Projecting 2015 season

[HardRightEdge]

PFF charged Shields with the TD pass to Lockette on that play.

I find that very interesting. One of the other purveyors of data I cited earlier (I'm too lazy to go back too see which one) mentioned they have a category they referred to as "uncovered" which they do not charge to anybody. That's a conservative approach in limiting assumptions, and subjectivity, regarding responsibility.

In any event, if somebody must be charged it should be Shields from the looks of it, but that charge does not reflect on his coverage ability per se. It's more about football smarts in the context of team play, as opposed to one-one-one skill on the island. So, if we're prone to think of QB-rating-against as a measure of how well a guy covers receivers and defends passes, that's not entirely the case, at least with PFF. In this case Shields was not even attempting to cover anybody.

There's another possibility, though. While I don't think this is the case, perhaps in this set and situation, the corner on the side Wilson is facing at the time of the Lynch fake is supposed to come up in run support for the QB run option. Under that scenario, if Wilson was facing right on the fake with the wide LB on the opposite side, perhaps Williams would be responsible for run support. It's possible Dix blew the play by not immediately coming over to take Lockette on the Lynch fake rather than reacting to Wilson's roll in the context of Shields being assigned to defend run immediately on the Lynch fake.

Isn't it odd that Shields would simply abandon his man nearly at the snap when this could have just as easily have been a play action throw? That's an incredibly risky move if he was not expecting the safety to take his man.

It strikes me that when presenting advanced analytics, it's most constructive to err on the side of caution in limiting subjective interpretation as much as possible. In this case, not knowing the defensive assignments and especially given that Shields was not attempting to cover anybody, it seems reasonable that it should be declared a "team error".

Further, having a "team error" category specifically for those cases where responsibility is not very clear would go some way in measuring team cohesion and scheme coherence.

Under such a scoring scheme, and to take one scenario, a team where the CBs have good individual ratings while the "team error" category is high would indicate the players are good one-on-one while not being so hot functioning as team.

The issue is perhaps clearer in the occasional circumstance where a corner releases the wideout, clearly expecting top side support, and the play ends up being a big gainer because the support never materialized. Who's to blame? Unless the coaches tell you how they graded that play, one resorts to speculation or interpreting body language when the yapping ensues in the end zone. When the yapping happens with regularity with one safety in one season, and then stops the next when he's gone, our suspicions are confirmed about who was to blame a good deal of the time. But that's long after the grades have been posted. At the time it was happening, the frequency with which DBs were still being moved around at the snap was also disturbing, but only provided circumstantial evidence.

 

[Sunshinepacker]

As a note, PFF is very open about the fact that they don't know play calls when they evaluate players. Over the entirety of the season you would imagine the majority of these kinds of 50/50 plays would wash out. In this instance maybe Shields was at fault, maybe he wasn't. Throughout the season I'm sure there are similar plays for every corner. It's not an exact science, more of an indication.

 

[Poppa San]

Back on topic or this gets locked!!

 

[Deleted member 6794]

I'll leave this topic for awhile with a couple of last observations on why my contention that Cobb was the "NFL most valuable 2014 receiver" defies most folks' eye test, if not their statistical tests, Packer fans included.

1. I noted above that Cobb made a first down on 56.3% percent of his targets, #1 among the top 13 yards-per-game leaders. What I did not note is that his percent was considerably higher than the league #2. Also, I did not note that Cobb made a first down on 78.0% of his catches. That's crazy good; the gap over #2 gets wider.

2. Despite those crazy good possession receiver metrics he still averaged 14.1 yds. per catch.

3. I already noted that he led those top 13 yards-per-game receivers in yards after catch per target. I did not previously mention that his 556 gross YAC was 4th. among all receivers or that 43% of his yards were after the catch. The only top receiver with a higher YAC % of total yards was Golden Tate (another underrated receiver).

What I draw from this is that certain receivers get extra credit from the highlight reels which tend toward the circus catches and blowing by a corner for a long ball.

Cobb catches a lot of balls in short range and runs for the first down or gets open in the intermediate range for first downs and adds yards after the catch, adding a fair number of TDs in the process. Less glamorous than the highlight reel exploits, but no less valuable. Yards, first downs and TDs have the same value regardless of how their achieved.

And again, he blew everybody away on a per target basis last season.

