Run pass option explained: making one defender wrong

An offensive lineman fires out of his stance and drives a defensive tackle backwards. The running back takes a handoff path towards the right guard's outside hip. Everything about the picture says run, and for about four tenths of a second it genuinely is one.

Then the quarterback pulls the ball out and throws a slant.

Run pass option explained in one sentence: it is a play that is a run and a pass at the same time, and stays both until the quarterback decides. That sounds like a gimmick and it is not. The RPO is the most efficient answer American football has found to a very old arithmetic problem, and the reason it took over the sport in a decade is that it does something no other concept does. It blocks a defender with a threat instead of with a body.

The mechanism is simple to state and much harder to defend: find one player, put him in a position where both of his jobs are live, and let him choose which one to get wrong.

The arithmetic problem the RPO solves

Start with the numbers, because everything else follows from them.

An offence lining up to run the ball has five offensive linemen plus, depending on personnel, a tight end and a fullback or a lead blocker. It cannot block everybody. A defence has eleven players and can put as many of them near the line as it likes, and the moment it commits one more defender to the box than the offence has blockers, somebody is running free at the ball carrier.

• Defenders in the box
• Blockers the offence can supply
• Defenders left unblocked

Six-man box660

Seven-man box761

Eight-man box862

Constructed arithmetic on stated assumptions, not measured data. The offence is assumed to have six available run blockers on the play, and the quarterback is treated as a non-blocker. The bars show what happens as the defence adds bodies to the box.

Show the numbers

Worked example: the blocking arithmetic that makes RPOs necessary

Item	Defenders in the box	Blockers the offence can supply	Defenders left unblocked
Six-man box	6	6	0
Seven-man box	7	6	1
Eight-man box	8	6	2

That constructed comparison is the whole reason the play exists. In the six-man box the run is fully blocked and the offence should run it every time. In the seven-man box there is one defender nobody can reach, and traditionally the offence had two choices: block him with a scheme that leaves someone else free, or accept that the run gets stopped.

The RPO adds a third choice. Do not block him. Threaten him instead. If the unblocked defender comes forward to make the tackle, throw the ball into the space he has left. If he stays back to cover that space, hand the ball off and run at a box that is now, functionally, one man short.

That is the entire concept. The defender is not beaten physically, he is beaten by being given two jobs on a play where only one of them can be done. And unlike a blocking scheme, this method does not get worse against better defenders. A faster, more aggressive linebacker is a more useful read, not a harder one, because he commits earlier and commits harder.

The read defender, and how a coach picks him

The read defender is usually called the conflict defender, and the name is the definition. He is the player whose assignment on this specific defensive call includes both a run responsibility and a pass responsibility that cannot be satisfied simultaneously.

Choosing him is not arbitrary. Three properties matter.

He has to be genuinely conflicted. A defender who has a run fit and no coverage duty is not a read, he is a problem. A defender in deep coverage with no run fit is not a read either. The read has to be somebody whose two jobs are actually incompatible on this snap, which means the offence is reading a defensive scheme as much as a person. Against a coverage that puts a linebacker in the hook zone directly behind the run fit he is responsible for, that linebacker is the read, and against a different coverage the same player might not be.

He has to be visible from the mesh point. The quarterback is standing with his back partly turned, extending the ball into the running back's stomach, and has perhaps four tenths of a second. He cannot scan. He can look at one place. A read defender who is behind the quarterback, or obscured by linemen, is unreadable no matter how conflicted he is, which is why so many reads are second-level defenders on the play side or overhang defenders on the edge where the quarterback's natural head position already points.

His movement has to be unambiguous. The best reads are defenders whose two possible actions look completely different in the first half-second. A linebacker who steps downhill into a run fit and a linebacker who drops into a hook zone are moving in opposite directions, which is a clean read. A defender whose run fit and coverage drop start with the same shuffle is a bad read, because the quarterback will be guessing rather than reading, and guessing is how RPOs turn into interceptions.

Which defender ends up carrying the conflict is largely determined by what the defence is playing. A single-high safety structure asks a lot of its second-level defenders, because there are eight defenders responsible for both run fits and underneath zones with only one deep player behind them. That is exactly the arrangement RPOs punish hardest, and the mechanics of it are set out in the piece on how a three-deep zone divides the field.

Pre-snap, post-snap, second level: three different plays that share a name

The term RPO covers three genuinely different concepts, and conflating them is why the play is so often explained badly.

