How defensive packages impacted yards per carry in 2015

Mike Clay analyzes the impact of each defensive personnel package on rushing production.

| 7 months ago
(AP Photo/Charles Rex Arbogast)

(AP Photo/Charles Rex Arbogast)

How defensive packages impacted yards per carry in 2015


When I sat down to write this annual piece last year, I decided to first take a look at the 2013 version to see if it was in any way predictive.

In a nutshell, it was.

DeMarco Murray, Mark Ingram, and LeGarrette Blount were among those who stood out, and the trio went on to break out during the 2014 season. That left me highly intrigued by those who fared well in last year’s study. Sure enough, Lamar Miller, who, like Murray in 2013, paced the NFL in yards per carry against opposing base defenses in 2014, took another big step forward last season. Meanwhile, the likes of Alfred Blue, Knile Davis, Toby Gerhart, Andre Ellington, and Chris Johnson showed poorly in the category and were similarly ineffective in 2015. Injuries annihilated the running back position last year, so we don’t have quite as many success stories, but we do have enough of a sample to believe that this is a worthwhile study.

Of course, if you’re new to PFF (or my work), you’re probably wondering what exactly it is that I’m studying. The idea here is that all runs are not created equal. Bigger/early-down backs tend to run against base defenses, while scat/third-down backs see more work against nickel and dime packages. As the below chart shows, this provides a major advantage for the latter.

# of DBs YPC Att.
1 0.6 330
2 0.8 828
3 2.4 3719
4 4.1 70953
5 4.6 30950
6 5.5 4239
7 7.9 111

The chart, which includes data from 2007 through 2015, show clear correlation between the number of defensive backs on the field and YPC. Note that I only used carries by tailbacks who lined up in the backfield. The sample is 111,130 carries.

Now that we have an “expected” YPC based on defensive personnel, we can apply it to each carry from the 2015 season in order to determine which backs’ actual YPC outperformed their expected by the largest margin. Only backs who carried the ball at least 80 times are included (sample = 53).

Rk Tailback Carries Actual Expected Diff
1 Thomas Rawls 147 5.6 4.2 1.4
2 Karlos Williams 91 5.5 4.2 1.3
3 Spencer Ware 85 5.5 4.2 1.3
4 Doug Martin 288 4.9 4.2 0.7
5 Le’Veon Bell 109 5.0 4.3 0.7
6 Charles Sims 106 5.0 4.3 0.7
7 Todd Gurley 226 4.9 4.2 0.6
8 Ryan Mathews 106 5.1 4.5 0.6
9 Tevin Coleman 86 4.6 4.1 0.5
10 Giovani Bernard 159 4.8 4.3 0.5
11 Mark Ingram 165 4.7 4.2 0.4
12 Darren McFadden 239 4.6 4.2 0.3
13 DeAngelo Williams 199 4.5 4.3 0.3
14 C.J. Anderson 206 4.6 4.4 0.2
15 LeSean McCoy 198 4.5 4.2 0.2

Holy youth movement, Batman. Following a year in which this chart was overloaded with veterans, a bizarre year for injuries at the position allowed many young, talented backs an opportunity to showcase their skills.

The headliner here, of course, is Thomas Rawls. The undrafted rookie stormed up the Seahawks depth chart and proved dominant when working in place of Marshawn Lynch. Lynch, whose effectiveness dropped off quite a bit in 2015, has since retired, which sets up Rawls for feature back duties. Rawls’ 5.8 yards per carry against base defenses was tops in the league among backs who carried the ball more than 45 times. Of his 147 attempts, 67 percent came when the defense had fewer than five defensive backs on the field (14th highest). Obviously super effective as a rookie, slow recovery from a broken ankle is all that could slow Rawls’ emergence into a superstar in 2016.

Doug Martin enjoyed a resurgence in 2015 and this study suggests that he was even better than his raw numbers suggest. Of Martin’s 288 carries, 78 percent came when the defense had fewer than five defensive backs on the field. That’s the league’s fourth-highest mark. Martin averaged 4.8 YPC vs base defenses (10th best) and 5.6 YPC vs. nickel (fourth best). Martin picked a fine season for a career year as he’s now headed to the open market.

