DYAR and the Dallas receiving corps...2007-09

[kramskoi]

...very sobering; Jimmy Johnson's assessment of Roy Williams...listening to him you wonder if we won't start seeing the real Roy Williams until next year or the year after...good thing he's young...

...in his 5 years in Detroit he had one elite-level year [2006] where he broke 1300 yds [7 tds] and his DYAR was an excellent 250, comparable to Owens that year [he was in the top 10 receivers that year and he consequently went to the Pro-Bowl]...it was also the only season where he started all 16 games...he was on pace for >1100 yds in 2007 but only started 12 games...ditto in 08 with the trade to Dallas in the first quarter of the season...

...this guy is a 1200yd/7 td receiver when he can make a full season but until he and Romo start clicking the Cowboys won't see that...and it's definitely looking unlikely this year...

...after last sunday, his DYAR is 5...by comparison, Crayton has climbed to 51 but he's well short of his stellar 2007 season where he reached 205 DYAR [21st in the league that year]...he did'nt have a lot of yards but he averaged ~ 1 TD per 100 yds [697/7 td]...very productive for a slot receiver...he was just as good as Donald Driver that year...he finished a dismal 70 DYAR during the 2008 campaign...

...curiously Owens, despite 10 TD's, rated only 75 DYAR last year and it's not hard to see why Dallas did'nt make the playoffs with the number 1 playing like a number 2...by comparison, Andre Johnson and Larry Fitzgerald both finished over 400 DYAR...such was the gap between Owens and the truly elite receivers last year...

...Witten's 205 DYAR was'nt enough to keep the passing game operating at an elite level in 2008...by the numbers; Dallas "seems" headed the same way this year...

...for comparison's sake we can compare the 13-3 Dallas offense of 2007 [by player] to 08 and 09 to date:

Player.........................2007 DYAR.................2008 DYAR...............2009 DYAR [thru week 11]

Jason Witten..........258 [21.3%][2nd].........203 [19.0%][3rd].........69 [8.0%][15th]

Terrell Owens.........448 [28.2%][3rd].........75 [-5.7%][46th].........-26 [-17.4%] [68th]- Buffalo

Roy Williams...........114 [1.0%] [43rd]......-29[-17.1%][73rd]...........5 [-11.6%] [56th]

Miles Austin............43 [33.8%][16th*]......106 [48.7%][4th*]........168 [22.6%] [15th]

Patrick Crayton.......205 [19.5%] [21st].....70 [0.9%][50th]............51 [-0.4%] [42nd]

Martellus Bennett...................................112 [20.8%] [13th].......-22 [-21.6%] [35th]

Sam Hurd................63 [8.9%] [10th*]...........not ranked..............26 [15.9%] [20th^]

=========================================================================
TOTAL DYAR...............1017...............................537............................297

RECORD......................13-3...............................9-7...............................?

=========================================================================
* denotes 10 - 50 passes
^ denotes 6 - 34 passes


...What a difference 2 years makes...total DYAR is well off 2007, with Witten, Bennett, Crayton and Hurd all down so far...incredibly, Williams is up slightly [relative to 08] along with Austin, who's clearly the best weapon Dallas has right now...his numbers are falling though...he was in the top 5 a few weeks ago...

...to date, not even Owens could save this offense in its current incarnation...not based on his 08 numbers...especially given his absolutely abyssmal season in Buffalo...note his incredible 07 season and the steep drop off in 08...what the heck happened last year?...

...i need not comment on Williams...the numbers speak for themselves...

...looking at these numbers it's clear to me that Dallas should be more of a power running/play-action team...the passing numbers just are'nt there...not to date anyway...they have less than two weeks to sort it all out before Manning, Rivers, and Brees come calling...and they still have to face McNabb and Campbell again...all good signal-callers...

...if they come off the mini-bye in the same offensive stupor of the past two weeks then it's going to be real ugly down the stretch...

...here's hoping but i have a sneaky feeling that if Dallas makes the playoffs this year it will be the defense that will get them there...

 

[Phoenix]

what in the blue h*ll is DYAR?

 

[Rudy]

what in the blue h*ll is DYAR?

Man, I was scared to ask the same thing. I'm guessing D-something Yards after Reception?