The pre-snap read. The quarterback looks at the defensive alignment before the ball is snapped, counts bodies or checks the leverage on a perimeter receiver, and decides. If there are six in the box, run it. If there are seven and the slot corner is playing eight yards off the receiver, throw the quick screen. This is the version most casual descriptions actually describe, and strictly it is not an option at all. It is a check. The decision is made before the snap, and after the snap the play is whatever the quarterback chose.

It is still valuable, and its value is arithmetic rather than deceptive. A pre-snap RPO guarantees the offence never runs into a loaded box and never throws into a defence built to stop that throw. The ceiling is low and the floor is high, which is a trade many offences will take.

The post-snap first-level read. The quarterback reads a defensive lineman or an edge defender at the mesh point. This is the closest RPO relative of the read option, and it produces the fastest throws in the concept family, usually to a receiver breaking behind the read defender's shoulder.

The post-snap second-level read. The classic RPO and the most dangerous. The quarterback rides the mesh point and reads a linebacker or safety who is seven to twelve yards off the ball. Because that defender is further away, the quarterback gets fractionally more time and the throw can go further downfield, which raises the ceiling of the play considerably. This is where glance and pop concepts live, and it is where the offence stops merely being efficient and starts producing explosive plays.

The three are not alternatives, they are layers, and a well-built offence uses the same run action with all three attached at different points in a game. The defence that solved the bubble screen in the first quarter is now the defence that gets hit by a glance route off identical run action in the third.

What it is not: play-action and the read option

Two concepts get confused with the RPO constantly, and the differences are mechanical rather than semantic.

Play-action is a pass with a run fake attached. The decision was made before the snap and the run was never going to happen. The running back's job is to sell the mesh and then protect, the linemen usually protect rather than drive, and the routes can develop for two or three seconds because nobody is required to throw quickly. Play-action attacks a defence's eyes and its run keys, and it can produce shots down the field that an RPO structurally cannot. The full mechanics of that concept, including why it works even when the defence knows the run threat is fake, are covered in the article on how play-action actually manipulates a defence.

The difference in one line: on play-action the run is a lie. On an RPO the run is true until the quarterback decides otherwise, and it is a handoff a great deal of the time.

The read option is a run with a run attached. The quarterback rides the mesh and reads an unblocked edge defender, then either hands off to the back or keeps the ball himself and runs. Both outcomes are runs. No forward pass is involved, which means none of the rules about ineligible players downfield apply, which means the play has no timing constraint at all. It is also the concept that exposes the quarterback to contact, which is the main reason professional offences use it sparingly and college offences use it a great deal.

The RPO borrowed its mesh point from the read option and its throw from the quick passing game, and inherited a legal problem from neither of them individually.

The one-yard rule is the real constraint, and everything about the play bends around it

Here is the part that decides what an RPO can and cannot be, and it is a rule rather than a tactic.

An offensive lineman is an ineligible receiver. NFL rules prohibit an ineligible player from being more than one yard beyond the line of scrimmage when a forward pass crosses that line. College football is more generous and permits three yards. In both codes the penalty is five yards.

• 1Yards an ineligible NFL player may be downfield
• 3Yards an ineligible NCAA player may be downfield
• 5Penalty yardage for the foul in both codes

Rule-defined figures taken from the NFL and NCAA rulebooks rather than measurements. The distance is measured from the line of scrimmage at the moment the forward pass crosses it.

Now consider what a run block looks like. An offensive lineman firing out of a three-point stance to drive a defender off the ball is trying to create movement, and creating movement means going forward. A yard is nothing. A guard executing a good down block will be past a yard before the quarterback has finished the mesh.

That single fact dictates the shape of the entire concept:

• The throw has to be immediate. Not quick, immediate. If the ball is not out within roughly the time it takes a lineman to take two steps, the play is illegal in the NFL, whatever else happens.
• The routes have to be short and flat. Slants, bubbles, hitches, quick outs, pops into the vacated linebacker level. Anything requiring a receiver to run for two seconds is unavailable, because the linemen cannot wait.
• The quarterback cannot progress through reads. He has one read, one throw, and no second option. If the read is ambiguous, he hands the ball off, because that is the only safe default.
• The blocking scheme has to be one that stalls naturally. Zone concepts are more RPO-friendly than power schemes for exactly this reason, because a zone blocker working laterally covers less ground downfield than a puller driving forward. The reasoning behind those blocking schemes is worked through in the piece on how zone blocking is taught.

