Projecting 2015 season

[Luca]

I'm highly suspicious of yards-per-coverage-snap data.

Passing offenses are "right handed" (or "left handed" in the rare cases of lefty QBs; the rest of these observations will assume a righty QB). I can't recall if I saw Rodgers' intermediate-to-deep 2014 throws charted in a link posted here or elsewhere, but even his throw frequency and completion rate on those passes is predominantly to the right side. The difference was quite striking, in fact. It's just plain more difficult to turn and throw left in a mechanically sound position, i.e., not across the body, than it is to the right side...there's simply fewer degrees of turn going right.

Further, the QB's vision dropping back from center is to the right side. They don't call the left side the "blind side" for nothing; if the QB is blind to the rusher, he's bound to be at least somewhat blind to the routes on that side as well.

You must be logged in to see this image or video!

You are right. Rodgers has been more accurate and graded better to the right side than the left side in 2014 (although his best area is 10-19 yards in the middle of the field) This effect also shows in 2012 and 2013. However this effect doesn't show up in the stats for 2009 and 2010. I am not sure why? Maybe our receiving core has something to do with it?

 

[HardRightEdge]

You must be logged in to see this image or video!

You are right. Rodgers has been more accurate and graded better to the right side than the left side in 2014 (although his best area is 10-19 yards in the middle of the field) This effect also shows in 2012 and 2013. However this effect doesn't show up in the stats for 2009 and 2010. I am not sure why? Maybe our receiving core has something to do with it?

For the purposes of my argument, the data would be more interesting if the the field was split right down middle into purely left and right. Or better yer, if it had 4 vertical columns with the middle split in half. Throwing to the right hash is marginally easier than throwing to the left hash, so it would be interesting to see how those middle throws are distributed.

That's not a criticism of your presentation...I get that's the way the data is typically compiled.

Why would '09 and '10 be different? Dunno. How different are they? Perhaps as the years go by he plays more to his strengths with fewer experiments at finding his limits. Hard to say.

It would be interesting to see how a more mundane QB does...Rodgers is remarkable no matter where he throws the ball, even if there are marginal differences.

Would it be possible for you to do a chart for Cam Newton in 2014? He finished smack in the middle of the QB ratings among qualifying QBs, and he's known for accuracy issues.

 

[Luca]

Why would '09 and '10 be different? Dunno. How different are they? Perhaps as the years go by he plays more to his strengths with fewer experiments at finding his limits. Hard to say.

You must be logged in to see this image or video!

Here are the figures for 2010. As you can see the difference in grades and especially QB-rating for the left and right side are absent in this one (2009 looks similar). I am not sure why? It's just an observation!

 

[HardRightEdge]

You must be logged in to see this image or video!

Here are the figures for 2010. As you can see the difference in grades and especially QB-rating for the left and right side are absent in this one (2009 looks similar). I am not sure why? It's just an observation!

A key point of my original argument was not just performance in left vs. right throws. Equally important in considering yards-coverage-snap with left corners vs. right corners is the frequency of throws left vs. right. Even in this 2010 data we see a meaningful frequency difference in right over left, just as with the 2014 data. The short yardage right throws with 6 TDs looks like a bread and butter throw in the red zone that season.

 

[Deleted member 6794]

Actually, yards-per-coverage-snap covers only half of what goes into the passer rating. There are big differentiators in the passer rating calculation involving TD and INT frequency. Those should be regarded as just as important with CBs as with QBs.

Can you name the other 3 cornerbacks and their respective ranks? I remain skeptical of statistics that make a positive assumption about why a player is not involved in a play.

Following on this logic, we should then rate receivers on yards-per-pass-snap. Do you endorse that?

Perhaps the best answer to this complex question is to look at both yards-per-coverage snap and passer-rating-against simultaneously...perhaps in a blended quotient. With the latter gauge, we see how he does when the ball is thrown at him, which supports or debunks the assumption about why offenses don't target him as often, or more often, as the case might be.

I´m sorry but there were only six cornerbacks ranked in both top 10. The other four were Demetrius McCray (4th in yards per coverage snap, 34th in passer rating against), Rashean Mathis (7th, 47th), Darrelle Revis (8th, 13th) and Xavier Rhodes (9th, 24th).

I agree that we shouldn´t take yards allowed per coverage snap as the only measurement of a cornerback´s performance but it´s a nice indicator combined with some other stats.