Here´s an interesting take by Football Outsiders on the topic, ranking receiver by a plus/minus system which takes a lot of different stuff into consideration. Cobb is ranked 9th in total +/- and doesn´t crack the top 10 per target.

http://www.footballoutsiders.com/stat-analysis/2015/2014-receiving-plus-minus

 

[HardRightEdge]

Here´s an interesting take by Football Outsiders on the topic, ranking receiver by a plus/minus system which takes a lot of different stuff into consideration. Cobb is ranked 9th in total +/- and doesn´t crack the top 10 per target.

http://www.footballoutsiders.com/stat-analysis/2015/2014-receiving-plus-minus

I wasn't interested enough to try to find out how they defined +/- in 2010.

Perhaps there are several dubious propositions that go into it, like this one:

"Meanwhile, a 16-yard gain on third-and-20 is also a failed completion, but if the receiver catches the ball 10 yards down the right sideline and then gains 6 YAC, on average that's going to be a more impressive play on his part. Both receivers helped their catch rate and gained the same amount of YAC, but the second receiver overcame more difficult circumstances to produce his play."

Besides, if getting a first down distinguishes "failed" from "successful" completions, why isn't Cobb in the top 10?

It all strikes me as too clever by half.

We do know this: "The next-to-last column, C%+, is the player's plus-minus divided by the targets." Therefore, Kenny Stills and Brandin Cooks are the finest wide receivers in all the land on a per target basis, doncha know.

 

[Deleted member 6794]

I wasn't interested enough to try to find out how they defined +/- in 2010.

Perhaps there are several dubious propositions that go into it, like this one:

"Meanwhile, a 16-yard gain on third-and-20 is also a failed completion, but if the receiver catches the ball 10 yards down the right sideline and then gains 6 YAC, on average that's going to be a more impressive play on his part. Both receivers helped their catch rate and gained the same amount of YAC, but the second receiver overcame more difficult circumstances to produce his play."

The article states that a five yard gain on 3rd-and-20 is a failed completion though.

Besides, if getting a first down distinguishes "failed" from "successful" completions, why isn't Cobb in the top 10?

It all strikes me as too clever by half.

We do know this: "The next-to-last column, C%+, is the player's plus-minus divided by the targets." Therefore, Kenny Stills and Brandin Cooks are the finest wide receivers in all the land on a per target basis, doncha know.

First of all Cobb is in the top 10 in total +/- at #9.

The statistic estimates how many passes a receiver caught compared to what an average receiver would have caught, given the location of those passes. Unlike simple catch rate, plus-minus does not consider passes listed as "Thrown Away," "Tipped at Line," or "Quarterback Hit in Motion."

Stills and Cooks both lined up on the outside for more than 70% of their routes yet ended up with either better stats per target or at least close to Cobb's stats.

Cobb ended up being the second highest ranked slot receiver behind Jarvis Landry. While I don't agree with that there's some merit to it as the Dolphins receiver has a higher catch rate and a better YAC percentage than Randall.

 

[HardRightEdge]

The article states that a five yard gain on 3rd-and-20 is a failed completion though.

So? I wouldn't expect every element of the methodology to be questionable.

First of all Cobb is in the top 10 in total +/- at #9.

But not on a per target basis.

The statistic estimates how many passes a receiver caught compared to what an average receiver would have caught, given the location of those passes. Unlike simple catch rate, plus-minus does not consider passes listed as "Thrown Away," "Tipped at Line," or "Quarterback Hit in Motion."

According to ESPN, Cobb's target count was 126. FO has a 125 count, with a scant single pass falling in those categories, which begs a question. Why does a QB intentionally throw a ball away? As often as not it is because the receiver in who's direction the ball is thrown did not get separation.

Another question is begged. The most obvious passes to be excluded are spikes. Clearly, the ESPN stats are flawed in this regard...every pass gets attributed to a receiver, but don't ask me how. It's interesting that FO did not make a point of this. According to PFF, about 1% of passes are spikes.

Stills and Cooks both lined up on the outside for more than 70% of their routes yet ended up with either better stats per target or at least close to Cobb's stats.

That's what FO says, which is what is being debated. I wouldn't even put these guys in the conversation without at least 100 targets. Why didn't they get thrown more balls? Cooks played only 10 games, but we don't give awards for excuses. You can't be in the conversation for "most valuable in 2014" if you're not on the field. Stills played 15 games, so he doesn't have an excuse.

Cobb ended up being the second highest ranked slot receiver behind Jarvis Landry. While I don't agree with that there's some merit to it as the Dolphins receiver has a higher catch rate and a better YAC percentage than Randall.

That's because Landry is a slot receiver is the more traditional sense...more short stuff leading to a higher catch rate and more YAC, as FO points out. There is no denying RBs rank first, slots and TEs next in those stats. Cobb was not running as many of these kinds of routes, as his 14 yds. per catch indicates. There are a lot of intermediate routes in the middle of the field in those stats. In Cobb's 2014 season, he combined WR stats out of the slot in yards per catch and TDs, while putting up slot numbers in catch %, FD % and YAC while having a meaningful number of targets. He was unique and highly valuable in 2014. There is another aspect to his value not captured in these stats...he may be the best play-extending receiver in the league.

On thing I note in the FO stats...they don't seem to attribute any extra value for TDs.

 

[Deleted member 6794]

According to ESPN, Cobb's target count was 126. FO has a 125 count, with a scant single pass falling in those categories, which begs a question. Why does a QB intentionally throw a ball away? As often as not it is because the receiver in who's direction the ball is thrown did not get separation.

Cobb lined up in the slot 87.3% of his routes and according to PFF had 106 of his 125 (86.4%) targets there. QBs rarely through the ball away over the middle of the field.

Another question is begged. The most obvious passes to be excluded are spikes. Clearly, the ESPN stats are flawed in this regard...every pass gets attributed to a receiver, but don't ask me how. It's interesting that FO did not make a point of this. According to PFF, about 1% of passes are spikes.

Rodgers didn´t spike the ball a single time last season.

That's what FO says, which is what is being debated. I wouldn't even put these guys in the conversation without at least 100 targets. Why didn't they get thrown more balls? Cooks played only 10 games, but we don't give awards for excuses. You can't be in the conversation for "most valuable in 2014" if you're not on the field. Stills played 15 games, so he doesn't have an excuse.

Taking Football Outsiders´ numbers for passes into consideration here´s how Stills and Cooks match up with Cobb based on per target production:

First downs: Cobb: 56.8% Stills: 59.5% Cooks: 39.1%
First down per catch: Cobb 78.0% Stills: 74.6% Cooks: 47.2%
Yards per catch: Cobb 14.1 Stills 14.8 Cooks 10.8

With Stills (79.7%) having a significantly higher catching rate than Cobb (72.8%) it makes sense to rate him ahead of Randall. Cooks has a fantastic catching rate (82.8%) but according to the numbers posted above he shouldn´t be ranked ahead of the Packers receiver.

That's because Landry is a slot receiver is the more traditional sense...more short stuff leading to a higher catch rate and more YAC, as FO points out. Cobb was not running as many of these kinds of routes, as his 14 yds. per catch indicates. There are a lot of intermediate routes in the middle of the field in those stats. There is another aspect to his value not captured in these stats...he may be the best play-extending receiver in the league.

I agree that Landry caught a ton of short passes with the ball being an average of 3.75 yards in the air per reception. While Cobb´s number is significantly higher at 8.03 it´s not close to most outside receivers. Just as an example Stills´ number was at 11.95 yards in the air per reception last year.

On thing I note in the FO stats...they don't seem to attribute any extra value for TDs.

That made me wonder as well.

 

[HardRightEdge]

Cobb lined up in the slot 87.3% of his routes and according to PFF had 106 of his 125 (86.4%) targets there. QBs rarely through the ball away over the middle of the field.

That's not the point. Eliminating all throwaways benefits WRs in those instances where the cause was failure to get separation. It goes to methodology.[/QUOTE]

Rodgers didnt spike the ball a single time last season.

That's not the point. To repeat, PFF has noted in one past season that 1% of passes are spikes. Why didn't they mention their exclusion? It goes to methodology.

Taking Football Outsiders´ numbers for passes into consideration here's how Stills and Cooks match up with Cobb based on per target production:

First downs: Cobb: 56.8% Stills: 59.5% Cooks: 39.1%
First down per catch: Cobb 78.0% Stills: 74.6% Cooks: 47.2%
Yards per catch: Cobb 14.1 Stills 14.8 Cooks 10.8

With Stills (79.7%) having a significantly higher catching rate than Cobb (72.8%) it makes sense to rate him ahead of Randall. Cooks has a fantastic catching rate (82.8%) but according to the numbers posted above he shouldn´t be ranked ahead of the Packers receiver.

Cobb had 126 targets. Stills had 85; Cooks had 70. These New Orleans guys are not in the conversation as 3rd. and 4th. options. FO used 50 targets as their cut off. While it makes for interesting comparisons under a bigger tent, Cobb was by far the more valuable player in 2014. I wouldn't consider a receiver to be in that discussion with anything less than 100 targets; 125 is probably more like it as a solid #2, or a #1 in a spread-it-around offense.