The three-yard college allowance is not a small difference. Three yards is roughly three times as much time, which means college RPOs can hold the mesh longer, throw deeper routes and read defenders further from the ball. That is why the concept matured in college football and arrived in the professional game in a compressed, faster form, and why plays that work every week on Saturday are illegal on Sunday.

The enforcement argument, which is the honest part of this subject

The rule says one yard. The rule as enforced is something else, and everyone involved knows it.

Judging the foul requires an official to watch five offensive linemen, note their depth relative to the line of scrimmage, and freeze that picture at the exact instant a forward pass crosses that line, while also officiating everything else happening on the play. There is no line painted on the field a yard beyond the scrimmage line. The official is estimating a yard, on grass, at speed, on a play designed to make the estimate difficult.

The practical result is that the foul has historically been called only when a lineman is obviously and egregiously downfield, which in practice means several yards rather than one. Offensive coaches have built their concepts inside that gap between the written rule and the enforced one, and defensive coaches have complained about it for as long as the concepts have existed.

Both complaints are reasonable, which is what makes this genuinely interesting rather than a simple grievance.

The defensive argument is that the offence is being permitted an advantage the rulebook explicitly denies it. A linebacker reading a lineman's aggressive run block is reading correctly, and if he steps downhill and the ball goes over his head, he has been beaten by an action the rules do not allow. He is being punished for playing football properly.

The offensive argument is that strict enforcement would not restrict the play, it would abolish it. A one-yard limit called precisely would make it essentially impossible to run block and throw a forward pass on the same snap, which would remove not just RPOs but a substantial portion of the modern quick passing game. Whether that is a bad outcome is a legitimate question, but it should be answered deliberately rather than arrived at by an officiating memo.

There is also a structural point that gets lost. The rule was not written with RPOs in mind. It exists to stop linemen from releasing downfield to block on a pass play, which is a different abuse from the one it is now being asked to police, and applying an old rule to a new concept produces exactly the kind of grey zone the game is currently living in.

What has changed recently is the mechanism rather than the rule. For the 2025 season the NFL expanded its replay assist system, and ineligible player downfield on a pass play is among the fouls the replay booth may now flag. A coach cannot challenge it, which matters: the booth initiates, so the foul is more likely to be caught on plays the booth is already looking at than to be policed consistently across every snap. The broader system, including what may be challenged and what may only be assisted, is set out in the piece on how the NFL's review process is structured. It is a meaningful shift, and it moves the enforced rule slightly closer to the written one without closing the gap.

The concepts: glance, bubble, slant and pop

Four route attachments do most of the work, and each is designed against a different defensive weakness.

The bubble. A slot receiver takes a step backwards and outwards, and receives the ball behind the line of scrimmage with two blockers in front of him. This is the least sophisticated of the family and the most forgiving. It reads an overhang defender on the edge of the formation, it is thrown behind the line so no ineligible downfield question arises at all, and its purpose is usually to make the defence account for the perimeter so the inside run has better numbers. Nothing about it is explosive. Its job is to make the defence widen.

The slant. A receiver breaks inside at a shallow angle and receives the ball in the window a linebacker has just vacated by stepping into his run fit. This is the archetypal second-level RPO. It is thrown flat and fast, the catch point is usually five to seven yards downfield, and everything depends on the quarterback releasing the ball as the linebacker's weight commits forward rather than after.

The slant also carries the concept's characteristic risk. The ball is being thrown into a zone that a defender occupied a moment ago, at a receiver moving across the field, from a quarterback whose eyes have been on the defender rather than the target. If the read is late by a beat, the ball arrives at a linebacker who never actually left, and RPO interceptions are disproportionately of this exact kind.

The glance. A backside receiver runs a skinny post, breaking inside towards the middle of the field behind the linebackers, and the read is usually a safety or the backside second-level defender. This is where the RPO stops being an efficiency play. Against a single-high structure, a glance route attacks the one deep defender directly, and if he has been influenced by the run action the receiver is running into an empty middle third with the ball already in the air.

The glance is also the concept the one-yard rule squeezes hardest, because the receiver has to travel further before the ball can arrive. It is the reason the college and professional versions of this play look so different.

The pop. A tight end or slot receiver releases vertically into the exact area a linebacker occupies before the snap. There is no route running to speak of. The receiver simply goes to where the defender was, and if the defender has vacated to fill a run gap, the throw is a short, flat pass into the hole he left. It is the purest expression of the concept: the receiver is not beating anybody, he is occupying the consequence of a decision.

