requires deciding what would have happened, and the rule decides that the catch would have been made. Every defensive interference foul is scored as a completion. The offence receives the full value of a reception without anyone having to establish that the receiver would have held on, kept his feet, or even reached the ball had he been left alone.

No other penalty in the sport pays out on a hypothesis. This one pays out on the most favourable hypothesis available.

What actually counts as interference, and what the rulebook lets go

The definition is shorter than most people expect. Interference is any act by a player more than one yard beyond the line of scrimmage that significantly hinders an eligible opponent's opportunity to catch a forward pass.

Two words in there are carrying almost all of the weight. Significantly, which is a threshold rather than a fact. And opportunity, which points at the receiver's route rather than at the collision.

The rulebook then lists the acts that constitute the foul, and the list is worth knowing precisely because it is not what people assume. Defensive interference includes contact by a defender who is not playing the ball and who physically restricts or impedes the receiver. Playing through the back of a receiver. Grabbing an arm in a way that restricts his chance to catch. Extending an arm across the receiver's body, whether or not the defender is playing the ball. Cutting off the receiver's path by making contact without playing the ball. Hooking a receiver in a way that turns his body before the ball arrives.

The list is explicitly open. The rulebook says these acts are included but not limited to, which means the list illustrates the standard rather than defining it. That is the first structural reason two plays that look identical on a screen can produce different decisions: the officials are not matching the contact against a checklist, they are applying a threshold, and the list is a set of worked examples.

The permissible acts matter at least as much and get discussed a tenth as often. Incidental contact by a defender's hands, arms or body is allowed when both players are competing for position and neither is looking for the ball. Feet tangling inadvertently is allowed when both are playing the ball or neither is. Laying a hand on a receiver in a way that does not materially affect him is allowed. Contact by a defender who has already gained position while making a genuine attempt to catch the ball is allowed. And contact that would otherwise be interference is permitted if the pass is clearly uncatchable, which is large enough to need its own section.

Read those together and the two questions an official is actually resolving come into focus. Was the defender making a bona fide play on the ball, or on the receiver? And did the receiver's opportunity change?

This is where the most durable myth about the rule needs killing. The broadcast line is that a defender who never turns his head to look for the ball has fouled by definition. He has not. Nothing in the rule requires a defender to find the ball. A defender who runs stride for stride with a receiver, never looks back, and never restricts him has committed nothing, and this happens on a great many plays every week. Conversely, a defender who does turn round can still commit the foul, because turning to look does not license him to climb through the receiver's back on the way.

Head position is a useful clue for an official deciding whether contact was incidental or directed. It is evidence, not a rule.

Face guarding is the mirror image of the same confusion. Waving a hand in front of a receiver's eyes without touching him is not a foul in the professional game, because the rule is built on contact and hindrance rather than on obstructing vision. The college and high school codes converged on the same treatment, and a defender who blocks the view without touching anyone is playing legally in all three. It looks outrageous on a replay. It is nothing.

The restriction runs on two different clocks for the two teams

This is the piece almost every explanation skips, and it changes how you watch a play.

Offensive pass interference rules apply from the moment the ball is snapped until the ball is touched. Defensive pass interference rules apply from the moment the ball is thrown until the ball is touched.

The windows are not the same length. A receiver is under the interference restriction from the snap. A defender is not under it until the quarterback lets go.

That is the whole reason a separate family of fouls exists to govern the pre-throw period, and it is why a defender's most physical work happens before the release. Until the ball is in the air, the worst that can happen to him is a five-yard foul. The instant it leaves the hand, the same act becomes the only penalty in the book with an unlimited price.

Both windows close on the same event. Once the ball is touched by any eligible player, offensive or defensive, the interference rules switch off completely. A tipped pass makes downfield contact legal from the moment of the tip, which is why a defender who wrecks a receiver on a deflected ball has done nothing wrong and why officials working deep have to register a deflection they may be twenty yards away from. Some of the least popular non-calls in the sport are tips that nobody in the stadium saw.

Then there is the line of scrimmage itself. There can be no pass interference at or behind the line, and no foul at all until one yard past it. A pass caught behind the line, a screen, a swing pass to a back, cannot generate interference no matter what a defender does to the receiver. He can be held, and defensive holding is available, but the big penalty is not.

That threshold is also the one the offence trips over most often, and it is the reason run-pass options sit permanently on the edge of legality. Linemen blocking more than one yard downfield while a forward pass is still live is offensive pass interference, and the entire concept depends on linemen who believe they are run blocking.

An uncatchable ball erases the foul, and nobody has defined uncatchable

Of all the permissible acts, this one decides the most arguments.

Contact that would ordinarily be interference is not a foul if the pass is clearly uncatchable by the players involved. A defender can grab a receiver's arm, turn his body and knock him off his route, and if the throw sails ten yards over both of them there is no flag and no yardage.

The rulebook does not define uncatchable. There is no distance, no cone, no test. The official watching decides whether the players involved could plausibly have reached the ball, and that judgement is made live, at full speed, on a throw that is still in the air when the contact occurs.

Two things make this harder than it sounds.

The first is timing. The contact and the flight of the ball are simultaneous. An official watching hands at fifteen yards is not tracking the arc of the throw, so the catchability judgement is often assembled after the fact from where the ball landed and where the players were.

The second is circular and genuinely unsatisfying. The contact itself can be the reason the ball looks unreachable. If a defender stops a receiver at eighteen yards and the throw lands at twenty-eight, the ball is uncatchable in the only sense that can be observed: nobody got near it. But the receiver did not get near it because he was being held. The exception is written around the players involved and their positions, which folds the effect of the foul into the defence against the foul.

Officials are trained to judge catchability against where the receiver would have been on his route rather than where the contact left him, which is the right instruction and an extremely difficult one to execute. It requires reconstructing a route that was interrupted, in real time, while also deciding whether the interruption crossed a threshold.

The exception is not a loophole. Without it, a pass thrown into the third row would still be worth thirty yards whenever a corner had his hand on a jersey, and offences would throw deliberate rubbish at covered receivers all afternoon. The rule needs it. It is simply the least tractable judgement in a rule already made of judgements.

Offensive pass interference is the same rule pointed the other way

Receivers foul constantly and get called for it rarely, and both halves of that sentence are true for structural reasons.

The offensive acts are the obvious inversions. Pushing or shoving off a defender to create separation. Driving through a defender who has established position. Blocking a defender in the vicinity of the intended receiver while the pass is in the air. Blocking more than one yard beyond the line of scrimmage before the pass is thrown.

The one that generates the most heat is the pick, or the rub. Two receivers cross, one of them collides with the defender covering the other, and the defender is left five yards behind. Whether that is a foul turns on a single question: was the receiver running a route, or was he setting himself to make contact? A receiver running his own route who happens to occupy the same square yard as a defender has committed nothing, and traffic is a legitimate feature of the passing game. A receiver who slows, squares up and absorbs a defender has blocked, and that is offensive interference.

The distinction is real and it is nearly invisible at full speed from twenty yards away, especially since the whole design of a rub is to make deliberate contact look accidental.

Enforcement pulls in the opposite direction to the defensive foul, and the ten yards from the previous spot has a bite the number understates. Offensive interference wipes out whatever happened on the play. A touchdown catch becomes first and twenty. The yardage is modest and the reversal is total, which is why offensive interference feels heavier than its number and why the two fouls are not really comparable instruments even though they share a name.

The same logic explains why offensive interference cannot sensibly be a spot foul. There is nothing to compensate. The defence lost no ground, because the play was undone.

Illegal contact and defensive holding are the same grab, priced differently

These two are constantly mistaken for interference, and understanding them is the fastest way to understand why the timing of the throw matters so much.

Illegal contact is a defender initiating contact with a receiver beyond five yards from the line of scrimmage, while the passer is still in the pocket with the ball, and the receiver is attempting to evade him. Five yards and an automatic first down. It has two conditions that kill it instantly: the ball must not have been thrown, and the passer must still be in the pocket. A quarterback who scrambles out extinguishes the foul, which is the reason defenders are coached to keep grabbing once the pocket breaks.

Defensive holding is broader. Grabbing, hooking or restricting a receiver in a way that impedes him, and it has no pocket condition attached. Five yards and an automatic first down.

Now put the three fouls on the same timeline. A cornerback runs down the sideline with his hand on a receiver's hip at twenty-two yards. If the quarterback still has the ball and is in the pocket, that is illegal contact, five yards. If the quarterback has scrambled right, it is defensive holding, five yards. If the ball is in the air and the grab restricted the receiver's chance to catch it, it is pass interference, twenty-two yards and a first down.

Same defender. Same hand. Same jersey. Three prices, and the variable is where the ball was.

Contact inside five yards of the line0

Illegal contact, passer in the pocket5

Defensive holding, passer scrambling5

Interference, ball in the air at thirty yards30

Constructed illustration. The contact is identical in each case; only the timing and the depth change. The thirty-yard figure is an invented route depth chosen to make the jump visible.

Show the numbers

Worked example: the same grab, priced four ways

Item	Yards awarded to the offence
Contact inside five yards of the line	0
Illegal contact, passer in the pocket	5
Defensive holding, passer scrambling	5
Interference, ball in the air at thirty yards	30

That chart is the article in one picture. The first three bars are set by the rulebook. The fourth is set by the receiver's route, and the receiver's route is decided by the offensive coordinator.

Officials are aware of the cliff, and it shapes how the crew talks to itself. The deep officials are watching bodies, not the ball, so they frequently do not see the release with their own eyes. Whether a downfield grab is a five-yard foul or a thirty-yard one can depend on information the flagging official did not personally observe.

One deep pass, from the throw to the flag

Every deep incompletion runs through the same sequence of questions. They are asked in order, and the first one answered against the flag ends the inquiry.

1. 1The pass is released down the right sidelineThe receiver is running a route roughly thirty yards downfield with a defender in trail position. The defender has not looked back for the ball.
2. 2Was the contact more than one yard beyond the line?If the contact happened at or behind the line of scrimmage, no interference foul exists at all. Defensive holding is the only thing available, and it is worth five yards.
3. 3Had the ball been thrown when the contact happened?If not, this is not interference. It is illegal contact if the passer was still in the pocket, or defensive holding if he was not. Either way it is five yards and an automatic first down.
4. 4Had the ball already been touched by anybody?A tip by a defender or by any eligible receiver switches the interference rules off instantly. Everything after the touch is legal. Officials working deep must register a deflection they may not have a clean view of.
5. 5Was the defender making a bona fide play on the ball?A defender who has gained position and is genuinely attempting to catch it is permitted contact that would otherwise be a foul. Not looking back is not, by itself, evidence either way.
6. 6Was the receiver's opportunity significantly hindered?This is the threshold, and it is about the route rather than the collision. A hand laid on a receiver that does not materially affect him is a permissible act. An arm across the body that turns him is not.
7. 7Was the pass clearly uncatchable by these two players?If the throw was never reachable, the contact is permitted no matter how obvious it looked. The official must judge this against where the receiver should have been, not where the contact left him.
8. 8The flag lands, and the spot is markedThe offence receives an automatic first down at the point of the foul. The distance is whatever the receiver had run. Nothing in the rulebook caps it.

The questions are resolved in this order. Any one of them answered the wrong way for the offence ends the sequence with no interference foul available.

Watch that sequence a few times and the pattern in the arguments becomes obvious. Almost nobody disputes step one, two or four. The shouting happens at the threshold question and at the catchability question, which are the two that have no factual answer.

Spot foul or fifteen yards: which one is fairer

The college game enforces the same foul differently, and comparing the two is the cleanest way to see what each system buys and what it pays for.

In college, defensive pass interference is a spot foul when the contact occurs less than fifteen yards downfield, and is capped at fifteen yards from the previous spot when it happens beyond that. Either way it carries an automatic first down. The professional game applies the spot with no ceiling.

• Professional enforcement
• College enforcement

Foul at 8 yards88

Foul at 20 yards2015

Foul at 35 yards3515

Foul at 50 yards5015

Constructed arithmetic applying the two published enforcement systems to four invented route depths. It shows the rules, not any team's results.

Show the numbers

Worked example: what one defensive interference foul is worth

Item	Professional enforcement	College enforcement
Foul at 8 yards	8	8
Foul at 20 yards	20	15
Foul at 35 yards	35	15
Foul at 50 yards	50	15

The two systems agree completely at short range and diverge without limit at depth. That divergence is the entire debate.

The case for the cap is proportionality, and it is a strong one. The size of the penalty in the professional game is decided by the length of the receiver's route, which is a decision made by the offence, not by the defender. A shirt tug at forty-eight yards is worth three times a clear tackle at fifteen. The severity tracks the geometry of the play rather than the severity of the act, and a system that prices marginal contact higher than blatant contact because of where each happened is doing something odd. It also means the variance in officiating is amplified by distance: a borderline call at fifty yards is a fifty-yard error, and borderline calls are the ones the rule is least able to get right.

The case for the uncapped spot is not about fairness at all. It is about incentives, and it is the reason no capped version has been adopted.

Cap the penalty and you create a rational foul. A cornerback beaten by three steps on a fifty-yard route has an arithmetic decision available: concede a fifty-yard completion, or tackle the receiver and concede fifteen. Under a cap the tackle is the correct play, comfortably, every time. That is not a hypothetical criticism of the college structure so much as a known property of it, and it is why the argument about capping the professional penalty recurs and never gets anywhere. Whatever the spot foul's flaws, it removes the incentive to commit the foul deliberately, because you can never buy the ball back for less than what you were going to give up.