 

[Mavs Man]

what in the blue h*ll is DYAR?

From Wikipedia:

DPAR/DYAR
DPAR (Defense-adjusted Points Above Replacement) and DYAR (Defense-adjusted Yards Above Replacement) are statistics that measure a player's total contribution over the course of all plays in a game or season. The increment of DPAR is the number of additional points a player adds for his team. For example, if a player worth 3.0 DPAR in a given game had not played that week, and had been replaced in the lineup by a typical back-up player, his team would (ostensibly) have scored 3 fewer points. This increment is cumulative -- where DVOA for each play is averaged together, DPAR is compiled over time. In addition, a player is not compared to the league average, but to the expected performance of a replacement-level player, defined as being about 13.3% less valuable than the expected contribution of a league-average player (i.e. -13.3% DVOA).

Prior to the 2008 NFL season, FO [Football Outsiders] decided to change from DPAR to DYAR, measuring value in yards rather than points. This is essentially a change in semantics, however, as the two stats essentially measure the same thing; Schatz stated that the change came about simply as an effort to make FO's stats more accessible to the average fan.

When used to judge the contributions of individual players, DVOA and DYAR are as yet unable to separate completely a player's value independent of his 10 teammates on the field with him. From FO's website: "That means that when we say, 'Larry Johnson has a DVOA of 27.6%,' what we are really saying is 'Larry Johnson, playing in the Kansas City offensive system with the Kansas City offensive line blocking for him and Trent Green selling the fake when necessary, has a DVOA of 27.6%.'" While this is a significant limitation, it should be noted that it is a limitation shared by virtually all individual football statistics.

As for the rest of it, you are on your own.

 

[Hypnotoad]

are they gerrymandering stats now?

 

[kramskoi]

...for the sake of brevity...it means performance above the level of a second-stringer...a positive number means above-average...a negative, below-average...compared to a second-string player...

...it means Dallas would'nt notice any dropoff with Crayton and Austin starting...

 

[Hypnotoad]

...for the sake of brevity...it means performance above the level of a second-stringer...a positive number means above-average...a negative, below-average...compared to a second-string player...

...it means Dallas would'nt notice any dropoff with Crayton and Austin starting...

So an offense which evenly distributes the ball would score less?

 

[kramskoi]

So an offense which evenly distributes the ball would score less?

...i don't follow...

 

[jterrell]

...for the sake of brevity...it means performance above the level of a second-stringer...a positive number means above-average...a negative, below-average...compared to a second-string player...

...it means Dallas would'nt notice any dropoff with Crayton and Austin starting...

No it doesn't "mean" that and it certainly doesn't PROVE it.

The problem with these types of stats is people use them well beyond what they actually measure.

You can't make Crayton into RW11. He doesn't get covered the same way or even play the same position.

The entire concept is complete hogwash. There is no such thing as a "typical" backup player.

Football Outsiders do some cool things but often, as is the case here, they take stats and try to make them mean more than they do.

RW11 is playing poorly. We don't need stats to tell us that. But comparing stats a part-time guy puts up versus an every down player is as erroneous as can be. Offensive production is dependent upon many things beyond a WR's control. Grading the offense as a whole makes sense. Grading a WR in this way is just goofy.

 

[FuzzyLumpkins]

No it doesn't "mean" that and it certainly doesn't PROVE it.

The problem with these types of stats is people use them well beyond what they actually measure.

You can't make Crayton into RW11. He doesn't get covered the same way or even play the same position.

The entire concept is complete hogwash. There is no such thing as a "typical" backup player.

Football Outsiders do some cool things but often, as is the case here, they take stats and try to make them mean more than they do.

RW11 is playing poorly. We don't need stats to tell us that. But comparing stats a part-time guy puts up versus an every down player is as erroneous as can be. Offensive production is dependent upon many things beyond a WR's control. Grading the offense as a whole makes sense. Grading a WR in this way is just goofy.

Hogwash? Tell the Boston Red Sox and Minnesota Twins that.

Setting a statistical norm to compare players with is a tried and true tool in professional sports. You can argue the semantics of the word 'typical' all day long. That does nothing to detract from the methodology.

