NFL Matchup Analysis: Read Yards per Play in Context

Two offenses can finish a game with 360 yards and leave very different statistical records. One might use 60 plays; another might use 75. Their yardage totals match, but their yards per play are 6.0 and 4.8 respectively.

That simple example explains why NFL matchup analysis should begin with a denominator. Before comparing any two numbers, ask what was counted, how many opportunities produced it and which situations were included. A more detailed statistic is useful only when you understand the question it answers.

Separate volume from efficiency

Volume describes how much activity occurred: plays, drives, attempts or total yards. Efficiency describes output relative to an opportunity: yards per play, points per drive or touchdowns per red-zone trip.

For a basic yards-per-play calculation, divide the relevant net yardage by the corresponding counted offensive plays. Use one provider's definitions consistently. Do not combine a passing-yardage total from one table with a play count from another without checking sack, penalty and play-filter treatment.

Here is an invented comparison:

Offense A generated more yards per counted play in this example. That does not establish which offense is stronger overall. We have not described the opponents, starting field position, turnovers or the situations in which those yards occurred.

Ask what the average is hiding

Averages can summarize very different distributions. Imagine two five-play sequences, both totaling 25 yards. One gains five yards on every play. The other gains 25 on one play and zero on the other four. Both average five yards per play.

To understand the difference, inspect the frequency of productive plays alongside the size of the biggest gains. If a table uses “success rate,” read its definition: a provider may define success using expected points or a down-and-distance threshold. Two identically named columns need not be interchangeable.

Keep the sample visible. A rate based on ten plays should not be described with the same certainty as a much larger sample, even if both are displayed to two decimal places.

Use EPA to add game-state context

Expected points added, or EPA, evaluates how a play changes the modeled scoring outlook rather than treating every yard as equally valuable. It depends on the underlying expected-points model.

For example, nflfastR's documented model uses inputs including down, distance, field position and time remaining. Its methodology describes evaluation on held-out seasons. That makes the model inspectable, but does not make every EPA estimate a fact independent of modeling choices. nflfastR model documentation.

A four-yard gain on third-and-three and a four-yard gain on third-and-ten illustrate why context belongs in the question. The yardage is identical; the situations are not. Use EPA as a complementary view of that context, and retain the provider and filters when comparing teams.

Put opponent and personnel context beside the rate

An offense's recorded efficiency is evidence about the games it played. It is not automatically an opponent-adjusted estimate of what it will do next.

For a useful matchup note, record the sample dates, opponents and relevant personnel availability. If you use an opponent-adjusted metric, identify the methodology rather than assuming the adjustment is identical across services. If no adjustment is available, say so.

The NFL's own tracking program distinguishes measured movement, classifications and model-derived metrics. That separation is useful when reading any dashboard: a measured event and an estimate built from events are different types of evidence. NFL Football Operations: Next Gen Stats.

Read red-zone rates with their counts

A rate of three touchdowns in four trips is 75%. Six touchdowns in ten trips is 60%. Those percentages describe the examples correctly, but their opportunity counts matter. One additional touchdown changes the first sample more sharply.

Check what “red-zone efficiency” means in the table: touchdowns, any score or another definition. Then read the numerator and denominator together. Avoid describing a short run of finishes as a stable team identity without examining a broader sample.

Build a research note before building a conclusion

Use a small comparison sheet with one row per metric: value, sample size, provider, date range, filters and interpretation. Add a final column for what the metric cannot establish.

For yards per play, that last column might read “does not describe consistency or opponent adjustment.” For EPA, it might read “depends on model and included plays.” For red-zone rate, it might read “few opportunities; check definition.”

The Cold Line Builder gives you a place to inspect and revise a matchup's inputs. Keep your reasoning connected to those inputs rather than treating a single leaderboard position as the conclusion. Review The Cold Line's research tools when you need weather or game context alongside the numbers.

Good NFL matchup analysis makes its comparisons reproducible. Another reader should be able to see which sample you used, why the denominator fits and where your interpretation could change.

Why TCL is a premier tool for NFL matchup research

TCL turns the research question into an inspectable workbench. The Builder shows the inputs behind an estimate, lets you examine their contribution and places your resulting line beside the market. With weather and learning tools available in the same platform, TCL is a premier choice for turning NFL statistics into a structured matchup analysis.

Explore the Cold Line Builder.