If Martin moves on, the Buccaneers won’t have trouble finding his replacement. Charles Sims is far from a household name, but he was one of the league’s most-effective backs in 2015. Unlike Martin, Sims benefited from handling most of his carries against nickel and dime defenses (62 percent – 10th highest). Like Martin, he easily outperformed his expected YPC. Sims averaged a solid 4.4 YPC vs base defenses and 5.8 YPC vs nickel (third best in the league). Sims might be best suited as a committee back, but he’s earned himself a shot at feature back duties.

Todd Gurley averaged a ridiculous 6.8 yards per carry against nickel as a rookie. That was easily tops in the NFL. His 3.8 YPC vs. base is underwhelming, but that figures to increase once the team improves its passing game.

Among our Top 15, we have four rookies and a converted fullback (Ware). Of the other 10, only two finished in the Top 15 in both 2014 and 2015 (Le’Veon Bell and C.J. Anderson). Both Jamaal Charles and Jerick McKinnon would’ve repeated as Top 5 in this category if they met our threshold of 80 carries.

Next, we have the other end of the previous chart. These backs were least productive in the league last season, underperforming their expected YPC marks by the highest margins.

Rk Tailback Carries Actual Expected Diff
53 Andre Williams 87 3.0 4.2 -1.2
52 Danny Woodhead 94 3.4 4.6 -1.2
51 Chris Polk 99 3.4 4.4 -1.0
50 Melvin Gordon 184 3.5 4.5 -1.0
49 Matthew Jones 144 3.4 4.3 -0.9
48 DeMarco Murray 193 3.6 4.5 -0.9
47 Jeremy Langford 145 3.7 4.4 -0.7
46 Joique Bell 88 3.4 4.0 -0.7
45 Darren Sproles 80 4.0 4.6 -0.6
44 Duke Johnson 104 3.6 4.2 -0.6
43 Jeremy Hill 232 3.5 4.1 -0.6
42 Frank Gore 256 3.7 4.3 -0.5
41 Javorius Allen 137 3.8 4.3 -0.5
40 Marshawn Lynch 117 3.7 4.2 -0.5
39 Alfred Morris 213 3.8 4.3 -0.5

As a rookie, Andre Williams ranked sixth on this list. He posted a 3.3 YPC, which was well below his 4.3 expected mark. This year, he was obviously worse, posting an ugly 3.0 YPC. Considering he’s been arguably the worst rusher in the league during the span, it’s fair to wonder how and why Williams has over 300 carries during his first two years in the league. Although Williams is a plodder, Ben McAdoo’s wide-receiver-heavy scheme allowed him to carry the ball with five-plus defensive backs on the field 52 percent of the time. That’s 17th highest among our sample of 53 backs. He averaged 3.7 YPC against base defenses and a horrific 2.8 YPC vs. nickel. It’s hard to imagine Williams making the Giants’ 53-man roster next year.

When one member of a team’s backfield shows up on a list like this, it’s fair to place the blame on that player. When both members of a two-headed attack appear in the Top 4, it’s fair to wonder if the scheme and/or offensive line were to blame. Danny Woodhead (-1.2) and Melvin Gordon (-1.0) both struggled to run the ball last season. Woodhead averaged only 3.4 YPC despite handling 83 percent of his carries against a nickel or dime defense (third highest). He averaged 3.4 YPC against opposing nickel defenses. Gordon, meanwhile, failed to score a touchdown as a rookie, but the problems go deeper. His 2.8 YPC against opposing base defenses was fourth-worst in the league. It’s a bit surprising to see that 69 percent of Gordon’s carries came against five-plus defensive backs, which is seventh-highest in the league. Gordon’s rookie season was certainly disappointing, but there are a few reasons for optimism, one of which is the fact that Woodhead also struggled badly with efficiency.

I haven’t been afraid to throw water on the Matthew Jones and Jeremy Langford hype trains and this study further adds to my arsenal.

Jones put together a few highlight reel plays, but a good chunk of that came on a small sample of receptions. The former Florida Gator was poor as a rusher, averaging 3.0 YPC vs. base defenses (eighth worst) and 3.5 YPC vs. nickel (fifth worst). Jones’ overall YPC was miserable despite the fact that 55 percent of his carries came against nickel defenses (13th highest).