As for different players being covered differently you completely miss the point. That is the whole reason why you establish a norm: so you can compare players. They aren't being compared head to head which is the crux of what your asserting.

 

[kramskoi]

No it doesn't "mean" that and it certainly doesn't PROVE it.

The problem with these types of stats is people use them well beyond what they actually measure.

You can't make Crayton into RW11. He doesn't get covered the same way or even play the same position.

The entire concept is complete hogwash. There is no such thing as a "typical" backup player.

Football Outsiders do some cool things but often, as is the case here, they take stats and try to make them mean more than they do.

RW11 is playing poorly. We don't need stats to tell us that. But comparing stats a part-time guy puts up versus an every down player is as erroneous as can be. Offensive production is dependent upon many things beyond a WR's control. Grading the offense as a whole makes sense. Grading a WR in this way is just goofy.

...high on the list of things a WR can control is catch rate right?...he can control how he reads defenses can he not...he can have a knack for working zone coverage are beating press...he can control the fluidity and timing of his routes...he can make the right choice on option routes correct?

...for comparison's sake...both have been targeted over 50 times this season [51 for Crayton and 56 for Williams] and Crayton's catch rate is 55%, Williams 42%...at the moment they are more comparable than they are not...Williams might be seeing more single coverage now that Austin is seeing coverage rolled his way...but that sure has'nt helped his catch rate...its still in the low forties...he simply drops more balls that hit his hands...and his performance is based on his ability to catch, first and foremost...without a catch there are no yards...he has the worst catch rate of the entire Dallas offense except for Deon Anderson and he's only been targeted three times...

...you don't like my assessment, fair enough but don't try and sell me on a WR not having anything he can control when the ball is snapped...

...the purpose of this write-up was not to compare Williams with Crayton...only to take a look at how these players have performed over the last 2 1/2 seasons...i have nothing further to add...

 

[cowboyz]

Hogwash? Tell the Boston Red Sox and Minnesota Twins that.

Setting a statistical norm to compare players with is a tried and true tool in professional sports. You can argue the semantics of the word 'typical' all day long. That does nothing to detract from the methodology.

As for different players being covered differently you completely miss the point. That is the whole reason why you establish a norm: so you can compare players. They aren't being compared head to head which is the crux of what your asserting.

name some football teams that use sabermetrics for player evaluation

 

[FuzzyLumpkins]

name some football teams that use sabermetrics for player evaluation

Name some that don't. I know for a fact that Parcells was very very big on using stats to find which ones correlated to wins. Its not that big of a leap of logic for him to use stats for other evaluations.

Thats just one example but quite frankly the methodology has a huge track record of success. Sports is just one minuscule example.

 

[BlindFaith]

...high on the list of things a WR can control is catch rate right?...he can control how he reads defenses can he not...he can have a knack for working zone coverage are beating press...he can control the fluidity and timing of his routes...he can make the right choice on option routes correct?

What he can not control are pass accuracy, if he is tightly covered and the ball is still thrown to him.

...for comparison's sake...both have been targeted over 50 times this season [51 for Crayton and 56 for Williams] and Crayton's catch rate is 55%, Williams 42%...at the moment they are more comparable than they are not...Williams might be seeing more single coverage now that Austin is seeing coverage rolled his way...but that sure has'nt helped his catch rate...its still in the low forties...he simply drops more balls that hit his hands...and his performance is based on his ability to catch, first and foremost...without a catch there are no yards...he has the worst catch rate of the entire Dallas offense except for Deon Anderson and he's only been targeted three times...

This is just BS. Tell me again how many drops Roy has been credited for. Show me the official stat. And while you are looking at these receptions to times targeted, look at the rest of the league. I posted this in the Roy flame thread. Tell me why ALL of our WRs are the lowest in the league. Only Witten has a good number.

When looking at these numbers lets also look at QB completion percentages. Romo is the 21st ranked QB.

So, with that being said, either ALL of our WRs suck or Romo is having a bad year. It's not just Roy.

The only variable that can't be answered by anyone on this board, try as they might, is are the WRs running the correct routes. If they aren't, then the bad completion percentage for Romo and the low RTT is clearly on them. If they are running the correct routes and making the correct reads, then the low numbers are on Romo. It's really that simple.

