FanPost

Roy Williams in the Red Zone: Under the Microscope

First, special thanks to OCC for helping me get started and providing invaluable advice on undertaking EPV analysis. Having looked at a very small part of the Cowboys offense I have an immense appreciation for the work that goes into his posts.

For the 2009 season Roy Williams had the 11th best average EPV per target among all NFL WRs with at least 10 red zone targets. For a primer on EPV see One Cool Customer’s post here. Roy Williams posted an average EPV per target of 0.35, better than Andre Johnson, Brandon Marshall, Larry Fitzgerald, and Randy Moss. After the jump: the top 30 red zone EPV/per target for all NFL players with at least 10 targets (including interceptions, as explained below).


Red Zone EPV per attempt including ints (minimum of 10 targets)

Rank

Player

Sum of RZ EPV

RZ Targets

Avg. of RZ Targets

1

V.Davis (SF)

 13.05

 13

 1.00

2

L.Coles (CIN)

 11.23

 13

 0.86

3

D.Clark (IND)

 11.85

 14

 0.85

4

D.Mason (BAL)

 10.92

 14

 0.78

5

C.Ochocinco (CIN)

 13.16

 17

 0.77

6

V.Shiancoe (MIN)

 11.82

 16

 0.74

7

S.Rice (MIN)

 14.25

 22

 0.65

8

J.Avant (PHI)

 8.14

 13

 0.63

9

F.Davis (WAS)

 7.86

 14

 0.56

10

D.Bowe (KC)

 7.73

 14

 0.55

11

H.Miller (PIT)

 8.93

 17

 0.53

12

A.Collie (IND)

 5.63

 14

 0.40

13

A.Gates (SD)

 5.09

 13

 0.39

14

J.Knox (CHI)

 3.80

 10

 0.38

15

V.Jackson (SD)

 4.71

 13

 0.36

16

P.Crayton (DAL)

 4.29

 12

 0.36

17

R.Williams (DAL)

 5.21

 15

 0.35

18

T.Gonzalez (ATL)

 8.52

 25

 0.34

19

S.Smith (NYG)

 7.50

 24

 0.31

20

J.Finley (GB)

 5.06

 17

 0.30

21

P.Harvin (MIN)

 3.49

 13

 0.27

22

A.Johnson (HOU)

 6.44

 25

 0.26

23

O.Daniels (HOU)

 2.53

 10

 0.25

24

R.White (ATL)

 5.04

 20

 0.25

25

W.Welker (NE)

 3.84

 16

 0.24

26

J.Addai (IND)

 2.47

 13

 0.19

27

K.Walter (HOU)

 2.19

 13

 0.17

28

C.Chambers (KC)

 1.34

 10

 0.13

29

M.Sims-Walker (JAC)

 1.82

 15

 0.12

30

G.Jennings (GB)

 1.30

 11

 0.12

 

Some of you may remember OCC’s post analyzing the Cowboy’s passing game. Disclosure: the impetus for this post was OCC’s post and my disagreement with the conclusions regarding Roy William’s red zone performance. Those who noticed the similarity may also have noticed the numbers in the initial post for Crayton, Witten, and Austin are different from the numbers in the table above.

Red Zone EPV per attempt excluding ints

Player

Sum of RZ EPV

RZ Targets

Avg. of RZ Targets

P.Crayton (DAL)

 8.24

 11

 0.75

R.Williams (DAL)

 5.21

 15

 0.35

J.Witten (DAL)

 1.56

 6

 0.26

M.Austin (DAL)

 3.59

 20

 0.18

 

The reason for the difference is how interceptions are treated. The OCC analysis excludes the impact of interceptions from the WR’s statistics while I have included the impact of interceptions. Either approach is defensible. I would argue that one of the main benefits of EPV analysis is that it captures the impact of both positive and negative plays and hence interceptions should to be included. The negative impact from incomplete passes is included where it may not be clear whether the QB or the WR is at fault; for consistency I have treated interceptions the same. The other side of the argument is that we can’t determine whether it was the QB or the WR that was at fault for an interception. Hence the impact should be excluded. In addition, especially in the red zone, interceptions have a disproportionately large impact. To conclude, I think either approach is defensible, however I have included interceptions in the calculation. BTW, in case you forgot, there were three red zone interceptions: Austin (picked by Champ Bailey), Witten (picked by Charles Woodson), and Crayton (picked by Asante Samuel in the final regular season game).

Returning to Roy Williams’ performance, I had two issues with the initial analysis. First, according to the statistics ex-interceptions, Roy Williams produced the 2nd best EPV among Crayton, Williams, Witten, and Austin. From that data, the analysis concluded that Roy William’s production could be easily replaced. I didn’t see that as a logical conclusion of the data.