I agree that Landry caught a ton of short passes with the ball being an average of 3.75 yards in the air per reception. While Cobb´s number is significantly higher at 8.03 it´s not close to most outside receivers. Just as an example Stills´ number was at 11.95 yards in the air per reception last year.

New Orleans used a classic passing attack with Graham and Cooks for ball control, Colston and Stills down the field.

Cobb was a dual threat in 2014...as a ball control receiver and a big play threat.

 

[Deleted member 6794]

That's not the point. To repeat, PFF has noted in one past season that 1% of passes are spikes. Why didn't they mention their exclusion? It goes to methodology.

Last year there were 66 spikes 0n 19,094 pass plays (0.35%). I guess that Football Outsiders excluded those passes out of the equation as part of the throw aways.

Cobb had 126 targets. Stills had 85; Cooks had 70. These New Orleans guys are not in the conversation as 3rd. and 4th. options. FO used 50 targets as their cut off. While it makes for interesting comparisons under a bigger tent, Cobb was by far the more valuable player in 2014. I wouldn't consider a receiver to be in that discussion with anything less than 100 targets; 125 is probably more like it as a solid #2, or a #1 in a spread-it-around offense.

New Orleans used a classic passing attack with Graham and Cooks for ball control, Colston and Stills down the field.

Cobb was a dual threat in 2014...as a ball control receiver and a big play threat.

Well, you tried to prove your point that Cobb was the most valuable receiver based on per target production. Stills, while having less targets, still had enough to be considered relevant and had better numbers than Cobb.

 

[HardRightEdge]

Well, you tried to prove your point that Cobb was the most valuable receiver based on per target production. Stills, while having less targets, still had enough to be considered relevant and had better numbers than Cobb.

Well, in my book you cannot be most valuable anything as what amounts to a part time player.

 

[Deleted member 6794]

Well, in my book you cannot be most valuable anything as what amounts to a part time player.

I didn't name Stills the most valuable receiver in the league. He had fantastic numbers per target though so I understand how he ended up on top of Football Outsiders' list.

 

[Megatron]

The only receiving duo in the league I'd take over Nelson and Cobb is Calvin Johnson and Golden Tate. If Rodgers had those two(especially Calvin) the offense would break all-time records and we'd win the SB rather easily.

 

[Deleted member 6794]

Well, there goes this year.

I'm sorry, that should have been posted in the notes thread.

 

[Joe Nor Cal Packer]

But still on many fewer targets! The point being, if Cobb's gross numbers moved him toward the very top of the league on a 15 - 18% team passing frequency adjustment, perhaps then the outrageously good productivity per target numbers might have gotten some notice.

If you break down the numbers on a per target basis, with Antonio Brown's stunning 44% more targets than Cobb, his numbers are equivalent to Randall Cobb + a lousy tight end. The spread between the two players on a per target basis is so wide as to not justify rating Brown as the better 2014 wide receiver given that Cobb had a significant number of targets (126, 25th. in the league) as opposed to a bit player like a Beasley.

I'll leave this topic for awhile with a couple of last observations on why my contention that Cobb was the "NFL most valuable 2014 receiver" defies most folks' eye test, if not their statistical tests, Packer fans included.

1. I noted above that Cobb made a first down on 56.3% percent of his targets, #1 among the top 13 yards-per-game leaders. What I did not note is that his percent was considerably higher than the league #2. Also, I did not note that Cobb made a first down on 78.0% of his catches. That's crazy good; the gap over #2 gets wider.

2. Despite those crazy good possession receiver metrics he still averaged 14.1 yds. per catch.

3. I already noted that he led those top 13 yards-per-game receivers in yards after catch per target. I did not previously mention that his 556 gross YAC was 4th. among all receivers or that 43% of his yards were after the catch. The only top receiver with a higher YAC % of total yards was Golden Tate (another underrated receiver).

What I draw from this is that certain receivers get extra credit from the highlight reels which tend toward the circus catches and blowing by a corner for a long ball.

Cobb catches a lot of balls in short range and runs for the first down or gets open in the intermediate range for first downs and adds yards after the catch, adding a fair number of TDs in the process. Less glamorous than the highlight reel exploits, but no less valuable. Yards, first downs and TDs have the same value regardless of how their achieved.

And again, he blew everybody away on a per target basis last season.

Wow, incredible stats. Thanks for taking time to put this together.

Good point about what makes the highlight reel v what really counts. I never would have guessed his YAC % to be that high. It's one thing to catch the ball, a hard enough task, but then to turn it into more, a lot more, is where a lot of a receiver's value lies. Thanks again. I'm glad he's playing for us!