Now I don't have the direct quotes, but I know for a fact that earlier in the year Brian Broaddus (sp) clearly stated that Roy was running the correct routes and he was getting open. The interception Romo threw to Baily was on him as everyone said Austin ran the correct route. Coaches have said that Roy is running the correct routes and making terrific catches in practice.

Crayton has made a comment that Romo didn't know the plays last year. Comments have come out about Romo flipping receiver sets. TO called out Romo and Witten and Garrett about plays. Romo has been questioned about his practice habits and his training and conditioning.

Like I always say, when there's smoke there's fire, however small a fire it may be.

Romo's completion percentages have been in decline since 2006.

But, as I have said before, feel free to blame ALL the receivers and Bennett and the running backs for not being productive.

Or you could point to Romo and Garrett, who are the two who have the most direct impact on how successful or unsuccessful these other positions play.

I am going to soften my stance on Romo though. His first couple of years were extremely productive, even though he still did struggle against top defenses.

So I don't know if he is just struggling this year or if teams have just figured out how to play him and this offense. The inaccurate throws are completely on him. The decisions he makes with the ball are on him. What plays are called and getting the matchups they want are on Garrett.

 

[BlindFaith]

Name some that don't. I know for a fact that Parcells was very very big on using stats to find which ones correlated to wins. Its not that big of a leap of logic for him to use stats for other evaluations.

Thats just one example but quite frankly the methodology has a huge track record of success. Sports is just one minuscule example.

The problem with using stats in football is that it's a total team game. There are so many variables involved on any individual stat that it really becomes difficult to define any clear baseline.

Take number of receptions. So a WR only has 20 receptions. The league leader has 65. What does that stat say?

Looking at it for face value it says that receiver isn't being very productive.

So you really need to look into it further. How many times was he targeted. How many drops did he have. Those are things you can validate.

What you still can not validate are is he running the correct routes. When he was thrown to, how open was he really. Are the plays called for him plays he is suited to run or play to his strengths.

All of the above mentioned variables play into the number of receptions stat. But the stat is meaningless unless you know definitively what the answer is to all of the other factors.

Baseball is an individual sport. Stats are what they are. If a guy hits .300, then he is a .300 hitter. No other variable impacts that unless you want to factor in injuries, but injuries would factor in to any sports statistic as well.

 

[JBond]

What he can not control are pass accuracy, if he is tightly covered and the ball is still thrown to him.



This is just BS. Tell me again how many drops Roy has been credited for. Show me the official stat. And while you are looking at these receptions to times targeted, look at the rest of the league. I posted this in the Roy flame thread. Tell me why ALL of our WRs are the lowest in the league. Only Witten has a good number.

When looking at these numbers lets also look at QB completion percentages. Romo is the 21st ranked QB.

So, with that being said, either ALL of our WRs suck or Romo is having a bad year. It's not just Roy.

The only variable that can't be answered by anyone on this board, try as they might, is are the WRs running the correct routes. If they aren't, then the bad completion percentage for Romo and the low RTT is clearly on them. If they are running the correct routes and making the correct reads, then the low numbers are on Romo. It's really that simple.

Now I don't have the direct quotes, but I know for a fact that earlier in the year Brian Broaddus (sp) clearly stated that Roy was running the correct routes and he was getting open. The interception Romo threw to Baily was on him as everyone said Austin ran the correct route. Coaches have said that Roy is running the correct routes and making terrific catches in practice.

Crayton has made a comment that Romo didn't know the plays last year. Comments have come out about Romo flipping receiver sets. TO called out Romo and Witten and Garrett about plays. Romo has been questioned about his practice habits and his training and conditioning.

Like I always say, when there's smoke there's fire, however small a fire it may be.

Romo's completion percentages have been in decline since 2006.

But, as I have said before, feel free to blame ALL the receivers and Bennett and the running backs for not being productive.

Or you could point to Romo and Garrett, who are the two who have the most direct impact on how successful or unsuccessful these other positions play.

I am going to soften my stance on Romo though. His first couple of years were extremely productive, even though he still did struggle against top defenses.