Second, we only knew the statistics for Dallas players. Hence there was limited basis for comparison. We needed the statistics for the entire NFL in order to have a basis for saying whether a player was above or below average for the NFL. So I calculated the EPV for every red zone passing play in the 2009 NFL season.

Having the figures for all NFL players produces a clearer picture of Roy Williams performance. My first conclusion is that with or without interceptions, Roy Williams performance in the Red Zone was respectable. It is apparent is that Roy’s 0.35 Red Zone EPV looked mediocre only because it was being compared with Crayton’s stellar 0.75 average EPV (ex-interceptions), which was 3rd best in the entire NFL for a WR with at least 10 targets.

My second conclusion is that red zone plays to players not named Crayton, Williams, and Austin were a real problem for Dallas. Going in I knew that Romo had one of the largest differences in the NFL between his overall completion percentage (63%) and his completion percentage inside the 10 (44%). After calculating the results I was really confused. How Dallas could have two WR in the top 20 in red zone efficiency and Romo simultaneously struggle inside the 10? Here’s the answer

Player

Sum of RZ EPV

RZ Targets

Avg. of RZ Targets

M.Austin (DAL)

 1.02

 21

 0.05

R.Williams (DAL)

 5.21

 15

 0.35

P.Crayton (DAL)

 4.29

 12

 0.36

Subtotal

 10.52

 48

 0.22

 

 

 

 

J.Witten (DAL)

 (3.97)

 7

 (0.57)

S.Hurd (DAL)

 (4.49)

 5

 (0.90)

F.Jones (DAL)

 (2.70)

 4

 (0.68)

M.Barber (DAL)

 (3.43)

 3

 (1.14)

M.Bennett (DAL)

 (3.82)

 3

 (1.27)

T.Choice (DAL)

 1.05

 2

 0.53

J.Phillips (DAL)

 (1.49)

 1

 (1.49)

Subtotal

 (18.85)

 25

 (0.75)

 

From Jason Witten through John Phillips, that’s 25 attempts producing negative 19 in EPV. Yikes!

One other minor point. I’ve noticed a tendency to discount or disregard the results of statistical analysis when it doesn’t conform to conventional wisdom. But isn’t that the point: to tell us when the conventional wisdom is wrong? The purpose of the statistical movement is to improve our subjective assessments by applying an objective process. Our recollections can be both imperfect and biased. For example, if one is frustrated by Roy Williams then the instances where he drops the ball are likely to be infuriating and remembered vividly. If one is going to read statistical analysis does it make sense to turn around and disregard it because of subjective perceptions?  Let me give an example. When I initially pointed out that the data suggested that Roy Williams had performed better than Miles Austin, several comments suggested that Austin saw worse throws and that’s why his EPV was worse.

Roy may have more [EPV] in the Red Zone, but how many of those passes to Miles in the red zone were actually catchable? Roy dropped very easy to catch balls while Miles caught the tough ones.

I remember several hitting Roy right in the hands in the end zone … I can distinctly remember a couple sailing over Austin’s head and a couple behind him in the end zone…To be fair, Williams caused at least one INT when the ball bounced off his hands

However, looking at the EPV’s and the play-by-play reminded me of Austin’s two worst plays which I had forgotten: A drop in the end zone against Kansas City and the Champ Bailey interception when Austin and Romo weren’t on the same page, neither of which were uncatchable passes. I had forgotten the drop against Kansas City until I read the play by play. Of course I remember it now because it hit Austin right in the hands and the announcers kept saying ‘imagine if Austin had caught that, he could have had 260 yards and 3 TDs’.  The point being that while it’s very difficult to remember every bad pass and interception over the course of the season, the impact of all incomplete passes and interceptions are going captured by the EPV analysis. Tread carefully when suggesting subjective assessments trump statistics.

For those that are interested, I’ve also posted the play-by-play for every Roy Williams red zone target and every Miles Austin red zone target. In addition to the play-by-play logs, I’ve also provided the following data in the link below (the data includes all players vs. only the top 30 included in the post).

1.  Red Zone EPV inc. ints. (players with at least 10 targets)

2. Red Zone EVP ex. ints. (players with at least 10 targets)

3. Red Zone EPV inc. ints. (all NFL players)

4. Red Zone EPV ex. ints. (all NFL players)

https://spreadsheets.google.com/ccc?key=0Aq_pzudn-VEDdGJhcGcyZWRzak5pbDNmOTNxcDhMZEE&hl=en

Another user-created commentary provided by a BTB reader.

X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Blogging The Boys

You must be a member of Blogging The Boys to participate.

We have our own Community Guidelines at Blogging The Boys. You should read them.

Join Blogging The Boys

You must be a member of Blogging The Boys to participate.

We have our own Community Guidelines at Blogging The Boys. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker