Last July, I wrote a post called “Why the running game is THE KEY to success for the Dallas Cowboys.” The analysis compared the 2013 and 2014 Cowboys, which essentially had the same roster on offense, and showed how the Cowboys’ commitment to the running game in 2014 made all the difference.
...success for the Dallas Cowboys comes when they have a powerful running game, and ONLY when they have such a running game.
To illustrate this point, we have a fantastic comparison at hand -- the 2014 versus the 2013 seasons.
One main reason to make this comparison is because the rosters on offense were nearly identical. The only starter who was added in 2014 was Zack Martin at RG. Otherwise, the offense was almost identical. Smith, Leary, Frederick, and Free across the line, Romo at QB, Dez, Williams, Beasley, and Harris at WR, Witten, Hanna, and Escobar at TE, and Murray, Randle, and Dunbar at RB. Clutts was a late season addition at FB in 2013 and carried over to 2014. The 5th WR was different, but made no significant difference. Parnell came in for Free in 2014 once the latter got hurt and did a bang up job.
The biggest difference is how the offense carried itself in 2014. IT RAN THE BALL.This was mostly due to a commitment by the coaching staff, and having a new offensive coordinator in Scott Linehan that Jason Garrett could trust. It certainly wasn't based off a major difference in personnel.
(Quick aside -- Yes Zack Martin was a Pro Bowl guard as a rookie. He received a run blocking grade of +4.8 in 1076 snaps from PFF. Mackenzy Bernardeau received a run blocking grade of +8.7 in 676 snaps in 2013 from PFF. So Martin clearly wasn't the difference in the Dallas run game.)
Certainly, the results in 2014 were far better.
2014 W-L record = 12-4, 1st in NFC East, 1 playoff win
2013 W-L record = 8-8, 2nd in NFC East, out of playoffs
The 2014 team was an astounding 8-0 on the road.
The 2013 team was 3-5 on the road.
After the introduction above, I inserted a series of tables to show how the running game made all the difference between 2013 and 2014. In this article, I’m going to update the data with 2015 numbers, and project 2016 numbers from seven games to sixteen.
Let’s look at the running stat differences over the four years.
|Total rush yards||1504||24||2354||2||1890||9||2638||1|
|Yards per attempt||4.4||8||4.6||3||4.6||6||4.9||4|
|Rush yards per game||94||24||147.1||2||118.1||9||164.9||1|
(Note, 2016 yards per game and per attempt are actual, but total yards, total attempts, and touchdowns are projected from seven games to 16).
As I wrote then about the differences between 2013 and 2014:
The Cowboys ran the ball 172 more times in 16 games, or almost 11 extra carries per game. The yards per game increased by 53.1, and the yards on the season increased by 850. They didn't run more because their yards per carry was suddenly better. 4.6 ypc versus 4.5 ypc is an improvement of about 2%. They ran more because they simply called more running plays.
In 2015, the running game was the same in yards per carry, but the attempts dropped by 100 compared to 2014. This was mostly because the Cowboys’ quarterbacking was so bad (see the quarterback stats at the bottom of this article) that the team fell behind and attempted more passes.
In 2016, the rushing attempts are on pace to exceed the 2014 numbers by 31, or 2 rushes per game. This is in large part because Dak Prescott has already rushed the ball 27 times (or roughly four times per game) for 105 yards. (That projects to 61 carries for 240 yards if he plays the whole year, which would be more attempts than Joseph Randle, the #2 rusher for Dallas in 2014.)
The 2016 Cowboys are rushing for 17.8 yards per game higher than in 2014, because of the higher yards per attempt (4.9 v. 4.6). But the biggest difference is in the touchdowns scored - 12 already in 2016 versus 16 total in 2014. That projects to 27 rushing TDs. Here, Dak Prescott’s four TDs make a big difference.
How does this translate to scoring?
|Points per game||27.4||5||29.2||5||17.2||31||26.9||7|
|Points per drive||2.18||4||2.49||2||1.5||29||2.63||2|
(Note: 2016 total points are projected, the other numbers are actual.)
Here’s what I said about 2014 versus 2013.
The overall points increase was relatively modest -- 28 extra points, or 1.8 ppg. This is a 6% improvement. This kept them 5th in the NFL both years. But they were a more efficient offense, ranking 2nd to Green Bay in points per drive, a 14% improvement over 2013. And they scored 9% more often on their drives.
In 2015, the bottom dropped out of the offense, with scoring down 12 points per game from 2014. Again, this was mostly on the quarterbacks, but the running game was also off, as rushing TDs dropped by 50% from 2014, and rushing yards were off by 29 per game, or 20%. The 2015 offense was also very poor in points per drive, at 1.5, a decline of 40% from the 2.49 points per drive of 2014. In 2015, the Cowboys scored on only 31.4% of drives, 28% below the 2014 rate.
