clock menu more-arrow no yes mobile

Filed under:

Do fast offenses makes bad defenses?

Part Three of the Washington Post's expose on the Redskins disastrous season was released today. It mostly offers Gregg Williams' side of the story, though differs greatly from the other pieces. Les Carpenter writes well and it is a must read.

One part of the article that troubled me most was when it tried to sell our defensive woes on the Al Saunders offense. And I quote:

When the Redskins hired Al Saunders to run their offense last winter, one NFL assistant said Williams was troubled by the move, not from any dislike for Saunders but because he worried that Saunders's frenetic offense, which often produced touchdowns quickly, would put pressure on Williams's defense. With the pace of the game accelerated, the defense would naturally give up more yards and points and its ranking would suffer.

"I can't do this," the assistant said Williams told him. The assistant asked that he not be identified by name because he considers both men to be friends.

Before you start worrying that this means there is a Williams-Saunders rift, the article goes on to express Williams satisfaction with Saunders, his respect for him, as well as the possibility that Saunders is in Washington because of Williams.

The driving force behind this criticism of the defense via offense comes from some guy who I don't particularly care much for:

Cowboys Coach Bill Parcells mentioned as much in a New York Times magazine article in the fall. While watching game tape of the Redskins, he noted how little regard Saunders seemed to have for his defensive coordinator. The NFL assistant who said Williams fretted about the hiring has watched the Redskins this season specifically to see how Williams would deal with the faster pace of the offense. He agreed with Parcells's assessment, adding that it was the biggest reason for Washington's defensive collapse.
Bill Parcells is a defensive minded coach, and a brilliant one. And while his assessment of our offense this year might be true, we'll discuss in a bit why it isn't an accurate criticism of Al Saunders generally. In any event, finally we have:
The same league assistant who spoke of Williams's stubbornness has broken down the Redskins' game tapes and said defensive players weren't getting time to rest because Saunders's offense was not putting together sustained drives, as it did in 2005. Either Washington scored fast or the complex system stalled, pushing the offense off the field after three plays. Neither was conducive to defensive dominance, the assistant said.

"Scoring points in this league is great," the assistant said. "But sometimes you can score too quickly."

Absurd premises; it cannot be great to score points and also not-great to score points. Better stated if points are scored it is best to score them slowly, but most drives in any offense don't have the liberty of deciding how long it takes to score. There are ways to lengthen TOP but these don't necessarily increase the amount of time that defenders rest. Play clock does not distinguish between runs or passes.

But there is a way to loosely examine this hypothesis, which probably has merits. More likely teams that cannot sustain drives put their defenses on the field. I question the scoring aspect, particularly since we weren't a potent scoring offense last year yet struggled despite the fact that we clearly were not "scoring too fast."

One way to measure this is Offensive Time Of Possession which would be as close to a pure measure of how much time an offense is leaving a defense on the field. Or it is at least as pure a measure of that as I am willing to examine.

Question: Does Offensive TOP correlate with the Redskins defense? Over the past 3 years the answer is: Maybe. 2004 (3rd best defense) we had the 8th best TOP. 2005 (9th best defense) we had the 5th best TOP. 2006 (31st best defense) we had the 16th best TOP. When our offensive TOP (and the higher this number, the more rest our defense is getting) is at its lowest, so is our defense.

Elsewhere in the league the correlation is even more remarkable. In 2006: Baltimore 1st in TOP, 1st in Defense. Jacksonville was 2nd in TOP, 2nd in Defense. New England was 6th in TOP, 6th in Defense. In 2005: Jacksonville 6th in TOP, 6th in Defense. Pittsburgh was 8th in TOP, 4th in Defense. In 2004: Pittsburgh is first in TOP, first in Defense. Denver 2nd in TOP, 4th in Defense. All of this should be intuitive as the longer an offense is on the field, the less the opposing offense is on the field (and, consequently, gaining yards).

But the plural of anectode is not data, and for every correlation from above there is a counterexample. In fact the best counterexample would be Al Saunders' Kansas City Offense which finished 3rd, 3rd, and 14th in TOP in '04, '05, and '06 respectively yet finished 31st, 25th, and 16th defensively in that same stretch. So, as KC's TOP decreased, it's defensive production increased.

There are many ways to account for this incongruity: coaching, personnel, etc. Simply stating that team's that score too fast hurt their defense is merely passing the buck. The purpose of offense is to score. The more believable complaint is with teams that cannot sustain drives, IE: teams that do not generate first downs, TOP, or yards. Does this criticism apply to Al Saunders? Recently, no. While Al Saunders runs a passing offense, he's shown a capability of holding on to the ball and running it proficiently. He was top 5 in TOP in both '04 and '05. In 2005 KC had the 6th most rushes per game and the Chiefs had more runs than passes on the season.

In other words, there is an incredibly complex relationship between an offense and a defense that cannot and should not be simplified to the point of saying "Offense X causes Defense Failure Y". A very bad offense (which we were not this year -- 13th overall) might be the difference between a 15th and a 20th ranked D, but it doesn't turn a 9th (2005) ranked unit into the 2nd worst (And no offense is insurmountably bad for a defense, or else we'll have to account for Oakland's 3rd best D despite their 32nd ranked offense.). The more likely culprits are disastrous personnel changes, injuries, and a stale defensive scheme (among other things). And all these factors effect each other in complicated, and frequently immeasurable ways.

The opposing argument to "fast offenses hurt defenses" is that some type of defense hurts an offense. One type of D that could make it more difficult to score is one that gets fewer turnovers than any other defense in NFL history including strike-shortened seasons. A lack of short fields makes it difficult to score (though this offense didn't do itself any favors with a miserable red zone effort).

Gregg Williams understands this, I hope. The article expresses as much:

Williams did not want to discuss the subject, saying he would never criticize another coach publicly. "We have a responsibility to play our defense," he said.

He said he has other beliefs as to why the Redskins' defense didn't work this year.

Such as improved NFC East offenses, injuries, etc. Include failed personnel decisions, to which Williams is accountable as well.

Assaulting the offense for our defensive woes is unaccountably revisionism. This defense simply sucked, and Al Saunders was brought to Washington to score points and has a proven history of doing so. There are legitimate criticisms of Saunders and his time in Washington, but the lack of production by our defense is not one of them. Like so many of our defensive backs this year, on what could have -- should have -- been crucial takeaways, this Defense Simply. Dropped. The. Ball. in 2006.