NFL

# How Does Offensive Tempo Affect Defensive Production in the NFL?

For every action, there must be an equal and opposite reaction; does this mean fast NFL offenses wear out defenses?

There are a lot of actions in the world that don't make sense, decisions that people make that don't seem to hold up under close scrutiny. For example, who in the world ever thought it was a good idea to write a movie about a bus that is rigged to explode should the driver go under 50 miles per hour?

Even worse, who looked that movie script and thought, "The dramatic tension in this script will really be perfectly sold by Keanu Reeves and Sandra Bullock." As if it couldn't get weirder, who was the genius who decided to ignore basic physics when launching the bus over a gap in a freeway?

That's why this is my PSA this week: Speed kills.

This little mantra does not apply only to your movie-going tastes, however; I want to test this little nugget in the football world as well. We've seen Chip Kelly's Eagles offense tear up the field on Sundays the past two years as if he had a cinder block on the gas pedal, and we've discussed here on numberFire how the speed of an offense will impact offensive production.

Now Iâ€™m curious about the other side of the ball. Will reduced time resting on the sidelines adversely affect fast teamsâ€™ defenses? Does quick pace-of-play lower defensive efficiency? If a defense is out on the field for a large volume of plays, will they wear out?

### Pedal to the Metal

To start this study, I compiled data from total NFL defenses from the last five years (2010 to 2014) on how much value they produced on a season-by-season basis.

To assess this, I used numberFireâ€™s signature Net Expected Points (NEP) metric as the foundation of this piece. NEP helps us take the numbers we get from the box score and assign them contextual value so they relate even closer to the game on the field. By adding down-and-distance value, we can see just how much each play and each team as a whole influence the outcome of games. For more info on NEP, check out our glossary.

I paired this offensive value data with pace-of-play data in terms of Defensive Seconds per Play (Sec/Play) -- compiled by Football Outsiders -- and total plays run. By taking the average value for the tempo of each teamâ€™s opposition and comparing it to their Adjusted Defensive NEP (adjusted for strength of opponent) and Adjusted Defensive NEP on a per-play basis, I was able to see if there truly was a relationship between speed of opposing play and defensive value.

The table below shows the Pearsonâ€™s r correlation between each category of defensive pace and value. This is essentially a guideline of whether or not there could be a relationship, rather than an indication of cause-and-effect. Remember, the closer a correlation is to 1, the more likely a relationship is; the closer to zero, the more unrelated the variables are.

D Sec/Play0.2880.289
Total D Plays0.1340.289

Our guideline for understanding r correlation is that anything from 0 to 0.19 correlation -- positive or negative -- is a negligible relationship, .20 to .29 is a weak relationship, 0.30 to 0.39 is moderate, and 0.40 to 0.69 indicates a high relationship between variables.

There is no strong and indicative relationship between any of these variables, but pace-of-play relationship is stronger for defense than it is for offensive speed to production. Since both pace relationships here are positive, what this is saying is that the higher a total is for Defensive Seconds per Play, the higher value the defense is produced; essentially, the slower each play is, the more value the defense creates.

Both pace-of-play relationships are also on the cusp of weak to moderate relationship, which suggests that the way an opposing offense pushes the speed of the game can affect an opposing defense. The strength of the relationship, however, doesnâ€™t appear to be a clear indicator of fact.

With the way that NFL athletes are conditioned these days, and the game demanding stamina and strength from all players, it appears that the idea of running an enemy defense ragged falls a little short.

### Rattle, Rattle, Thunder, Clatter

Just to show this effect in a clearer visual, the table below depicts the top-five of defenses required to play quickly last year. We can see from this chart that there is no exact relationship between defensive ineffectiveness and the speed they are forced to play at, in terms of per-play speed and their rankings in Adjusted NEP and per-play Adjusted NEP.

Sec/Play RankTeamSec/PlayAdj NEP RankPer Play Rank
1stHOU25.611st1st
2ndIND26.2216th16th
3rdGB26.2215th15th
4thCLE26.4619th18th
5thCIN26.5712th12th

If you force a defense to play fast, there is no guarantee that they will struggle. In fact, we can see that each of the top five fastest defensive units ranked in the top two-thirds of the league in Adjusted Defensive NEP last season. Even if we look at the fastest team each of the last five years, only one defensive unit ranked outside the top five in defensive value.

This is possibly due to those teams getting ahead quickly and forcing others to go hurry-up late in the game, but with such a large sample size, there has to be something more to it than that.

### Pop Quiz, Hotshot

I had one other question about how pace-of-play affects defensive units: what about a teamâ€™s own offense? If your own team is forcing you back out onto the field with regularity, because they score so fast, will you and your defensive companions wear down? Or will the desperation of the opposing offense to keep up make it easier for your defense to dominate?

I also correlated teamsâ€™ average Time of Possession (T.O.P.) and each teamâ€™s Offensive Seconds per Play as well, to see how own offenses affect defensive production. If we expect quicker offensive play -- and therefore less Time of Possession -- to cause Adjusted Defensive NEP to suffer, then we will see a meaningful negative relationship between Adjusted Defensive NEP and Offensive Seconds per Play, but a positive relationship between defensive value and Time of Possession. What do we find?