Defensive Smackdown

The statistic Rex Ryan and many Jets fans like to look to in measuring the effectiveness of Jets defenses during Ryan's tenure is total yards allowed.  Under that metric the Ryan defenses have been very good, though declining every year of Ryan's reign.  In 2009 the Jets defense was indisputably the best defense in the NFL.  It was #1 in yards allowed, #1 in points allowed, #1 against the pass.  That was a truly dominant defense.  In every year since, however, the defense has regressed in terms of total yards allowed.  From #1 in 2009, the defense has trended ever lower, to #3 in 2010, #5 in 2011, #8 in 2012 and finally #11 in 2013.   That is not the best of all possible trend lines, but it still reflects a pretty good defense even at its worst.

However, there is more to defense than just not allowing yards.  After all, the team that allows the fewer yards is not declared the winner after 60 minutes of play.   It's points that count, right?  In this metric Ryan's defenses have not fared nearly as well.  From #1 in 2009 the Jets defense rapidly diminished, to #6 in 2010, #20 in 2011, #20 in 2012, and #19 in 2013.  Ugh.  That progression is downright ugly.  But is it a better, more accurate reflection of the Jets' defensive prowess?  I would argue it is not, for one simple reason: turnovers.  Specifically, the Jets have turned the football over at an alarming rate in every year other than 2010 of the Rex era.  Teams with offenses that constantly hand the ball to the defense are obviously going to give up more points than teams that rarely turn the ball over, all other things being equal.  So even though at a very basic level one could argue that all that matters is points allowed, if you really want to get at which teams have the best defenses it would seem necessary to account for how often the offense turns the ball over.

This conundrum might sway you to revert back to yards allowed as the best measure of a defense, since yards allowed is in theory solely affected by defensive prowess, without being distorted by the effects of offensive turnovers. But is there a better way?  After all, it is certainly possible to envision a defense that gives up chunks of yardage between the 20s while doing two things well: creating turnovers and keeping the opposing team out of the end zone.  This in fact is pretty much what a typical Tampa 2 defense is designed to do.  If effective, such a defense could conceivably be the best defense in football without necessarily ranking all that high in terms of yards allowed.  So, how do we get a better measure of defensive prowess, one that not only takes into account the most important factor of all, points allowed, but also accounts for differing levels of difficulty based on how often the offense turns the ball over, thus increasing the odds of the defense giving up points for reasons that are all about the ineptitude of the offense, not the effectiveness of the defense?

Such a metric would have to solve this important puzzle: how many points is a turnover worth?  If we can quantify how much damage an offensive turnover does to defensive points allowed, then we can devise an "adjusted points allowed" metric which would in theory roughly project how many points every defense "should have" allowed, had every offense committed zero turnovers.  By placing every defense on a level playing field in this theoretical zero offensive turnover universe, we can in theory rank each team's true effectiveness in preventing opposing teams from scoring.  This system would have the virtue of largely eliminating the unfortunate distortions caused by offensive turnovers of ranking defenses by the simpler points allowed method.  It would also have the virtue of eliminating the distortions caused by ranking defenses with wildly divergent defensive philosophies by yards, where a system that is designed to give up yards but not points is unfairly penalized in comparison to a defense that may give up few yards but is inefficient at preventing points.  Finally this new metric would indirectly account for the increased defensive prowess of defenses that generate bunches of turnovers, in that the adjusted points allowed metric is a scoring metric, not a yards metric.  As such this metric will reward defenses that may give up a few more yards but are in fact more effective because they generate lots of turnovers and thereby prevent the opposition from  scoring.

This brings us to the all important question for the purposes of this new metric: how many points is a turnover worth?  If we can't answer this question with a reasonable degree of accuracy then the whole exercise is doomed.  Fortunately for us, there have been several studies that have done fairly sophisticated statistical analysis of this question, and have come up with a fairly consistent answer.  It turns out a turnover is worth just about 4 points.  Some studies place that number slightly higher, some slightly lower, but all come up with close to that same 4 point answer.  How very fortunate for us the math nerds have answered this all important question.  If you're interested in the details of how the mathematicians have answered this question, I refer you to the excellent articles found at Football Outsiders,  SportsRatings, and Sportsquant which detail some of the analysis that went into this answer.  Armed with this information we can proceed to build a metric which should be better than either points allowed or yards allowed in measuring the effectiveness of NFL defenses.

Here's how we build our model.  First we rank all NFL teams by points allowed, since the bottom line in the NFL is points.  Then, for every time each NFL team's offense turned the ball over, we subtract 4 points from the defense's points allowed total.  This simple calculation will in theory remove distortions caused by the offensive turnovers of each team and project just the points the defense should be held accountable for if all offenses had been turnover free.  In effect it levels the playing field for the defenses, and we are left with what should be something approximating a "pure" measure of defensive prowess, free from the distortions of offensive turnovers.  Since I have never seen this metric used anywhere before, unless somebody can point me to a prior author, I get to name this metric.  Yay!  Let's call this metric the Defensive Smackdown ratings.  These ratings should be better than either yards allowed or points allowed in measuring the effectiveness of NFL defenses.  So, without further verbosity, I present the Defensive Smackdown ratings for all NFL defenses during the Rex era.

<

Jets

<

<

<

<

There are a few things of note here.  First, it is interesting that according to the Defensive Smackdown ratings, the Jets defense has been good but not really elite in any year since 2009.  In fact the Jets defense, despite a near complete overhaul in personnel, has been surprisingly consistent in its ranking.  From 2010 through 2013 the Jets defense ranked #11, #10, #9 and #14.  For Rex detractors these numbers can be used to perhaps show how after the Jets overload blitzes took the league by storm in 2009, the league quickly caught up and the vaunted Rex defense has been just a little better than average ever since.  On the other hand Rex supporters also can draw something from these numbers, as the Jets defense is the only defense in the entire NFL to never rank in the bottom half of the league over the last five years.

Interestingly the Jets defense in 2013, despite being repeatedly burned deep, would likely have come in at just about the same level as the prior three years if it had only been able to generate turnovers at closer to a league average rate.  As it was the Jets were the second worst team in generating turnovers, and although these charts do not capture the net effect of that deficit, it is not difficult to see how the Jets lack of defensive turnovers would have led to more points for opposing offenses.

It may also be interesting to note that since 2009 the Jets defense has not even been the best in the AFC East, ceding that title to the Miami Dolphins every year since, and in 2013 even the lowly Patriots had a better defense than the Jets.  A big part of that picture is the Jets inability to generate a consistent pass rush since 2009 and a decreasing ability to generate turnovers, particularly since Revis left town.  GGN's own Xes&Os wrote a series of insightful fanposts regarding the Jets defense, the pass rush, and turnovers last year which may provide significant insight into these issues. You can check these out here:  Wrong Kind of Defense?, Rex's Pass Rush,  No Turnovers, Ryan Blitzes.

How about you?  Do you think the Defensive Smackdown ratings present an accurate picture of the Jets defense over the years?  What do you see in the numbers and on the field?  Is there something the Jets should be doing differently?  Do you expect to see major improvements in any areas in 2014 and beyond?  Will the Jets ever return to defensive dominance?   Let us know your thoughts in the comments.