Sympathy for the Devils -- Can Roy E. and T.O. Coexist?

Tony Romo must sometimes feel like Larry from the movie Animal House.  You can just see him in the huddle, with one devil, let's call him T.O., on one shoulder, whispering temptations in his ear.  At the same time, you can see another devil, let's call him Roy E. doing the same:

T.O.:  Throw it!  Throw it to me!  Throw that cross.  Throw that fly!

Roy E.:  Tony, I'm surprised at you!  Throw it at me!

T.O.:  Don't listen to that jackoff!  Look at these hands!  You'll never get a better target!

Roy E.:  If you ignore this magnificent pass catcher, you'll despise yourself forever!

T.O.:  What?  You're throwing to the tight end again!?!

Roy E and T.O.:  You homo!

Both of these guys are notorious for calling out their QBs and OCs, on the sidelines and in the press, in the search for more attempts.  It's the sameness of their selfishness -- and that's not necessarily a bad thing in a receiver, mind you -- that probably has Jerry Jones in a similar Larry-like pickle this offseason.  You can just as easily envision Jerrah with an angel on one shoulder -- let's call him Stephen -- and a second on his other shoulder.  Let's call that one Jason:

Stephen:  Cut T.O., Jerry!  Do it!

Jason:  You know you want to!

Stephen:  You'll never get a better chance!

Jason:  If you let that T.O. get his talons deeper into your poor, helpless QB... you'll despise yourself forever!

There are all types of angels and devils hovering about Valley Ranch these days, and that's before we get to Ed Werder and ESPN.  The question is whether Jerry heeds any of these voices and makes a significant wide receiver move? 

The solution is not as easy as it might seem.  Both Terrell Owens and Roy E. Williams have lots of qualities to recommend them.  Both also are football devils;  they have foibles that can wreck coaching careers. 

Williams, nonetheless, is a lock to remain.  Dallas forfeited three draft picks, one of them the 20th overall selection, to get him.  Jones gave him a gaudy long-term deal.  He's set. 

He also proved to be a big disappointment in '08.  Williams averaged a weak 10.4 yards per reception in a Cowboys uniform.  That slots him behind every other Cowboys receiver or tight end who had significant passing attempts:

• Miles Austin -- 21.1 YPA
• Terrell Owens -- 15.2 YPA
• Martellus Bennett -- 14.2 YPA
• Patrick Crayton -- 14.1 YPA
• Jason Witten -- 11.2 YPA
• Roy E. Williams -- 10.4 YPA

I'm willing to give Roy E. a mulligan on '08.  It's difficult for a player switching teams in mid-season to learn and thrive quickly in a new system.  He also had to split time between quarterbacks -- his first three Cowboys starts coincided with Brad Johnson's three woeful starts.  Williams had three respectable games when he played with the pinky-splinted Romo, catching 8 passes for 123 yards, an Owens-like 15.4 YPA.

Then, just as Romo's splint came off, Williams suffered a plantar fascia injury to a foot.  These are very painful and can affect players who rely on explosiveness to do their jobs.  Look at tape from the Steelers game or any other December game.  Williams could not separate from corners and burn up field.  Opponents were content to stick their second corner on him and forget about Roy E.  His only catches came on slants and hitch routes, where he could use his size to box out corners. 

They're tiny sample sizes, but look at the pre-plantar Roy E. and the nail-in-the-heel Roy E.:

• November:  8 catches, 123 yards, 15.4 YPA
• December:  8 catches, 37 yards, 4.6 YPA

The first line typifies a quality NFL receiver.  The second line is what you expect from Lousaka Polite.

The bigger question, given all the hubbub surrounding T.O., is whether Dallas got Owens' successor or a slightly less expensive reprise of Joey Galloway's deal?

To answer that question, we need to assess Owens' and Williams' positions on the NFL's receiving totem. 

What is a Number One?

You see the terms thrown around a lot.  "This guy is his team's number one receiver."  "This guy is a number two."  Since the Cowboys paid a steep price for Williams, I first want to determine if they got a player with the skills to be THE primary target.

To guage that, I need to define what makes a number one.  I'm going beyond a receiver's position in his team's passing chart.  For me, a number one is a receiver who can beat top cornerbacks consistently.  He's a guy who can produce even though the opposition is scheming to take him out of the game. 

A number one receiver is consistent Pro Bowl caliber.  I'm not interested in the Hall of the Very Good Here.  I'm talking about the pantheon.

For a clearer statistical measure of a number one, I'm turning to Scientific Football's measure of quality corners and receivers.  KC Joyner now has five full seasons of passing stats to lean upon.  Joyner used to supply a simple core stat, YPA, showing how many times a receiver was called up and how productive he was on those throws.

Last year, Joyner added a second, color-coded metric to his stats.  He had long noted that passing attacks rely on matchups, on pitting your strongest player or players against an opponent's weakest ones.  Last year, he began charting how receivers performed against three levels of cornerbacks -- reds, yellows and greens.

7.0 YPA has been the median for corners and receivers alike, so Joyner used this as his baseline.  Cornerbacks who average 7.0 or less on passes thrown their way make up the red level.  Cornerbacks with YPAs between 7.0 and 9.0 are yellow level, the average middle.  Corners with YPAs over 9.0 are green.  These are the marginal players, the raw rookies or the career 3rd and 4th CBs. 

I'm flipping the numbers for wide receivers.  A WR gets "number one" status in my system if he meets two criteria:

1.  The WR averages more than 7.0 YPA against red-level competition.  This should be obvious.  If red-level CBs hold the opposition to under 7.0 YPA, a WR should be top this figure;

2.  The WR has an overall YPA of 8.5 or greater.  WRs don't always face top corners.  There simply are not that many red-level CBs in any given year.  In '07, only 31 reached that benchmark.  When I averaged CB stats over the four years '04 through '07, I found that only fourteen NFL CBs carried averages of 7.0 or lower.

For this reason, quality receiver should have overall YPAs significantly higher than 7.0, since they're going to face yellow and green-level CBs a lot.  If they can't beat the Anthony Henrys and Jacques Reeves of the world, they're not elite.

My filters produced similar small numbers for WR.  In 2007 33 WRs had overall YPAs over 8.5.  Twenty one of them had YPAs versus red corners of 7.0 or greater.  Only ten receivers were on both lists, though several other big names just missed:

1. Reggie Wayne
2. Terrell Owens
3. Larry Fitzgerald
4. Roddy White
5. Wes Welker
6. David Patten
7. Braylon Edwards
8. Marques Colston
9. Deion Branch
10. Brandon Stokely

And the near misses:

1. Andre Johnson
2. Dontae Stallworth
3. Jerricho Cotchery
4. Randy Moss
5. Chad Johnson

Understand that no single year is definitive.  There are lots of variables that can raise or lower a receiver's numbers.  He could be injured.  His QB could be injured.  His supporting skill position players may be poor or injured.  His offensive line may be weak, etc. etc.  I found extreme fluctuation among the CBs when I created four-year averages and expect the same would be true of receivers.  Even some of the best ones would have a down year in one of the categories.  Take Steve Smith in the '07 totals.  His overall YPA was uncharacteristically low, but he abused top level CBs.  He bounced back with a very strong overall '08.

Because I only have a small body of data to analyze, I'm using these numbers to establish capability:  high numbers in both categories, especially against top CBs, tells me a receiver had the talent to pace his passing attack. 

The '07 numbers are troubling, because they don't cast Roy E. Williams in the best light.  Here are two sets of '07 stat lines.  See if you can tell who they profile:

Table I

Table II

All five players have similar raw YPAs but the first two were outstanding against big time corners.  Their performance remained constant regardless of the opponent or scheme.  The other three had noticeable drop-offs when they got extra attention.  In every case their averages dropped at least 40%.  Player Es' performance dropped out of sight.

Player A is Terrell Owens.  Player B is Larry Fitzgerald.  They and Reggie Wayne had the most impressive overall lines from the year before last.  And the last three?

Player C is Anquan Boldin.  Note that his raw YPA is identical to his teammate's, but there's no comparison when they faced elite cover men. 

Player D is Roy E. Williams.  From a bargain standpoint, it seems that it doesn't make that much difference whether Dallas acquired one or acquired the other.  They seem to be the same player.  The downside to that comparison is that neither had a number one's line, at least not in '07.  Which brings me to Player E.  That's Patrick Crayton's '07 line.  

We can see the benefit Crayton derived playing with an elite receiver and tight end.  Conversely, we can see what happened when he faced those rare secondaries who had a quality second corner.  Taken as a group, I have to ask, would Dallas have significantly improved itself with either Williams or Boldin taking Crayton's place?

Since the question is no longer a hypothetical, I must ask, is Roy Williams capable of performing at Owens' '07 level?

Devil or Angel, I Can't Make Up My Mind

"If there is one player I believe is going to have a breakout year in 2007, it's Roy Williams.  Mind you, Williams made a tremendous amount of progress raising his YPA from 7.6 in 2005 to 9.0 in 2006.  [Mike] Martz said the biggest difference in Williams last year was that he finally got it right more than one play in a row...

It wasn't just that Williams got his head in the game either.  Martz emphasizes conditioning because he tends to find a weak spot in a defense and continually go after it, usually with his best player.  Had Williams been out of shape, this would have been hard to do."

-- Roy Williams' profile in Scientific Football 2007

Going into [the week 12 Packers] contest, Williams had five games where he had 11 or more passes thrown to him versus zero for [Calvin] Johnson.  In that game the roles were reversed and Johnson saw fifteen passes to Williams' seven...

Keep that in mind as a reference point for Williams' post-week 12 comment that he didn't feel included in the offense.  It was so bad that Williams actually sat down with Mike Martz and spoke to him about it...