So I don't know if he is just struggling this year or if teams have just figured out how to play him and this offense. The inaccurate throws are completely on him. The decisions he makes with the ball are on him. What plays are called and getting the matchups they want are on Garrett.

Lots of good stuff there. One thing that I can see with my own eyes, without the use of stats is Romo is making the rookie mistake of staring down receivers. My opinion is that he is very concerned with turnovers (as he should be) but because of that he looks to certain players on certain plays and defenses seem to be picking up on it. Just my useless 2 cents.

 

[kramskoi]

The problem with using stats in football is that it's a total team game. There are so many variables involved on any individual stat that it really becomes difficult to define any clear baseline.

Take number of receptions. So a WR only has 20 receptions. The league leader has 65. What does that stat say?

Looking at it for face value it says that receiver isn't being very productive.

So you really need to look into it further. How many times was he targeted. How many drops did he have. Those are things you can validate.

What you still can not validate are is he running the correct routes. When he was thrown to, how open was he really. Are the plays called for him plays he is suited to run or play to his strengths.

All of the above mentioned variables play into the number of receptions stat. But the stat is meaningless unless you know definitively what the answer is to all of the other factors.

Baseball is an individual sport. Stats are what they are. If a guy hits .300, then he is a .300 hitter. No other variable impacts that unless you want to factor in injuries, but injuries would factor in to any sports statistic as well.

...i really hate to bump this because the thread is'nt really about Williams exclusively...per Bob Sturm, this is the distribution for the entire season...and one must remember that all players caught balls primarily from the same quarterback...only Anderson has a lower catch rate but he's only been target three times...

...i want to be clear that i still think they can figure out how to get the most out of Roy...it just looks like it will take more time...but the results below speak to his production value to the offense currently...his catch rate is not allowing him to make the most of his opportunities, for whatever reasons...


Name...Targets..Catches....%.......YardsFD/TD/INT

Witten.......71......53.......75%.....48125/1/3
Austin.......62.......35.......56%.....67918/7/2
Crayton.....51.......28.......55%.....42712/3/1

Williams...56.......24......42%.....42815/3/0

Choice......21.......15........71%....1327/0/0
Bennett....24.......13........54%.....1397/0/0
Hurd........10.........6........60%.....1253/1/1
Barber......16.......13........81%.....983/0/0
Jones........6.........5........83%....562/0/0
Anderson...3.........1......33%.....50/0/
Ogletree....4.........4.......100%....583/0/0
Phillips.......2.........1........50%....00/0/

Totals......326.....198.......61%....262498/15/7

 

[viman96]

sorry man but it is annoying you start each sentence with ...

 

[BlindFaith]

...i really hate to bump this because the thread is'nt really about Williams exclusively...per Bob Sturm, this is the distribution for the entire season...and one must remember that all players caught balls primarily from the same quarterback...only Anderson has a lower catch rate but he's only been target three times...

...i want to be clear that i still think they can figure out how to get the most out of Roy...it just looks like it will take more time...but the results below speak to his production value to the offense currently...his catch rate is not allowing him to make the most of his opportunities, for whatever reasons...

That I agree with. But to say "Roy sucks, bench Roy, Roy's a bum" is a mistake. Not saying you are saying this ... directly. But others are.

The problem is this offense and the inability to effectively and consistently get anyone not named Witten involved in the passing game. This goes for Felix, Austin, Roy, Bennett.

I don't believe that all of these players are below average. They may not be perennial pro bowlers, but how their production statistically compares to others is far too low for what their abilities should provide.

 

[BlindFaith]

Lots of good stuff there. One thing that I can see with my own eyes, without the use of stats is Romo is making the rookie mistake of staring down receivers. My opinion is that he is very concerned with turnovers (as he should be) but because of that he looks to certain players on certain plays and defenses seem to be picking up on it. Just my useless 2 cents.

Agreed. Something has changed with Romo. I fear that he excels at "playground ball" and struggles sitting in the pocket and reading defenses.

He has been at his best when a play breaks down and he buys time.

In the pocket, he does seem to lock in on a receiver and not read the defense to find the best match up.

But then why not more role outs and boot legs? This goes back to Garrett and not designing the offense around the talents of the players, but rather sticking to his plays hell or high water.