Langford, meanwhile, followed a similar script, but was even worse. Langford showed some flashes when Matt Forte was out of the lineup, but volume, not efficiency, was the driving force. Langford posted a 2.7 YPC against base defenses, which was worst in the NFL. This was a distinction held by Alfred Blue in 2014 and Ray Rice in 2013. Langford saw quite a bit of nickel and actually did okay (4.3 YPC) on those runs. This study, combined with his inability to break tackles, poor blocking, and drop issues, makes Langford’s early-offseason ADP of 25th overall (10th RB) laughable.

Barring major improvements, neither Jones nor Langford are going to be able to hold down feature back duties in Chicago. They are two players to ignore the hype on in 2016 fantasy drafts.

Next, we have the players who faced fewer than five defensive backs on the highest percentage of their 2015 carries.

Rk Player Att <5 >4 <4 Base Nickel Dime
1 Antonio Andrews 143 90% 10% 6% 85% 9% 1%
2 Carlos Hyde 115 86% 14% 3% 83% 13% 1%
3 Tevin Coleman 86 80% 20% 3% 77% 20% 0%
4 Doug Martin 288 78% 22% 1% 76% 22% 0%
5 Tim Hightower 96 76% 24% 1% 75% 24% 0%
6 Jeremy Hill 232 73% 27% 7% 66% 27% 0%
7 Adrian L. Peterson 350 71% 29% 3% 68% 28% 1%
8 Karlos Williams 91 69% 31% 4% 65% 27% 3%
9 Mark Ingram 165 68% 32% 2% 66% 29% 2%
10 Jonathan C. Stewart 267 68% 32% 5% 63% 31% 1%
11 Todd Gurley 226 68% 32% 1% 67% 32% 0%
12 LeSean McCoy 198 67% 33% 1% 66% 31% 2%

Last year, I highlighted the fact that Kyle Shanahan’s offense in Cleveland set his running backs up for inefficiency because they were almost always running against a tougher defensive package. It appears that he has not corrected the issue in Atlanta. Those invested in Tevin Coleman in dynasty leagues should not cut bait just yet. Coleman, as we saw earlier, exceeded his expected YPC by a solid margin. We see here that his 4.6 YPC was actually better when we dig deeper. The rookie faced a base or tougher defense on 80 percent of his carries. His 4.5 YPC against base was 14th best at the position. Devonta Freeman, meanwhile, carried the ball against four or fewer defensive backs on 65 percent of his tries, which was 17th highest. Freeman averaged 3.6 YPC against base (38th) and 4.6 vs. nickel (17th). This suggests Coleman may take on more early-down work in 2016, with Freeman sustaining his passing-down gig.

Defenses clearly didn’t fear Tennessee’s passing game last season, especially when Antonio Andrews was on the field. Andrews faced fewer than five defensive backs on a league-high 90 percent of his carries. Although this certainly impacted his YPC, Andrews’ 3.6 mark was still well below expected (4.0). He’s unlikely to open 2016 as the team’s lead back.

One of my favorite NFL players, Carlos Hyde is probably the best running back in the NFL in terms of shrugging off tacklers. Injuries aside, he’s been terrific thus far in his two-year career, but he clearly had his hands full in 2015. Hyde faced a base defense on 83 percent of his carries, which trailed only Andrews for highest in the league. Hyde managed a solid 4.2 YPC vs. base, but racked up only 15 carries against nickel. Now entering Chip Kelly’s offense, Hyde is guaranteed to see a ton of nickel (more on this later). During his two seasons, he’s racked up 177 yards on 35 carries (5.1 YPC) against nickel defenses.

The next chart shows the players who had the benefit of seeing nickel, dime, or quarter defenses on the highest percentage of their carries.

Rk Player Att <5 >4 <4 Base Nickel Dime
53 Darren Sproles 80 14% 86% 0% 14% 79% 8%
52 Ryan Mathews 106 17% 83% 0% 17% 82% 1%
51 Danny Woodhead 94 17% 83% 0% 17% 74% 9%
50 DeMarco Murray 193 22% 78% 0% 22% 77% 1%
49 James Starks 167 26% 74% 5% 21% 73% 1%
48 Eddie Lacy 211 31% 69% 1% 30% 67% 2%
47 Melvin Gordon 184 31% 69% 1% 30% 64% 5%
46 Rashad Jennings 195 36% 64% 2% 35% 62% 2%
45 Jeremy Langford 145 37% 63% 3% 34% 61% 1%
44 Charles Sims 106 38% 62% 3% 35% 61% 1%
43 C.J. Anderson 206 39% 61% 3% 36% 56% 3%
42 Ameer Abdullah 143 41% 59% 0% 41% 59% 0%

Darren Sproles faced a nickel or dime defense on an NFL-high 86 percent of his carries last season. Ryan Mathews finished second at 83 percent. DeMarco Murray was fourth at 78 percent. This is obviously notable considering all three players worked in Chip Kelly’s offense in Philadelphia. In fact, Eagles’ backs faced five or more defensive backs on 82 percent of carries in 2015, which was easily highest in the NFL (Packers were second at 72 percent) and a massive increase from 43 percent in 2014.

Earlier, I posted the best and worst backs in terms of actual vs. expected YPC. You may have noticed that all three Eagles backs appeared. Mathews (+0.6) took advantage of the “easier” defensive fronts. He averaged 5.9 YPC against nickel, which was second best in the league. Hilariously, he managed only 22 yards on 18 carries against base, although eight of those carries did come inside the opponent’s 10 yard line. To no one’s surprise, Murray (-0.9) struggled badly in Kelly’s offense. His 4.2 YPC on 42 carries against base was fine, but he averaged 4.9 in 2014 and an NFL-high 5.7 in 2013. Of course, Murray’s 3.4 YPC on 149 attempts against nickel was third-worst in the league last season. He sat at 4.9 in 2013 and 4.4 in 2014. Finally, Sproles (-0.6) also took a big step back in 2015. He averaged 4.3 YPC on 63 tries against nickel and posted 19 yards on 11 attempts against base.

What does this all mean? Well, if we needed some evidence that certain players are better in certain schemes, this might be it. Still competent against base (which is where he dominated in Dallas), Murray figures to bounce back in a more conventional offense under Doug Pederson this season. Mathews averaged 4.3 YPC vs. base prior to 2015 and thrived vs. nickel this past season, so he will be right on Murray’s heels. Sproles’ age very well could be the factor here, but he remained an explosive piece of the team’s passing game. Hyde, as noted earlier, will transition from seeing a third linebacker to a third cornerback on a majority of his runs in San Francisco.

We’ve talked a lot about this already, but this next chart shows the backs (min. 40 carries) with the highest YPC against Base last season.

Rk Player Att YPC
1 Le’Veon Bell 45 7.0
2 Thomas Rawls 96 5.8
3 Karlos Williams 59 5.8
4 Spencer Ware 43 5.6
5 DeAngelo Williams 89 5.3
6 Eddie Lacy 63 4.9
7 Giovani Bernard 75 4.9
8 Darren McFadden 149 4.9
9 Adrian L. Peterson 237 4.8
10 Doug Martin 220 4.8
11 Denard Robinson 41 4.7
12 Mark Ingram 109 4.6

The sample here is super small with Bell, but this is further evidence of his dominance. He ranked fourth in this category on 135 tries (4.9 YPC) last season. Eddie Lacy had a rough season, and sees a lot of work against nickel, but how about his 4.9 YPC against base? It’s actually a giant improvement on his 3.5 mark in 2014. Giovani Bernard’s strong play was somehow overshadowed by Jeremy Hill’s struggles, but Bernard was a significantly-more effective runner. His 4.8 YPC was well above expected, and his 4.9 YPC against base defenses was seventh-best in the league. This is notable as he enters a contract year.

Here we have the backs with the poorest 2015 production against Base defenses.