In 2016, the total points and points per game are not yet back to 2014 levels (29.2 in 2014 v. 26.9 in 2016). Some of this could be related to turnovers and short fields, as the 2014 squad got 31 turnovers, while this year they are projected to get 23. The quarterback play in 2014 was also stronger over the season than it has been so far this year (with Romo having a fabulous second half). (See the table below.) Yet the 2016 offense has been more efficient than 2014’s offense was, with 2.63 points per drive (versus 2.49 in 2014), and scoring on 48.6% of drives (versus 43.6% in 2016). This could stem in part from the better rushing attack, as Zeke is leading the NFL in success rate per run at 57%. DeMarco Murray was also excellent here, with a 54% success rate, so that can’t be the full reason. Could it stem from the fewer turnovers by Dallas? Dak has two picks to this point, while Romo had six.
How does this help the defense?
|Time of possession||48.33%||26||54.69%||1||51.88%||13||55.24%||1|
|Points per game||27||26||22||15||23.4||16||18.6||7|
(Note: 2016 points per game and time of possession are actual; defensive plays, points allowed and turnover differential are projected.)
Here’s what I said about 2013 versus 2014.
The Dallas D was on the field for almost 11% fewer plays. As a result, points against Dallas declined by 80, or 5 ppg, which was much greater than 28 extra points, or 1.8 ppg gain on offense. And this difference cannot be explained by turnover differential, which often is a HUGE factor, because it was identical in 2013 and 2014.
In 2015, the time of possession suffered, but the overall number of plays on defense wasn’t much higher than 2014 — 997 versus 978, or less than 2 plays per game. The points allowed were slightly higher (374 versus 352). This actually pretty amazing given how atrocious the 2015 Cowboys were at generating turnovers. They went from plus 6 to minus 22. Indeed, it’s pretty clear the 2015 defense was more stingy than the 2014 defense because it had to stand up despite an inept offense and a horrific turnover differential.
So far in 2016, the defense has taken a great leap forward. Time of possession on offense has the Cowboys back on top of the NFL, at 55.24%. Turnover differential is up by a lot from 2015, and even a little from 2014 (mostly because of fewer Cowboys turnovers, rather than more forced turnovers). Points allowed per game are down 3.4 from 2014, and 4.8 from 2015. This is not a factor of the offense controlling the ball, as defensive plays are actually projected to be higher than the last two years. The defense is just better.
The running game makes the quarterback better
This point was easy to show from 2013 to 2014 because Tony Romo was the quarterback both years, and he went from being a top-10 quarterback to being the most efficient quarterback in the NFL with the running game he had in 2014.
(Note: All numbers are actual. There is no QBR number for 2015 because no aggregate number was made for the four quarterbacks Dallas used last year.)
In many ways, this the most important table in this article, because it shows how the running game and quarterback play are so intertwined for the Cowboys. Take a top-10 quarterback like Tony Romo and give him a running game that forces defenses to load the box, which opens up more one-on-one opportunities in the passing game, and his efficiency will go through the roof.
But not just any quarterback or running back will do. If you are trotting out Brandon Weeden, Matt Cassell, and Kellen Moore, instead of Tony Romo or Dak Prescott, it doesn’t work. Same is true with Joseph Randle and Darren McFadden as lead rushers instead of DeMarco Murray or Ezekiel Elliott.
Dak Prescott’s 2016 numbers are not yet what Tony Romo achieved in 2014. He’s run very close over seven games, but from this point forward in 2014, Romo only had one bad game and a lot of tremendous ones. Dak might improve too, but we’ll have to see.
What’s pretty clear, though, is that Ezekiel Elliott and the Dallas rushing game is helping Dak Prescott have the start he’s having, just like DeMarco Murray helped Tony Romo in 2014. We don’t have the same direct evidence we had between 2013 and 2014 where it was the same quarterback both years and more direct comparisons could be made. But we can look at the weekly output of Dak and Zeke and see how Dak’s worst game was also Zeke’s worst game - the opener against the Giants. Tiny sample size? Yes. Still, it seems apparent that the rushing game is the engine driving the Dallas offense.
What’s less clear is how much it is helping the defense, as most of the improvement on defense has to be attributed to the defenders themselves.
Let me conclude by tweaking the conclusion I drew last year.
Better running game and high commitment to running game has led Dallas to:
- Vastly more efficient QB
- More efficient scoring offense
- Higher time of possession
- More wins overall.