Who Were The Cowboys' Most Productive Defenders In 2014?

When Cowboys linebacker coach Matt Eberflus served as defensive coordinator in Missouri under head coach Gary Pinkel from 2001-2007, the Tigers used a production points system to evaluate and grade their defensive players. A production points system awards a certain number of points for each defensive play like a tackle, a sack, an interception or a defensive score.

Nick Saban at Alabama also uses a production points system (as do many other college programs) which means former Alabama players like Rolando McClain are familiar with it. Saban may have used the system when he was briefly the coach of the Miami Dolphins, and Jason Garrett may have come across the system there.

Whether he did or not ultimately doesn't matter. What matters is that the Cowboys are today using such a system to evaluate their linebackers. In November 2013, Jason Garrett praised Kyle Wilber's performance after the 24-21 win over the Giants, and referenced production points.

"I think he played 25 plays in the game last week against the Giants and had 24 production points," Garrett said. "That doesn't mean anything to you guys but that was a really positive performance. He was around the ball a lot ... He is certainly still learning the position. He doesn't always pull the trigger as quick as he needs to just because he hasn't seen the looks as much, but he is playing better and better."

After the regular-season finale against the Redskins in 2014, Garrett again talked about production points when he pointed out that Bruce Carter set a team record with 56 production points in Washington. Carter had two interceptions in that game, and the coaching film credited him with 10 tackles, two assists, three tackles for loss and two pass deflections. In addition to his two interceptions, Carter had two more drive-ending plays, both tackles.

"Bruce played an outstanding game yesterday," head coach Jason Garrett said on Monday. "Matt Eberflus, our linebackers coach, does a great job with those guys, and one of the ways we evaluate our linebackers is by production points. Bruce set the record in production points (Sunday). Not only with the interceptions he had, but a number of tackles. He really showed up throughout the game. He really played an outstanding game for us."

Production point systems are usually unique to the team using them, as each coach will tweak the point totals to his preferences, his scheme, or his points of emphasis. So we do not know exactly how the Cowboys are allocating points in their system. But a little digging in the interwebs unearthed a "Production points system" used in youth football that uses the following scoring values:

Defensive Touchdown = 10 points Forced Fumble = 5 points Fumble Recovery = 5 points Interception = 5 points Sack = 5 points QB Hit = 5 points Tackle for a Loss = 3 points Solo Tackle = 3 points Assisted Tackle = 3 points Pass Deflection = 2 points Missed Tackle = -1 point

Using the point system above, Wilber would get exactly 24 points for his four solo tackles, three assists and one TFL (though we'd be double-counting the TFL). Carter would get 53 points for his effort with the point system above, three points shy of the total points the Cowboys awarded for his performance. 

Those two point totals are close enough to the actual Cowboys total to make me wonder how the other defenders would look using this particular production point system.

At the danger of repeating myself: In no way am I inferring that the Cowboys are using the exact same production points system. And that both Wilber's and Carter's score come back so close to the Cowboys' total may be pure coincidence. But I was intrigued enough by this coincidence to calculate the production points for all Cowboys defenders based on the stats we have available.

We know from previous Cowboys statements that they measure the effectiveness of their defensive personnel by how productive each player is per snap played, and Garrett reiterated that when he talked about Carter.

"It’s all relative to the number of plays that you play," Garrett explained. "You have positive production points and negative points that take away from the total. At the end of the day, 56 is pretty darn good. Again, it was much more than just the interceptions – those were huge plays for us in the red zone and critical moments in the game. He just made a number of tackles and really showed up as both a run defender and pass defender."

If you combine the production points with the total snaps played, you'll get a metric which we'll call Production Points per Snap played, or PPS.

A few notes on the data we'll use: Snap count and missed tackles are taken from Pro Football Focus. All other stats are official NFL stats, with QB Hits manually compiled from the NFL game books.

I'm also going to slightly change the youth football scoring system we looked at above. It doesn't make sense that a QB Hit counts the same as a sack, so I'm reducing the points for a QB Hit to three instead of five. Also, because football stats in general count a tackle for loss as both a tackle and a tackle for loss, a TFL would actually count six points in the production points system above, versus the five points for a sack. That doesn't make sense either, so I'm allocating four points for a TFL and won't double count it as a tackle. The impact of these two changes on the final numbers for each player are marginal, but they are significant for my peace of mind.

Finally, we'll only be looking at Cowboys defenders who played at least 100 snaps over the 16 regular season games. With all of that out of the way, here is the data for the linebackers:

Player	Snaps	Tackles				Pass Rush		Pass Def			Fumbles			Prod. Pts	PPS
		Solo	AST	TFL	MT	Sacks	QB Hits	PD	INT	TD	FF	REC	TD
Justin Durant	336	31	18	3				4	1		2	1		178	0.53
Bruce Carter	527	48	20	4	13	1	3	8	5	1				260	0.49
Kyle Wilber	216	16	11		3	1.5	3	2				1		103.5	0.48
Anthony Hitchens	541	59	16	3	7		2	4	1					240	0.44
Rolando McClain	654	65	16	10	11	1	3	3	2		1			277	0.42

Justin Durant's season came to an abrupt end when he tore his right biceps against the Redskins in Week 8. Until then he had been a key element in a defense that had helped the Cowboys get off to a 6-1 start. In the week leading up to the Redskins game, DC Rod Marinelli had heaped praise on Durant:

"He’s a terrific linebacker. Everybody is seeing it right now," Cowboys defensive coordinator Rod Marinelli said. "He is our captain. He’s a heck of a ball player. He’s aware, heady, fits our system to a T."