Rk Player Att YPC
43 Jeremy Langford 50 2.7
44 Tre Mason 56 2.7
45 David Cobb 42 2.8
46 Melvin Gordon 56 2.8
47 Shaun Draughn 54 3.0
48 Marshawn Lynch 61 3.0
49 Duke Johnson 66 3.0
50 Matthew Jones 58 3.0
51 Chris Polk 48 3.1
52 Frank Gore 120 3.2
53 Justin Forsett 87 3.2
54 Christine Michael 46 3.3

Had I included players with at least 75 carries (instead of 80) in our earlier actual vs. expected list, Tre Mason would’ve come in dead last. He averaged 2.8 YPC, which was a hefty 1.4 below expected. Like Langford is this year, Mason was generating a ton of fantasy hype before the Rams picked Todd Gurley in the first round of the 2015 draft. Maybe they knew something we didn’t. It’s possible the Bears make a similar move this April.

So far, our focus has been on players. Our final chart splits the data by team.

Tm Att <5 >4 <4 Base Nickel Dime
TEN 315 80% 20% 3% 77% 18% 1%
SF 305 74% 26% 2% 72% 25% 1%
BUF 368 70% 30% 2% 68% 27% 3%
MIN 434 70% 30% 4% 66% 29% 2%
ATL 377 68% 32% 1% 67% 31% 1%
CLV 304 67% 33% 1% 65% 33% 0%
TB 399 66% 34% 2% 64% 33% 1%
SL 358 66% 34% 1% 65% 34% 1%
DAL 369 65% 35% 3% 62% 34% 1%
CAR 400 65% 35% 5% 60% 34% 2%
CIN 396 64% 36% 6% 59% 34% 2%
JAX 294 64% 36% 4% 59% 35% 1%
NO 359 64% 36% 3% 61% 34% 2%
SEA 409 62% 38% 2% 59% 36% 2%
OAK 331 55% 45% 5% 50% 39% 6%
KC 377 55% 45% 5% 50% 40% 5%
NFL Avg 11880 54% 46% 3% 51% 43% 3%
BLT 350 54% 46% 1% 53% 44% 2%
ARZ 440 52% 48% 3% 49% 44% 4%
PIT 379 51% 49% 5% 46% 43% 5%
MIA 289 51% 49% 0% 50% 46% 4%
WAS 411 50% 50% 3% 47% 47% 2%
HST 426 49% 51% 2% 47% 45% 6%
CHI 415 49% 51% 2% 47% 50% 1%
IND 326 48% 52% 4% 44% 49% 3%
DET 296 47% 53% 4% 43% 52% 1%
NYJ 383 45% 55% 3% 42% 47% 8%
NE 349 44% 56% 3% 41% 39% 17%
DEN 460 43% 57% 3% 40% 54% 2%
SD 368 34% 66% 1% 33% 60% 6%
NYG 379 33% 67% 3% 30% 63% 4%
GB 407 28% 72% 2% 26% 70% 2%
PHI 407 18% 82% 0% 18% 79% 2%

This shouldn’t come as a surprise, but this chart inversely correlates well with how often teams have a third wide receiver on the field. The Eagles, Packers, and Giants have been near the top of the league in that category for several years, and the Titans, 49ers, and Bills were on the other end in 2015.

The most interesting team here is the Patriots. Defenses have a fourth cornerback (sixth defensive back) on the field 17 percent of the time when New England is running the ball. This is intriguing considering that the Patriots have a second tight end on the field on 41 percent of their pass plays, which is third-highest in the league. This suggests that New England does a good job disguising run vs. pass despite the fact that they called pass on 86 percent of James White’s snaps and 77 percent of Dion Lewis’ snaps in 2015.

Follow Mike Clay on Twitter: @MikeClayNFL

  • YouBarkIBite

    Ironic that Thomas Rawls is at the top yet Marshawn Lynch was near the bottom, even though they essentially ran behind the same (very poorly graded) offensive line. Even more interesting is that PFF graded Lynch higher this year. I’d be very interested to see someone take a stab at explaining this huge disparity (I’m not saying that as a troll either, just honestly curious).

    My suspicion is that PFF grades of RB are almost exclusively based on YAC and missed tackles (which Lynch is great at) with little consideration for picking the correct hole and getting there fast (which seems to be Rawls strength and is obviously more subjective and time consuming to determine). Lynch had some great runs by showing extreme patience (like the TD run against the 49ers in the NFC championship game in 2013) but he could also get dropped for losses by looking for the big run instead of taking the short gain (see 2014 safety against Denver).