Stat Explainer

The Run-Pass Option (RPO): One Play, Two Outcomes

The QB reads one conflict defender after the snap and either hands off or throws - taking whatever the defense gives, bounded by the linemen-downfield rule.

By The NFL Analytics Editorial Team · Published June 14, 2026

One Play, Two Outcomes, Decided After the Snap

The run-pass option, or RPO, is one play that contains two: a run and a pass packaged together, with the quarterback deciding which to execute after the ball is snapped by reading a specific defender. Instead of choosing run or pass in the huddle and hoping it matches what the defense does, the offense lets the defense declare itself first - then takes whatever the read tells it the defense gave up. It is the closest thing football has to making a defender wrong no matter what he does.

The mechanic is elegant. The line blocks as if it is a run. The back takes a run path. A receiver runs a quick route. The quarterback puts the ball in the back's belly and watches one defender - typically a linebacker or safety on the edge of the box. If that defender commits to stopping the run, the quarterback pulls the ball and throws into the space the defender vacated. If the defender drops to take away the throw, the quarterback hands off into a box now short a man. The conflict defender cannot be right.

The short version: an RPO is a run and a pass on the same snap. The quarterback reads one "conflict" defender post-snap and either hands off or throws - taking whatever that defender's reaction leaves open. It stresses defenses by punishing them for committing either way.

The Read: Putting a Defender in Conflict

Every RPO is built around a single defender the offense chooses to "read" - the player the blocking scheme deliberately leaves unaccounted for so the quarterback can use him as a decision key. Because the offense is not blocking him, that defender is free to make a play; the trick is that whichever play he makes, the quarterback counters it.

  • Read the box. If the read defender (often a second-level linebacker) flows downhill to fill against the run, the box is now outnumbered against the pass behind him, and the quarterback throws.
  • Read the flat or seam. Other RPOs read a defender responsible for a throwing window - if he widens or sinks to cover the route, the run lane opens and the quarterback hands off.
  • Pre-snap and post-snap layers. Many RPOs stack a pre-snap look (count the box; if there are light numbers, lean one way) on top of the post-snap read, giving the quarterback two chances to find the soft spot.

The routes attached are almost always quick-developing - slants, hitches, bubble screens, quick seams - because the throw has to come out fast, before the run-blocking up front becomes illegal downfield action. The whole concept lives in a narrow timing window: long enough to read the defender, short enough to stay legal.

The Linemen-Downfield Rule

RPOs exist in the form they do because of one specific rule, and understanding it is essential to understanding why these plays look the way they do. Offensive linemen are not allowed to be more than a fixed short distance downfield - one yard beyond the line of scrimmage in the NFL - when a forward pass crosses the line. The penalty is "ineligible receiver downfield" (illegal man downfield).

The tension: on an RPO, the line is run-blocking, which means firing out and driving defenders - moving downfield. But if the quarterback then throws a forward pass with linemen too far past the line, it is a penalty.

This single rule shapes the entire concept in three ways:

  • The throw must come out quickly. The quarterback has only the brief window before his linemen drift past the legal limit. That is why RPO throws are short, fast routes rather than developing downfield concepts.
  • The downfield distance is the enforcement battleground. A generous interpretation lets offenses run more aggressive RPOs; a strict one curtails them. The exact downfield distance allowed is the lever rule-makers use to expand or rein in the concept, and it has been a recurring point of officiating emphasis.
  • It distinguishes legal RPOs from illegal ones. A perfectly designed RPO that the quarterback holds a half-second too long becomes a flag - not because the design was illegal, but because the linemen got too far downfield before the ball came out.

So the linemen-downfield rule is not a footnote; it is the boundary condition that defines what an RPO can be. The college game's more permissive downfield allowance is a large part of why RPOs flourished there first and arrived in the NFL in a more compressed, quicker-trigger form.

Why It Stresses Defenses

RPOs are not just clever - they attack structural weaknesses that defenses cannot easily resolve. The pressure they apply is mathematical as much as schematic:

The box-count problem

A defense has a fixed number of players. Loading the box to stop the run leaves windows for the quick pass; dropping to cover the pass leaves the box light. The RPO forces that trade-off to be made on every snap and immediately exploits whichever way it tips.