The professional game has one other lever, held in reserve and almost never pulled. Where a team presents an apparent certain scoring opportunity and the defence commits a palpably unfair act, the referee may award a score outright. It exists precisely to cover the case where no ordinary penalty is severe enough to remove the benefit of cheating, and its rarity is the point.

There is a third structure nobody runs, which would split interference into two fouls: a lesser one for contact that restricted a receiver marginally, and a major one for contact that clearly destroyed the play. That gets proportionality and keeps the deliberate foul unprofitable. It also imports a brand new judgement call about severity, made by the same official, at the same speed, in the same second. Every proposed fix for this rule turns out to require more judgement rather than less.

Why the NFL pass interference rule is not reviewable, and the season it was

For one season, it was reviewable. The experiment is the most informative thing that has ever happened to this rule, and it did not go the way anyone expected.

The mechanism: coaches could challenge both interference calls and interference non-calls during the first twenty-eight minutes of each half, using the same two challenges they already had. In the final two minutes of each half, and in overtime, reviews were initiated from the replay booth instead. The standard for overturning was clear and obvious visual evidence that the on-field decision was wrong.

It lasted a single season and was not renewed.

What killed it was not that it produced bad answers. It was that it barely produced answers at all. The standard was applied so strictly that the great majority of reviews confirmed whatever had happened on the field, including plenty of plays that looked to everyone watching like the sort of thing the rule had been introduced to fix. The apparatus added stoppages, added a new category of argument, and changed very few outcomes.

Three lessons came out of it, and they generalise well beyond this one foul.

Replay is built for facts, and interference is not one. Every review that works well in the sport resolves a binary physical question. Did the foot land in bounds. Did the ball cross the plane. Was the knee down. There is a correct answer sitting in the video, and the camera either shows it or does not. Interference asks whether a receiver's opportunity was significantly hindered, which is a threshold applied to a continuum. Slowing the picture down does not reveal the answer, because the answer was never in the picture. Review replaces one person's judgement with another person's judgement, taken from better angles.

Making a foul reviewable turns every non-call into an accusation. Before the experiment, a play with contact and no flag was a decision. Afterwards it was a claim awaiting adjudication. Since contact occurs on a large share of pass routes, the population of plays with a plausible case is enormous, and coaches had to start spending challenges on lottery tickets. The rule did not just add reviews. It changed what a quiet play meant.

The evidence threshold has no natural setting. Set it low and a large fraction of contested downfield throws get re-officiated in slow motion, which slows the game and moves authority from the field to a screen. Set it high, which is what happened, and almost nothing is reversed, which makes the whole structure theatre. Both failure modes are real and there is no principled point between them, only an arbitrary one that somebody has to choose and then defend every week.

The Canadian game has kept a version of interference review and took several seasons to arrive at a standard its officials and coaches could live with. That is the honest counterexample, and it carries the same moral: it can be done, slowly, by an organisation prepared to spend years defining what it wants reviewed. The professional American game wanted the problem solved by the following autumn.

The general architecture of what can and cannot be challenged still reflects that conclusion. Reviewable items are overwhelmingly factual. Judgement fouls are not on the list, and the one time an exception was made it was withdrawn after a year.

What the seven officials can actually see

A crew is seven people, and understanding where each of them is standing explains more about interference calls than any amount of argument about bias.

The referee stands behind the passer. His job is the passer, which is why he is the authority on when a hit on the quarterback crosses the line, and it means he knows to the frame when the ball was released and almost never sees the contact thirty yards away. The umpire works behind the defensive line, watching blockers, which makes him the official most likely to catch a lineman a yard and a half downfield on a pass. The down judge and line judge work the two sidelines at the line of scrimmage. The field judge, side judge and back judge work deep, and they are the ones who call this foul.

Here is the part that surprises people. An official covering a deep route does not watch the ball. He watches the players. If he looks up to track the flight, he loses the hands and the feet, which is where the foul lives. So the deep officials are deliberately looking at the one thing they are being asked to judge and are deliberately not looking at the throw that determines what their judgement is worth.

The result is a division of labour with an unavoidable seam. The official with the best view of the contact often has no clean view of the release or of a tip at the line. The official who saw the release has no view of the contact. That is not incompetence and it cannot be trained away, because a human being cannot focus on two things twenty-five yards apart.

Two further distortions are worth understanding before shouting at a replay.

The camera lies about depth. The standard broadcast angle is a long telephoto shot from the sideline, which compresses distance and stacks two players who are actually a yard apart into apparent contact. Light incidental brushing looks like restraint. The official, standing side on with real depth perception, is seeing a different play from the one on your screen, and it is not obvious that his version is the worse one.

