Inside the Numbers: The 2009 Cowboys Passing Game (Part II)

Ask ten different people about the state of the Cowboys receiving corps these days and you'll likely get eleven different answers. And it would definitely be a mixed bag of answers. Some might focus on the very public struggles of Roy Williams, others on Miles Austin's breakout season, others again might lecture you on Dez Bryant or Golden Tate or Jacoby Ford - or on all three.

The simple answer would be to look at the usual stats. Here's what the raw stats for the four top Cowboys receivers looked like in 2009:

2009	Receptions	Yards	Average	TDs
Miles Austin	81	1,320	16.3	11
Jason Witten	94	1,030	11.0	2
Patrick Crayton	37	622	16.8	5
Roy Williams	38	596	15.7	7

We've long held that the NFL is becoming more and more of a passing league. With a casual glance at the table above, it may not have struck you as particularly noteworthy that the 2009 Cowboys had two 1,000+ yards receivers, but in fact, this is only the fourth time in franchise history that this has happened. The other three seasons were 1979 (Tony Hill, Drew Pearson), 2006 (Owens, Glenn) and 2007 (Owens, Witten).

Were the Cowboys’ receivers individually as good as the overall stats suggest? Does Miles Austin really stand head and shoulders above the rest, is Patrick Crayton only half as good as Austin, how big a disappointment was Roy Williams and whatever happened to Jason Witten's touchdowns? Follow me as we turn the all seeing eye of Expected Points Value (EPV) on our receivers to uncover the truth behind the stats.

Go back to one of my previous posts if you would like a refresher on EPV.

Before digging into the individual receiver stats, please keep in mind that the EPV I'm using in these posts is the value of each play, not the value of each player. As I look at the player stats, what I'm actually doing is adding up the value of each play a player was involved in according to the play-by-play of the NFL game books. This is important because football is a team sport, and the credit or blame for each pass is shared by the receiver, the O-line, the coaching, opposing defenses and many others.

There is one important area in which I deviate from this rule as I look at receivers: Interceptions and turnovers on downs. I gave the full penalty of an interception to the overall passing game numbers, but am leaving it out of for the receivers. That's because you don't always know, on an interception, who the ball was intended for. In the game books, the scorer has to make a judgment call, but when the ball is tipped or overthrown for instance, who really knows where that ball was supposed to go?

More often than not, an interception really is a poor decision by the QB, despite the opinion of QBs that the receiver may have run the wrong route. And there's no way I could 'credit' Witten with that fluke bounce of the ball off his foot into the arms Kenny Phillips in the first Giants game last year.

The overall passing game numbers shake out as follows: The official records show 550 pass attempts. Two spikes, nine interceptions and two passing turnovers on downs - out the window. Leaves us with 537 pass attempts for the analysis, of which five passes have 'n.a.' as a target, because not even the official scorers had the slightest clue who those balls were intended for, and fans in the stands don't count as receivers.

The value of the Dallas Cowboys receivers

My good friend M. Excel and I have charted every single pass play of the 2009 season. The numbers below are the average value of a pass thrown to each Cowboys receiver. Rather than simply comparing quantitative stats like numbers of catches or total yards (not very helpful for comparing players), these figures show the average value that each play involving a Cowboys receiver generated for the team. And everything's factored in: those blown routes by Roy Williams? In there. Miles Austin's team leading 17 first down conversions on third down? Got 'em. The incomplete short lob to Martellus Bennett in the end zone against the Panthers? Look no further.

EPV by Receiver
Player/Position	Pass Att.	EPV/PA	Player/Position	Pass Att.	EPV/PA
All Receivers	537	0.39	Halfbacks/FB	82	0.06
			Marion Barber	35	0.14
Wide Receivers	290	0.54	Tashard Choice	22	0.07
Kevin Ogletree	8	1.25	Felix Jones	22	-0.02
Miles Austin	122	0.66	Deon Anderson	3	-0.33
Patrick Crayton	65	0.62	Tight Ends	160	0.32
Sam Hurd	10	0.56	Jason Witten	121	0.44
Roy Williams	85	0.23	John Phillips	9	0.17
			Martellus Bennett	30	-0.09

 

So what can we deduce from this chart? Miles Austin rocks, Roy Williams is overpaid, Kevin Ogletree should see more playing time, Jason Witten had a disappointing year and that throwing to our running backs is not the best idea? On the surface, yes. But stay with me as we break the numbers down a little further.