If this was an issue with Williams last year, it could be even more of one this year as the pass volumes drop.  Martz questioned Williams' work ethic before and when that is combined with the political issues, it becomes clear why Detroit was shopping him.  They are tired of dealing with the headache."

-- Roy Williams' profile in Scientific Football 2008

Does that second passage chill you?  The complaints about not seeing balls?  The sit down with the OC.  Sound all too familiar? 

And it raises the question, what happened in '07?  Williams started very well, posting big games against teams with good corners -- the Eagles, Redskins, Bears and Vikings.  Then, in the second half, his game came apart.  Was that promising '06 the fluke, or did Williams, as his supporters contend, simply wear down in the NFL penitentiary that is Detroit?

First, how good was Williams 2006?  Let's put it side by side with his 2007:

Williams was the real thing in 2006. He hit the marks for a number one -- he topped 9.0 overall and 7.0 against red-level corners.  (This variance shows why I argue for using these numbers to measure a player's potential ceiling, not his absolute value.)  Williams is absolutely capable of being a number one. 

I say that with several caveats.  First, Roy E. needs a position coach who will show him tough love, as his OC Mike Martz did in Detroit.   If Dallas is to see the return of the good Roy, Ray Sherman needs to ride him the way John Garrett rode Martellus Bennett this summer, the way Todd Haley rode Larry Fitzgerald in Arizona. 

I also question whether Roy E can co-exist with T.O.  Look at his '06 attempts -- his 152 throws translates to 9.5 a game.  It's the exact number of throws T.O. got that year.  Now, Todd Haley did manage to keep Owens, Terry Glenn and Jason Witten happy that year.  He threw an average of 9.5 times to Owens, 7.0 times a game to Glenn and 5.6 times a game to Witten. 

Jason Garrett may be able to employ a similar pass blend next year, but that would take throws away from Witten.  With Wade Phillips asking for more rushes in 2009, throws will be fought over more intensely, assuming the skill position personnel remain unchanged.

The bottom line is that Roy E. is most certainly capable of being a number one in Dallas' scheme.  Will Dallas give him that role, or ask him to be a bigger Terry Glenn in their offense? 

A lot will depend on how tolerant the team remains of Owens' baggage. An even bigger question is whether Terrell Owens is still capable of playing at his '07 level?

But I Got Wise... You're the Devil in Disguise!

The short answer is no.  T.O. 2007 versus T.O. 2008*:

*(Totals are for 14 games, not including the second Redskins and Eagles games.)

**(Red totals are for eight games of Dallas' season.)

These are raw, incomplete totals.  KC Joyner was kind enough to pass them on, despite being weeks away from starting SF '09.  Even though they're incomplete, they suggest somebody in the early stages of a steep decline: 7.5 defines middle of the pack.  That 2.0 average was computed over half a season, but it makes Patrick Crayton's '07 look stellar.

Owens could argue that he didn't have a healthy sidekick for much of the season.  Williams was hurt, as was Miles Austin.  Felix Jones only played in six games and Dallas went through December with Tashard Choice as its only healthy running back.  Owens looked like Superman in camp and he destroyed a very good Eagles secondary in week two.  But the Packers gave him bracket coverage the next week and T.O. faced that nearly every week the rest of the way. He never found a way to shake free.

The counter-arguments, however, accumulate.  Owens himself was never hurt.  He got the same rough number of balls in '08 as in '07.   Romo threw 9.3 passes his way in 2007, and 8.9 passes per game in 2008. 

The truly amazing stat comes against red level competition.  A 2.0 average is horrifying.  A team simply cannot earmark 140-150 passes a season for a receiver who can't evade top-level coverage. 

The NFC East has the highest concentration of red-level corners in football.  The Eagles had two this year in Sheldon Brown and Asante Samuel.  The Redskins had two in perennial Shawn Springs and Carlos Rogers.  New York's Corey Webster had an elite season. 

If you don't have it any more, the division will expose you in a hurry, and this year Owens was exposed. 

Was '08 an anomaly?  Here are Owens' raw YPAs the last five years:

• 2004 -- 9.3  - Philly's Super Bowl year;
  
• 2005 -- 8.7 -- the Eagles suspend T.O.
  
• 2006 -- 8.0 -- T.O. and Todd Haley/Parcells;
  
• 2007 -- 9.7
  
• 2008 -- 7.5

To my eyes there's a slow, but definite decline starting in '05, the year the Eagles cut him loose.  '07 was the year of Wade and Jason, when he was free of Reid, McNabb, Parcells and Haley.  T.O. responded, but if you're running the Cowboys, do you trust the one great year or the four declining ones?  Which one(s) is/are more likely to predict '09?

Will the Cowboys brass dismiss this as the product of scheme, of a failure to properly use Owens?  Or will they conclude that he's begun the inevitable decline?  How much will his overbearing personality factor into the team's ultimate decision?

Which devil has your ear, Jerry?