Slow motion lies about intent. Every hand becomes deliberate at a quarter speed. A defender fighting for balance looks like a defender grabbing. This is precisely why the replay experiment produced so few reversals: officials in the booth, looking at the same slowed pictures, kept concluding that what looked damning at quarter speed was ordinary at full speed.

The relationship to coverage structure matters too. The deep sideline is the hardest space on the field for a defender to contest without contact, which is exactly why two-deep zone shells leak there, and it is also the space where an official has the worst angle and the biggest yardage consequence riding on the call. The rule's largest penalties are concentrated in the geometry where the officiating is hardest.

The honest criticism: the value is set by the route, not by the act

Everything above is the rule working properly. The criticism worth making is not that officials get it wrong. It is that the rule's design makes their errors unusually expensive and unusually arbitrary.

The price is decided by the offence. A defender controls whether he commits the foul. He does not control how far downfield he is standing when he commits it, because the route was drawn by somebody else. Two defenders committing exactly the same act on the same afternoon can be charged twelve yards and fifty-one yards.

The penalty pays out on an assumption. Compensation requires a counterfactual, and the rule always resolves it in the offence's favour. The catch is assumed. The receiver is assumed to have kept his feet, held on, and stayed in bounds. Whether he actually would have is the sort of question the catch rule spends thousands of words failing to settle when the ball is actually caught, and the interference rule disposes of it by assumption when it is not.

It converts a threshold judgement into a large swing. The whole call turns on the word significantly, which is a dial rather than a switch. Any judgement near the middle of that dial is close to a coin flip, and the rule multiplies coin flips by route depth. A borderline decision at eight yards costs eight. The identical borderline decision at forty-eight costs forty-eight.

It changes what a bad throw is worth. This one gets said least and matters most. If a quarterback throws deep into tight coverage, the possible outcomes are a completion, an incompletion, an interception, and a flag worth forty yards. That fourth branch changes the arithmetic of the decision, and offences would be irrational to ignore it. Contested deep throws are not purely attempts to complete passes. Some of them are attempts to generate a flag, and the rule makes that a defensible use of a down.

There is no symmetrical instrument for the defence. Offensive interference is capped at ten yards from the previous spot. No foul the offence can commit hands the defence a comparable swing. The largest single-play penalty in the sport runs in one direction only.

The available fixes each fail in a different way. Capping it makes deliberate deep fouls profitable. Reviewing it was tried and expired. Tightening the catchability standard adds a judgement rather than removing one. Splitting it into two grades of foul requires officials to sort severity live. And leaving it alone preserves a rule where a marginal decision by a tired official on the last play of a half can be worth more than three quarters of offence.

That is not an argument that the rule is wrong. It is an argument that there is no version of it that is right, which is a different and more uncomfortable conclusion. The foul that keeps the deep passing game honest cannot be priced honestly.

How to judge a pass interference shout live

The useful version of all this is a short sequence you can run in the two seconds between the flag landing and the referee switching his microphone on. Watch a few weeks of how the moving parts of American football fit together at this level and pass interference stops being a mystery and becomes an argument you can actually have.

Find the release before you find the contact. Whether the ball was in the air is the single largest variable in the whole rule, and it decides whether you are watching a five-yard foul or a forty-yard one. If you cannot tell from the replay, neither could the official standing thirty yards away from the quarterback.

Check the line of scrimmage and check for a tip. Contact at or behind the line cannot be interference. Contact after any touch cannot be interference. Both facts are easy to establish and both remove the possibility of a flag entirely, and both are missed constantly by people who have already decided what they saw.

Watch the hands, not the head. A defender who never turns round has not fouled by that alone, and a defender who does turn round has not earned immunity. The question is whether his arms restricted the receiver, and it is answered at the wrists.

Ask whether the route changed, not whether there was a collision. Two players fighting for the same yard of grass is football. The threshold is whether the receiver's opportunity was significantly hindered, so the thing to look for is a receiver knocked off his line, turned early, or unable to extend. Contact that leaves the route intact is a permissible act and is supposed to be.

Ask whether the ball was reachable, and then ask why it wasn't. If the throw was never getting there, the contact is legal. But check whether it failed to get there because it was badly thrown or because the receiver was stopped from running at it. The difference is the whole foul, and it is the judgement that generates the most disagreement precisely because there is no way to observe the route that did not happen.

Then look at the spot, because the spot is what decides the game. By the time the ball is marked, the argument about whether there was contact is finished and irrelevant. What matters is that a threshold judgement, made in a fraction of a second, at a bad angle, about a continuum, has just been converted into a specific number of yards that nobody wrote down in advance. The flag is the small part. The distance is the rule.