The curious case of Jason Witten. Despite posting 1,030 receiving yards ('merely' the 19th best TE season in the Super Bowl era), there seems to be this nagging meme in Cowboys nation that Witten had a sub-par year. Perhaps this is because he had only 2 TD receptions last year, perhaps it's because of the uncharacteristically high seven false start penalties he accrued, perhaps it's something else altogether. And his EVP of 0.44, a lot lower than Austin and Crayton, seems to bear this out.

But the devil is in the details and the details in this case clearly are his lack of touchdowns. If we were to level the playing field between our receivers even further than we're already doing with EVP, and exclude all touchdown receptions from their figures ("No fair", a cry arises from in front of multiple monitors the world over - but it's just to illustrate a point), we end up with the following EVPs for the four top receivers:

1. Crayton: 0.44, 2. Witten: 0.42, 3. Austin: 0.37, 4. Williams: 0.07

We'll return to Crayton's and Williams' numbers a little further down, but it's pretty clear that Witten's numbers suffered from a lack of touchdowns. Why the Cowboys chose to ignore one of their most potent offensive weapons on touchdown plays continues to baffle me.

In an interview with CBSSports.com Witten said that only scoring two touchdowns in 2009 was frustrating, but overall he felt he had a tremendous year, finishing with the second best totals of his career in catches and yards. "I thought I had a good year," Witten said. "I didn't get in the end zone as much as I liked. You want to be that guy that makes those critical plays."

Kevin Ogletree or how to get great EVP numbers. As always with this data beware of the very small sample size behind some of these numbers. But still, one thing is undeniable - the few times Ogletree was called on, he delivered. Of eight passes thrown his way, he caught seven. The seven catches all came in situations with nine or more yards to go. Five of the seven catches resulted in first downs. We'll see if this short but impressive résumé is enough to get him more playing time in 2010.

Roy Williams: Red zone threat? As we've seen above, outside of his seven touchdowns, RWs performance really wasn't anything to write home about (of course, I haven't factored in his downfield blocking, which by now has attained superhero proportions). Now, if RW were a legitimate red zone threat, perhaps we could all learn to live with him, but is he?

2009	Crayton	Williams	Witten	Austin
Pass Attempts	11	15	6	20
Red zone TDs	4	6	2	3
Red Zone EPV	0.75	0.35	0.26	0.18

Again, beware of the low sample sizes, but it looks like Roy Williams' production in the red zone, from a purely numerical point of view, could be taken up by Crayton and Witten without the Cowboys missing a beat.

Are our halfbacks really that bad? The EVP numbers seem to suggest that the Dallas halfbacks add little value to the passing game. And that is true. But it's not because the Dallas halfbacks can't catch the ball, far from it:

	Pass Att.	CMP	CMP %	1st down	1st down%
Barber	35	26	74%	10	29%
Jones	22	19	86%	5	23%
Choice	22	15	68%	6	27%

All three halfbacks have very good to excellent completion percentages. But the passes to the halfbacks result in a first down (or TD) on less than a third of the passes. For comparison, Austin and Crayton are both close to 50 percent. And because EPV places a premium on conversions for a first down, the EPV numbers for our halfbacks are low.

I have no idea how many of the passes to the halfbacks were actually designed plays, my guess would be not that many. Now, if the majority of the passes to the halfbacks were thrown their way because no one else was open, then the EPV number is actually quite ok, because it effectively says that throwing to the halfbacks did not negatively affect the overall Cowboys game, but in fact maintained the value of the drive. Not at all shabby for the fourth or fifth read in the QBs progression.

The amazing Patrick Crayton. There. I said it. Patrick Crayton had an excellent year in terms of EPV, and whatever else you might think, there's little doubt that as a slot receiver he added significant value to the Cowboys games.

Going in to this analysis I had expected Patrick Crayton to come away with so-so numbers. Boy, was I surprised. The plays that involved Patrick Crayton generated nearly as much value for the Cowboys as plays involving Miles Austin. This does not mean that he would be able to replicate these numbers as the number one or two receiver on the team, but as a slot receiver, he did an exceptional, if under-appreciated, job.

In the next and final installment on the passing game we'll break down our top receivers further. Who was Romo's security blanket last season, who was the guy who got it done up the middle, and why all the love for Patrick Crayton?

[Hat tip to Brian Burke at advancednflstats.com for providing the EPV data]