Each of these is attached to run action, and the run action matters. The same slant behind inside zone and the same slant behind a gap scheme create different conflicts, because the two run plays ask the read defender to fit differently. Coaches build these as packages rather than as plays, which is why an offence can present the same picture two dozen times in a game and produce a different answer to whatever the defence did last.

36%27%36%

• Front players with a run fit and no coverage duty4
• Second-level and overhang defenders with both a run fit and a coverage assignment3
• Perimeter and deep defenders in coverage with no run fit4

A constructed illustration of one common arrangement against a single-high structure, not a prescribed defence and not a measurement. The point is the middle group: those are the players an RPO is designed to attack, because they are the only ones with two live jobs.

Show the numbers

Worked example: how a defence's eleven divide against an RPO

Item	Value
Front players with a run fit and no coverage duty	4
Second-level and overhang defenders with both a run fit and a coverage assignment	3
Perimeter and deep defenders in coverage with no run fit	4

An RPO, step by step

1. 1The call arrives with two plays attachedThe huddle or signal contains a run concept and a route concept together. Nobody on the offence except the quarterback knows which one will happen, and every other player executes his own half regardless.
2. 2Pre-snap: count and leverageThe quarterback counts bodies in the box and checks the alignment of the defenders over the route side. Some offences let him kill one half of the play here, which turns the option into a simple check.
3. 3The snap: everybody commitsThe offensive line run blocks. The running back runs his path. The receiver runs his route. None of them are waiting to find out what the play is, and none of them can afford to hedge.
4. 4The mesh: the ball goes into the stomachThe quarterback extends the ball into the running back's midsection and rides it, holding the handoff open. This is the moment the play is genuinely both things at once, and it lasts a fraction of a second.
5. 5The read: one defender, first movementThe quarterback's eyes are on the conflict defender and nowhere else. He is reading the first unambiguous commitment: weight forward into a run fit, or hips turning into a drop.
6. 6The decision: default is the handoffIf the read is clean and the defender came forward, pull the ball. If the read is ambiguous, give it. An RPO quarterback who guesses on unclear reads produces interceptions, and the handoff is always the safe answer.
7. 7The clock the offence cannot seeWhile all of this happens, the offensive linemen are moving downfield. One yard in the NFL, three in college. Nothing about the quarterback's process changes this deadline and nothing extends it.
8. 8The throw: flat, fast, no second optionThe ball comes out on a line to a receiver who is already at his break point. There is no progression and no scramble drill. If the throw is not there, it should not have been pulled.
9. 9The consequence for the read defenderWhichever way he moved, he is now wrong. If he filled the run, the ball is behind him. If he dropped, the run has an extra blocker for the box.
10. 10The next snapThe offence has learned how this defender behaves under this run action, and the same picture will be shown again with a different attachment. The read defender's next decision is now being made with more doubt than his last one.

The sequence a coached RPO runs through. Timings are approximate coaching descriptions of a very short interval rather than measured values, and different offences teach different mesh depths and reads.

Why it punishes an over-committed run defence, and why that is the point

A defence stops a modern run game by being sound in the box: enough bodies, gaps accounted for, defenders fitting downhill quickly. Every one of those requirements creates the conflict the RPO feeds on.

Committing an extra defender to the box means one fewer in coverage, which is the pre-snap RPO's entire justification. Playing aggressive run fits from the second level means linebackers who move forward on their first read, which is exactly what a slant or a pop is waiting for. Rotating a safety down to add a body against the run means a single-high structure, which is what makes a glance route viable.

Playing the run well and defending an RPO are therefore not merely difficult to do at once. They are, at the level of individual assignments, directly opposed. This is why an RPO offence does not need to guess correctly. It needs only to be right about the structure, and the structure is visible before the snap.

The consequence for defensive coaching has been substantial. Sub packages that were designed to add speed against passing situations are now on the field on standard downs, because the defence needs second-level players who can cover as well as fit, and the trade-offs of those personnel groupings are worked through in the article on why defences play with five and six defensive backs.

There is a limit to the punishment, though, and it is worth being honest about it. Most RPOs produce moderate gains. A bubble screen against soft coverage is a few yards. A slant into a vacated hook zone is a first down at best. The concept is an efficiency machine rather than an explosive one, and an offence that lives entirely on RPOs tends to produce long drives that stall, because the play distribution is narrow and the yardage per snap is capped by the design. The exception is the glance concept, which is the only member of the family with a genuinely high ceiling and also the one hardest to run legally.

