In the wake of the DeMarco Murray signing, we should expect to hear a lot about how he was critical to the Cowboys offensive identity in 2014 (this is certainly the narrative coming out of Philadelphia, where Eagles fans, anxious for some sign of sanity from their front office, are crowing that the Iggles have "torn the heart out of the Cowboys offense"). A key part of the story goes like this: the Cowboys ran Murray on first and second downs so that they could repeatedly get themselves in manageable third down situations.
Indeed, after the week seven win against the Giants, both Tony Romo and Jason Garrett extolled the virtues of the run game in precisely this regard. Here's Romo:
"This is the best we have ever been on third down," Romo said. "That is really changing the game. It is our ability to consistently stay patient and run the ball over and over again. If we continue to be good on third down, we can continue to get a lot of reps on the ground. That helps the process."
"Typically you’re in more manageable third-down situations when you run the ball effectively..."
But it's not only players and coaches who felt this way. Here's an example of this line of thinking from someone I respect:
The key here is the phrase "Conversion rate too critical to W-L." This assumes that a woefully undernourished defense defense was protected by the running game that converted third downs, sustained offense and kept the defense off the field. Indeed, I myself hammered this narrative pretty hard all season, in numerous posts and on our weekly podcasts.
The problem with this story is that it's not really true, or at least it wasn't true for the entirety of the season. Let's start by looking at this game-by-game chart, which tracks Dallas' average yards to go on second and third downs over the course of the season, with third down conversions and conversion percentage thrown in for added measure:
Avg. To Go
Avg. To Go
The first thing I'd like to point out is that the Cowboys 2014 offensive campaign consisted of two distinct seasons. In the first, comprised of games one through seven, they converted third downs at an alarmingly successful rate, never lower than a 50% clip and, in five of the seven games, in excess of 55%. To put that in context, only one team in the last decade - the 2011 Saints - finished with a third down conversion percentage higher that 50%. Dallas was not only converting at a historic rate, but shattering previous highs. That was some pretty special early season sauce.
After the first Giants game, however, the Cowboys never again rose above 50%, and only reached the .500 mark on two occasions. In short, a team that converted first downs at an historic rate in Part I of the 2014 campaign fell back to earth in Part II, converting at a meagre 38% clip that, had it been extended over the course of the season, would have put the Cowboys 24th in the league (about where they finished in 2013, when Dallas converted at a 35% rate, good for 26th in the NFL). Thanks to this moribund second half, the Cowboys finished the season at 47%, which was second in the league behind the Saints.
One of the 2014 season's overriding narratives concerned the Cowboys' ability to sustain offense by running on first and second down to set up highly convertible third downs. But it's clear that, for over half of the season, they weren't much better at converting on third downs than they had been the previous season, when the Dallas offense registered its lowest figure in over a decade. Aside from 2013's ignominious low, the 38.3% conversion figure from games nine through sixteen is the lowest percentage since Tony Romo took over as quarterback in 2006.
Curiously, the chart also shows that there was little to no correlation between the average third down distance and the actual conversion rate in each game. In the Monday nighter against the Redskins, for example, the Cowboys enjoyed their best third down to-go distance of the season, yet suffered one of their worst third down conversion rates. Conversely, Dallas' best conversion rate in 2014 occurred as they faced their second-worst average third down distance. And, finally, there was no correlation between down-and-distance and winning; of the six instances in which the Dallas offense found itself most egregiously "behind the chains" - with second down and eight or more yards to go - five resulted in victory.
It was the combination of these factors - the poor third down conversion rate in games 8-16 and the lack of correlation between to-go distance and conversion rates (and to winning) - that induced me to question the notion of the manageable third down as somehow key to the Cowboys' success in 2014. To figure this out, I want to start by contextualizing the 2014 season by comparing its trends to those developed over the last five years. In a later post or posts, I'll offer some game-by-game trends and take a deeper look at Murray's specific contributions.
Lets kick off this investigation by looking at the Cowboys average yards to go figures going back to 2009, the last time they were a legitimate playoff team:
|Season||Avg. 3rd Down Yards To Go||NFL Ranking||NFL Average|
A couple of things here: first, it's clear that the Cowboys generated a much more manageable third down yards-to-go average last season. The other thing that jumps out is that the worst of these was in 2009, the last season the Cowboys enjoyed a top-five running game. That offense, you may recall, was much more boom-or-bust; they combined explosive afternoons (37 points against Atlanta) with dry outputs (seven-point totals against the Packers and Redskins).
As the example of 2009 suggests, these averages are impacted by penalties and negative plays; a couple of third-and-25s can skew the numbers. In 2009, you may recall, the Cowboys suffered from an inordinate number of "drive killers," particularly penalties and sacks, that surely led to less manageable third down situations. Indeed, as O.C.C. pointed out at the time, every single time the 2009 Cowboys faced more than two third downs, the drive ended without a TD. They were an explosive but inefficient lot.
With that in mind, let's look at the percentage of manageable third downs each of these seasons. For our purposes, I'll consider any third down of three or fewer yards to be "manageable." And, to provide context, we'll look at the percentage of other down-and-distances faced:
Clearly, there was a big spike in the percentage of manageable first downs in 2014, at a nearly seven percent clip. This resulted in a small decline in the percentage of 4-6 yard-to-go third downs and a more significant decline in the percentage of third downs from seven to ten yards in distance. What is curious is that the least manageable category, eleven or more yards, hit a three-year high, which is significant when we look at the conversion percentages for each of these categories over the same time frame:
Looking at this, you'll notice that the Cowboys 2014 conversion percentage in 0-3 and 4-6 yard-to-go situations is comparatively strong, but still in line with their achievements in recent seasons (they were better in 2009 in the zero to three range, and were excellent in 2012 from four to six yards, for example). Where their performance on third down appeared to improve most noticeably was in the less manageable ranges: in the 7-10 range, they were just shy their highest mark, and their conversion percentage in the 11+ range crushes their 2013 performance and registers as a six-year high.
Indeed, the Cowboys' performance in the less manageable situations is a key part of the 2014 story - and they were most proficient at it in games one through seven, when they were 13-27 from seven to ten yards (48.1%) and 8-18 on third and eleven or more (a staggering 44.4%), including conversions of 14, 15, 18 and 20 yards. So, while it's clear that the Cowboys did a better job in 2014 of navigating manageable third downs, a key aspect of their ability to sustain offense was an unearthly performance converting the least desirable situations.
Indeed, as Romo noted in the post-game presser that I referred to at the top of the page:
"I don’t know what the numbers were today, but on third-and-8 or third-and-10, we were converting some of those. Third-and-3 feels the same way as third-and-8 or 9 for us right now."
No wonder, given those conversion percentages on third-and-long.
In the next installment, we'll look more closely at the game-by-game breakdown to ascertain, if we can, just how much the running game contributed to the Cowboys' ability to convert and to sustain.