Disruption is Production
The debate is over. Disruption is production.
You’ve probably witnessed Twitter debates on this subject but we’re here to end this once and for all.
Yes, sacks are great. 100 percent of sacks result in a negative play for the offense. And that’s obviously the goal of every defense. Sacks can end drives and change momentum. So before we dive in further, understand this: we would rather have a player who made 5 sacks and zero hurries than a player with zero sacks and five hurries. They are not perfectly equal statistics.
However, sacks do not occur at the same rate as hurries. So what we’re going to demonstrate here is that just because a player isn’t generating a high rate of sacks, doesn’t mean he isn’t productive.
CFB Film Room’s analysts have charted roughly 10,000 dropbacks by quarterbacks from the 2015 season. And from these numbers we can prove that disruption is production.
Take a look at the numbers from quarterbacks with time to throw versus their production under pressure. The numbers should speak for themselves:
When a pass-rusher generates a non-sack pressure, the quarterbacks’ stats drop significantly across the board. The odds of a completion drop by ~30 percent, yardage gained drops by ~20 percent. But perhaps most importantly, the odds of generating an interception nearly doubles.
Without pressure, quarterbacks throw an interception once every 43.8 attempts. That number rises to once every 27.9 attempts under pressure.
Even when you extend this to include all pressured dropbacks (because some pressures obviously result in scrambles, rather than pass-attempts) we found that quarterbacks threw an interception once every 32.9 non-sack hurries.
That’s an important statistic to keep in mind, because the top-level pass-rushers in the college game generate far more than 33 non-sack hurries per season. Joey Bosa, for example, despite playing just 11 full games in 2015 generated 48 non-sack hurries. Which means, assuming average quarterback play from his opponents, his non-sack pressures were likely responsible for at least one turnover.
Now the pro-sack crowd is probably thinking to themselves, well, sacks can generate fumbles. Yeah, but not very often. We found that sacks resulted in a lost fumble once ever 25.1 sacks—which means, on average, even your elite pass-rushers at the college level will only generate a turnover on a sack once every other season.
So what’s the breaking point with pressures? How many non-sack pressures do you need to generate to make up for the fact that you didn’t finish them off with a sack?
There isn’t a perfect mathematical answer to that question, but we can get in the ballpark by dumbing it down a little.
The easiest way to set a benchmark is to look at the odds of forcing the offense into a negative play (zero or negative yards gained). A sack is obviously 100 percent. But we also know that non-sack pressures generate a negative play 47.1 percent of the time (note: this is lower than the incompletion rate in the table above because we have to add in the times a quarterback scrambles for positive yards).
So a sack generates a negative play roughly twice as often as a non-sack pressure. That makes the math pretty simple, but it would be underselling the impact of the non-sack pressures.
We’ve already established that due to the rate of non-sack pressures, pass-rushers are far more likely to generate a turnover on the non-sack pressures than with a sack, so this is where the math gets somewhat derailed. Unless we can clearly define the value of a turnover (we can’t) then it’s impossible to put a number on the value of a non-sack pressure in comparison to a sack.
But for the sake of keeping things simply, let’s just stick with our negative play assessment and say a non-sack pressure is worth .5 sacks. How does that change how we view defensive players?
In 2015, we credited Clemson’s Shaq Lawson with 12 sacks and Joey Bosa with 5. That makes Lawson look substantially more productive.
But we credited Bosa with 48 non-sack pressures compared to Lawson’s 34. Conveniently, with our crude assessment that a non-sack pressure is worth half a sack, when you add it all up we get 29 total “negative plays” for both players.
Certainly makes you wonder about how many players we undervalued in the past before advanced stats emerged for football, doesn’t it?