That efficiency-versus-explosiveness trade shows up clearly in drive-level measures of value, where a series of six-yard gains and a single thirty-yard gain are worth quite different things depending on down and field position. The framework for that comparison is set out in the piece on what expected points actually measure.

The defensive answers, ranked by how well they actually work

Defences have spent a decade building responses, and they fall into five families of varying honesty.

Remove the conflict by reassignment. The cleanest structural answer. Assign the read defender exclusively to the pass and fit the run with somebody else, usually by having a different player scrape over the top or by spilling the run outside to an unblocked defender who is coming anyway. This does not defeat the play so much as refuse to participate in it: the quarterback reads a defender who no longer has a decision to make, so the read gives no information. The cost is that the run fit now belongs to someone further away, which means the run gains a couple of yards more than it otherwise would.

Disguise and rotate. Show one structure before the snap and become a different one after it. A safety who is fourteen yards deep at the snap and eight yards deep a second later has invalidated the quarterback's pre-snap count without ever being in position to be read. This is the most effective answer against pre-snap RPOs specifically, and it does nothing at all against a clean post-snap read, because a post-snap reader is looking at the defender's actual movement rather than at his alignment.

Attack the timing. Jam the RPO receivers at the line. The concept depends on a receiver arriving at a specific spot at a specific moment, because the throw cannot wait, and a receiver who is two tenths of a second late has broken the whole play, since the quarterback has already had to give the ball. This is why press coverage on the slot has become so much more common on standard downs.

Blow up the mesh. Penetration by an interior defender arriving at the mesh point removes the option entirely, because there is no longer a moment in which both plays are live. The difficulty is that pressure vacates the zones RPOs throw into, so an all-out blitz can be the worst possible call against a good RPO quarterback. Simulated pressures that create penetration without emptying the second level are the more precise answer, and the general family of those calls is covered in the piece on how pressure packages are built.

Play light and make them run. The most honest answer and the least satisfying. Play two deep safeties, keep second-level defenders out of conflict, take away everything the RPO throws into, and accept that the offence can run the ball for reasonable gains all afternoon. The bet is that an offence needing twelve plays to score will make a mistake before it gets there. It is a genuine strategy, it works more often than its reputation suggests, and it requires a defensive coordinator to accept being visibly out-gained on the ground, which is a professional risk of its own.

Ask for the flag. The last and least reliable option: point at the offensive linemen and hope. As enforcement mechanisms tighten, this becomes marginally more viable, but a defence that is relying on an officiating judgement to stop a play has not solved anything.

The pattern across all of these is worth naming. None of them beat the RPO on its own terms, because on its own terms the play is unbeatable by design, and a genuinely conflicted defender genuinely cannot do both jobs. Every workable answer instead changes the premise: eliminate the conflict, hide the structure, break the timing, or refuse the trade and take the run. That is what modern defensive American football is largely about, and it explains why defensive innovation of the last decade has been so heavily focused on disguise and post-snap rotation rather than on new coverages.

What to watch, and what tells you an RPO is coming

Four things are visible on a normal broadcast angle, though the camera makes the third one hard.

Watch the offensive linemen, not the quarterback. If they are firing out low and driving on a play that ends in a forward pass, you have watched an RPO, whatever the commentary says. Linemen who set and retreat are pass protecting, and that is play-action or a straight dropback. This single tell separates the two concepts more reliably than anything else on the screen.

Find the conflict defender before the snap. Look for a second-level defender aligned inside a slot receiver with a run gap in front of him. He is the read on a large share of snaps, and once you have identified him you can watch him rather than the ball and see the play being decided.

Time the throw. An RPO pass is out almost as soon as the mesh ends. If the quarterback holds the ball for more than a beat after pulling it, either the play was not an RPO or something has gone wrong and he is now improvising with linemen illegally downfield.

Check the safety count at the snap and a second later. If it changes, the defence is disguising, and a pre-snap RPO decision has just been made on information that is no longer true. Whether the offence has an answer to that tells you most of what you need to know about how well the concept has been installed.

None of this requires a coaching background. The RPO is one of the most legible plays in the sport once you know to watch the linemen and one linebacker rather than the football, and the moment the read defender commits is visible in real time. He takes one step, and the play he was trying to stop is the one that happens.