Conflict defenders

The read defender is asked to do two contradictory jobs - fit the run and cover the pass - and is punished for choosing either. Repeated RPOs wear on these players and make their pre-snap alignment a liability.

Low-risk efficiency

Because the quarterback only throws when the read says the pass is open, RPO throws tend to be high-percentage, low-sack, low-interception plays. They generate steady, on-schedule gains rather than boom-or-bust shots.

Eye discipline tax

Like play-action, RPOs punish defenders who peek at the wrong key. The run action pulls eyes into the backfield exactly when a defender needs to be reading his coverage responsibility.

RPO usage rates and their efficiency vary by team and by season.

The defensive counters mirror the offense's logic: disguise the box count so the quarterback cannot reliably read it pre-snap, drill "pass-key" eye discipline so conflict defenders react to the right thing, and use simulated pressures that change who the conflict defender even is. As with most schematic battles, the result is an equilibrium - RPOs remain efficient, but defenses make the reads harder rather than impossible.

How RPOs Show Up in Play-by-Play

For analysts, RPOs create a genuine measurement headache, because the play is ambiguous by design. The same call can end as a run or a pass depending on a post-snap read, which breaks the clean run/pass split that most metrics assume.

  • The result, not the call, gets recorded. Standard play-by-play logs the outcome - a handoff is a rush, a throw is a pass attempt. An RPO that is handed off looks identical to a designed run in the data; an RPO throw looks like a quick pass. The shared decision behind them is invisible.
  • RPO identification requires charting. To study RPOs as a category, charting services have to tag plays as RPOs from film - watching for the run blocking, the mesh point, and the quarterback's read - because no automatic field in the box score marks them.
  • They distort run/pass tendency stats. A heavy-RPO offense's raw pass rate and pass rate over expected are muddied, because many of its "runs" were one read away from being passes and vice versa. The offense's true pass-or-run identity is blurred.
  • Efficiency reads need the RPO tag. Evaluating whether RPOs help an offense requires the charted flag plus an expected-points framework - comparing RPO plays' EPA against the offense's other plays. Without the tag, the effect is smeared across the rush and pass buckets and disappears.

The practical upshot: when you see RPO usage or efficiency numbers, they came from someone watching the film and tagging plays, not from a clean box-score field. That is worth remembering before treating any RPO statistic as precise - and a reason different sources' RPO counts can diverge.

Caveats and How to Read It

  • RPO is a charting category, not a box-score fact. Counts and rates differ by source because identifying an RPO is a film judgment.
  • The run/pass label is misleading. Don't treat RPO handoffs as pure run-game volume or RPO throws as pure dropbacks - they're a third thing, and tendency metrics that ignore that are distorted.
  • Legality is a live variable. Because the linemen-downfield rule defines the play, officiating emphasis on that rule can expand or shrink what RPOs are practical from season to season.
  • It is not a cure-all. RPOs trade explosiveness for efficiency and quick completions; an offense cannot live on them alone, and disciplined, well-disguised defenses can blunt the reads.
The practical read: treat RPOs as a packaged run-pass concept that punishes a conflict defender, and remember that any RPO statistic depends on someone having charted the play. Judge their value with EPA on charted RPO plays - and read team run/pass tendencies skeptically when an offense runs a lot of them.

The bottom line

The run-pass option packages a run and a pass on the same snap and lets the quarterback decide between them after the ball is snapped by reading one "conflict" defender, taking whatever that defender's reaction leaves open. It stresses defenses by forcing the box-count trade-off on every play and punishing the read defender for committing either way, generating efficient, low-risk gains. The whole concept is bounded by the linemen-downfield rule - offensive linemen may be only a short distance past the line on a forward pass - which forces RPO throws to be quick and short and makes officiating emphasis on that distance the lever that expands or limits the play. In play-by-play data, RPOs hide: they are recorded by their outcome as ordinary runs or passes, so studying them requires charting, which is why RPO usage and efficiency figures vary by source and should be read as estimates rather than exact counts.

Further reading

About the author

The NFL Analytics Editorial Team

Our editorial team researches and writes these explainers from public play-by-play and tracking data, primary sources, and the published methodology behind each metric. Every figure is checked against a verified data pull before publication; we never invent statistics.