And the numbers bear this out. The most productive Cowboys defender on a play-by-play basis is Justin Durant. As a comparison, Sean Lee had 0.48 PPS in 2013, so Durant's number is pretty good, even if it was achieved on a relatively low snap count. Perhaps even more surprising is seeing where Bruce Carter is ranked in this metric. After a lackluster 2013 season where he only recorded a 0.35 PPS, 2014 was a good year overall for Carter, even if the season did feature a few up and down games for him. Imagine if the guy could put it together for 16 full games.

Rolando McClain doesn't come off favorably in this metric, but keep in mind that McClain started the season slowly. He was dealing with various bumps and bruises as he got back into playing shape, and then played through a knee injury at the end of the season along with concussions. In between, McClain was about as dominant as a linebacker can be.

On to the defensive line

Player	Snaps	Tackles				Pass Rush		Pass Def			Fumbles			Prod. Pts	PPS
		Solo	AST	TFL	MT	Sacks	QB Hits	PD	INT	TD	FF	REC	TD
Jack Crawford	143	6	2	1	1	2	1				1			42	0.29
Anthony Spencer	384	12	10	2		0.5	4	1			1	1	1	104.5	0.27
George Selvie	515	19	11	2	6	3	8	1			1	1		137	0.27
Henry Melton	433	11	4	2		5	7	4				2		111	0.26
Jeremy Mincey	724	19	17	4	6	6	13				1	1		185	0.26
Tyrone Crawford	643	22	11	6	4	3	13	1			1			162	0.25
Terrell McClain	329	14	5	3		1	2				1			76	0.23
Nick Hayden	585	17	21	4	1		3	1						128	0.22
DeMarcus Lawrence	223	6	3	1	1									27	0.12

As football fans, many of us are completely fixated on sacks as the ultimate - and often only - metric by which to measure a pass rusher. But that can become problematic when you use a dozen or so individual plays to evaluate a guy who played 700 snaps in a season. There's no denying that the Cowboys defensive line came up way short in terms of sacks, but looking at the production score suggests that there may be more ways to look at defensive linemen than just sacks.

Aided by his low snap count and a handful of splash plays, Jack Crawford ends up at the top of this list. Whether his level of play is sustainable with more snaps remains to be seen, but at least the Cowboys still have a chance to find out as he's under contract for 2015.

Beyond that, the edge rushers predictably rank higher than the inside guys as they generally get more chances to make plays. Having said that, it's disappointing to see that the edge rushers weren't able to get much separation from the defensive tackles in terms of PPS, and it's certainly something the Cowboys need to address this offseason.

We'll give DeMarcus Lawrence a Mulligan for his rookie season. With his low snap count, a few positive plays would have had a significant impact on his PPS score. In the two playoff games, Lawrence played 56 snaps and recorded 3 tackles, 2 sacks, 2 QB Hits, 1 forced fumble and 2 fumble recoveries. That adds up to a PPS of 0.71, and if that's going to be his level of production in 2015, everything's going to be just hunky dory.

But the biggest story here is one that isn't immediately obvious. in 2013, DeMarcus Ware (0.31 PPS), Jason Hatcher (0.31), George Selvie (0.28) and Everette Brown (0.28 on only 159 snaps) were at the top of the table for defensive linemen. Every other lineman was at 0.20 or lower. This year, every single lineman with the exception of the rookie (whom I've already made excuses for) is above that threshold. The D-line may not have been as good at the top, but it was much better in depth.

On to the secondary:

Player	Snaps	Tackles				Pass Rush		Pass Def			Fumbles			Prod. Pts	PPS
		Solo	AST	TFL	MT	Sacks	QB Hits	PD	INT	TD	FF	REC	TD
Jeff Heath	130	16	1		3									48	0.37
Barry Church	914	77	20	2	15			6	2		1	1		310	0.34
J.J. Wilcox	997	62	12		15		1	5	3		1	2		250	0.25
Sterling Moore	752	35	11	1	4			13			1	2		176	0.23
Orlando Scandrick	885	50	5	2	9	1	1	9	2		2			204	0.23
Morris Claiborne	151	5	2		2		1	2	1					31	0.21
Brandon Carr	1028	45	9	1	15	1	3	8				1		183	0.18

This metric is probably not particularly suited to assess defensive backs. As a rule, the DBs play a little further away from the ball, so it's no surprise to see lower overall production scores here, especially for the corners.

Barry Church played a lot closer to the line of scrimmage than any other DB did, which explains why he is ranked so high. Heath, working off a small snap, count tops the list and Wilcox shows up with a lower number in part because he was lined up as the free safety more often. Also, note that this metric is biased towards the stats that show a player making a play, but it doesn't include all the mistakes a defensive player made, so it is what it is.

Another interesting aspect as you look at the three different units is how they progressed from the previous season. Last year at around this time, we looked in detail at the Production Points for the 2013 season, and here's how this year's numbers compare to the 2013 numbers at a unit level:

	2013 PPS	2014 PPS
Linebackers	0.41	0.47
Defensive Line	0.24	0.24
Secondary	0.27	0.25

Despite the absence of Sean Lee, the linebacking unit improved significantly in terms of production, and when you look at the overall defensive turnaround in 2014, the linebackers had a much bigger role in that than many people realize.

The defensive line played with better depth, even if they had to make do with lesser talent. But in terms of production, the secondary regressed to a point where it is almost the worst unit on the defense.

We've established that the Cowboys use a similar metric as the PPS as one part of their player evaluations. We also know that there is much more that goes into player evaluations than a simple youth football metric like the one I used today. But there is a natural limit to how accurately we can assess player performance without access to the team data on performance, assignments, technique and more.

Yet even with those limitations, in many cases the data above can be a pretty telling indicator of a player's performance, even with sample size issues and the absence of negative plays to contend with.