Success Rate: The NFL's Consistency Metric, Down by Down

The Question Success Rate Answers

Most box-score numbers tell you how much an offense gained. Success rate asks a narrower and, in many ways, more useful question: how often did a play do its job? It throws away the size of each gain and keeps only a yes/no verdict - was this play "on schedule" or not. Summed across hundreds of snaps, that simple binary turns into one of the most stable and intuitive efficiency measures in football.

The appeal is that a single 60-yard touchdown can lift a yards-per-play average for an entire game, papering over a dozen stalled drives. Success rate refuses to be fooled by one explosive play, because every snap counts exactly once. That makes it the natural complement to magnitude-based metrics like EPA: one tells you how big, the other tells you how often.

The short version: success rate measures consistency and floor - the share of plays that stay on schedule - while EPA measures magnitude and explosiveness. Read together, they describe completely different things about the same offense.

The Down-by-Down Definition

The most widely used definition of a "successful" play is a yardage threshold that scales with the down. The standard, popularized by Football Outsiders and now baked into public play-by-play tools, is:

1st down: gain at least 50% of the yards needed for a first down
2nd down: gain at least 70% of the yards needed
3rd or 4th down: gain 100% - i.e., convert for the first down (or score)

The logic behind the rising bar is staying ahead of the chains. On 1st-and-10, picking up five yards leaves a very manageable 2nd-and-5, so half the distance counts as a win. By 2nd down you need to have made more progress to keep the next down comfortable, so the bar climbs to 70%. On 3rd and 4th down there is no next play in the series to bail you out - either you convert or the drive is in jeopardy - so nothing short of the full distance counts.

Success Rate = Successful plays ÷ Total plays

A handful of refinements show up in different implementations: plays near the goal line are sometimes judged on whether they score or set up an easy score rather than a strict percentage, and some versions tighten the late-game definition. But the down-by-down 50/70/100 rule is the core that almost everyone shares.

A Worked Example: Same Yards, Different Success

Consider a single drive's worth of plays. The yardage gained can be identical while the success verdicts diverge sharply, because the down-and-distance context is everything.

Situation	Yards needed for "success"	Gain	Verdict
1st & 10	5 (50%)	6	Success
2nd & 4	2.8 (70%)	2	Failure
3rd & 2	2 (100%)	2	Success
1st & 10	5 (50%)	4	Failure

Notice that the 2-yard gain on 2nd-and-4 is a failure (it needed 2.8), while the very same 2-yard gain on 3rd-and-2 is a success (it converted). The yard count is identical; the situation flips the verdict. That context-sensitivity is exactly what raw yardage ignores and success rate captures.

Consistency vs. Explosiveness

The defining feature of success rate is that it is blind to magnitude. A 2-yard conversion on 3rd-and-1 and a 70-yard touchdown both count as exactly one success. This is not a flaw - it is the entire point. Success rate is built to isolate consistency, the trait that explosive-play metrics are built to ignore.

This is why an offense's success rate and its explosiveness can pull apart. Picture two units (figures below are hypothetical, chosen to illustrate the contrast):

Illustrative: consistency vs. explosiveness

Hypothetical figures for illustration only - not any real team's stats.

Offense A - "Boom or bust"

Lots of explosive plays, strong yards per play
But a mediocre success rate (~42%)
Many negative or short-of-schedule plays between the big ones

Offense B - "Stay on schedule"

Few explosive plays, modest yards per play
But a high success rate (~52%)
Rarely off schedule; methodically moves the chains

Both can be good offenses, but they win differently and they are vulnerable differently. Offense A's production is fragile - take away two long touchdowns and its profile collapses. Offense B is dependable but can struggle to score quickly when trailing. Success rate is what separates these two profiles; yards per play alone might rate them similarly.

Because of this, success rate is best read as a measure of an offense's floor rather than its ceiling. A high success rate means the offense is reliably staying out of long-yardage trouble. It says little about how often the unit breaks a game open.

The Relationship to EPA

Success rate and EPA are close cousins, and understanding the link clarifies both. Expected Points Added assigns every play a value: how much it changed the offense's expected points. There is a closely related "EPA-based" definition of success - a play is successful if its EPA is greater than zero, meaning it left the offense better off than before.

EPA-based success: play is "successful" if EPA > 0

This EPA-positive definition and the 50/70/100 yardage definition usually agree, because a play that stays on schedule almost always nudges expected points upward. They are two ways of drawing the same line. The key relationship is this: success rate is EPA with the magnitude stripped out. EPA records both the direction and the size of each play's value; success rate keeps only the direction.

Why analysts pair them: EPA per play tells you how much value an offense generates; success rate tells you how consistently. A team can have strong EPA from a handful of explosive plays while posting a middling success rate - and that gap is itself a finding, flagging a boom-or-bust profile that may be more fragile than its EPA suggests.

Practically, success rate tends to stabilize faster over a season than raw yardage and is harder to distort with one fluky long gain, which makes it a useful "is this real?" sanity check on a hot or cold start. If a team's yards-per-play is soaring but its success rate is ordinary, the production may be riding a few big plays that are unlikely to repeat.

Caveats and How to Read It

Success rate is robust, but it is not the whole story, and treating it as a single grade misses what it is for:

It deliberately ignores big plays. An offense built on explosiveness will be undersold by success rate alone. Always pair it with a magnitude metric.
The thresholds are conventions, not laws. The 50/70/100 rule is a sensible, widely adopted standard, but it is a chosen definition - different sources tweak the goal-line and late-down handling, so two published success rates may not be computed identically.
Game script colors it. A team protecting a big lead may run conservative, low-success plays on purpose; a team trailing may take low-percentage shots. Neutral-situation splits are cleaner for judging true quality.
It is a rate, not a volume. A high success rate on few plays is less meaningful than the same rate over a full season.

The practical read: use success rate to answer "how consistent and on-schedule is this unit?" - then reach for EPA per play or an explosiveness measure to answer "and how much damage does it do when it hits?" Neither question answers the other.

The bottom line

Success rate reduces every play to a single yes/no - did it stay on schedule? - using a threshold that rises with the down: 50% of needed yards on 1st down, 70% on 2nd, and a full conversion on 3rd and 4th. By ignoring magnitude entirely, it isolates consistency and an offense's floor, the exact thing that explosive-play and yardage metrics obscure. It is closely tied to EPA: a play is roughly "successful" when its EPA is positive, so success rate is essentially EPA with the size stripped away. Read it alongside EPA - how often versus how much - and never as a standalone grade, because by design it cannot see the big play.