BTW it also works for wide receivers as well. Last season the top 10 included A.J. Green, Demariyus Thomas, Odell Beckham Jr, Julio Jones, Dez Bryant, Antonio Brown, Jordy Nelson, Emmanuel Sanders, T.Y. Hilton and Calvin Johnson.

 

[Sunshinepacker]

I think that yards allowed per coverage snaps in combination with TDs and interceptions is a pretty good indicator of a cornerback's performance.

But then the question becomes, "was he covering the #1 receiver on each snap?"...all are good indicators and all give a great idea. I'm just saying that it's somewhat imprecise to try and figure these things out with the limited information that most fans get.

 

[Sunshinepacker]

For those interested in discussing cornerback performance, Football Outsiders has a terrific article up right now that delves into best starting corners ranked by adjusted success rate, defined as: share of targets on which the corner prevented a successful gain (45 percent of needed yards on first down, 60 percent on second down, and 100 percent on third down). Sam Shields, by this metric, ranks as the 7th best corner in the NFL (among #1 ccorners). Pretty interesting article. Certainly many could argue that interceptions should count more than just preventing a successful gain but I would argue that interceptions tend to be unpredictable and having a consistent corner that can simply prevent the offense from succeeding is a good thing. Anyway, here's the link for the article:

http://www.footballoutsiders.com/stat-analysis/2015/best-cornerback-stats-2014

 

[HardRightEdge]

For those interested in discussing cornerback performance, Football Outsiders has a terrific article up right now that delves into best starting corners ranked by adjusted success rate, defined as: share of targets on which the corner prevented a successful gain (45 percent of needed yards on first down, 60 percent on second down, and 100 percent on third down). Sam Shields, by this metric, ranks as the 7th best corner in the NFL (among #1 ccorners). Pretty interesting article. Certainly many could argue that interceptions should count more than just preventing a successful gain but I would argue that interceptions tend to be unpredictable and having a consistent corner that can simply prevent the offense from succeeding is a good thing. Anyway, here's the link for the article:

http://www.footballoutsiders.com/stat-analysis/2015/best-cornerback-stats-2014

Not giving some additional weight to INTs is a distortion to some degree. I agree that a team level in any particular game, especially in the playoffs, leaning on ball hawking is a dubious proposition. But over the course of a season and when looking at individual players, INTs should factor in some way, otherwise the value associated with ball skills is diminished. There's the old saw that says some DBs became DBs and not wide receivers because they have bad hands. We see that when a DB drops a ball thrown right at him, a not so uncommon occurrence. (It seems more notable with LBs, but that's another story). TDs surrendered are not considered either, which seems counter intuitive.

There's an interesting caveat in this piece that goes to a point not previously discussed:

"We also report the number of targets that each cornerback faced, as well as a metric we call "estimated target percentage." This stat uses snap counts to estimate what percentage of possible targets were thrown at this player when he was on the field. The "possible targets" part of that metric leaves out the passes noted in the previous paragraph, as well as those passes listed with "Uncovered" or "Blown Coverage," though the total of possible targets does include "Hole in Zone" passes.

Uncovered and blown coverage passes are sensibly omitted if responsibility cannot be assigned according to the eye test. One would hope other purveyors of analytics take the same conservative approach, and don't just make a guess and force a responsibility. But somebody on the field usually often is to blame, even if it's impossible to tell without knowing the defensive call, acknowledging that at times the problem is the call, no the players. To some degree, players who more frequently blow an assignment so badly he can't be assigned blame will get away with a better rating than those that don't.

This also begs the question of how "hole in zone" passes are assigned responsibility whereas it is not with uncovered and blown coverage throws. That's a fine line. There are surely guys who are better in man than in zone; one supposes the argument for including "hole in zone" throws would not reflect a guy's weakness in zone. Nonetheless, there is a question here that goes to consistency and reliability in the stats based on the one analyst's "eye of the beholder" calls compared to another's when judging responsibility in zone.

One thing I like particularly in this approach is counting pass interference throws against the DB. Too much NFL data omits penalties.

 

[Half Empty]

Even I can determine if he gets five or more INTs, but if someone digs this out next year, what constitutes having 'a better season'?

So, apparently, whoever favors one CB will be able to prove their point, but so will the person backing the other one?

 

[HardRightEdge]

I´m sorry but there were only six cornerbacks ranked in both top 10. The other four were Demetrius McCray (4th in yards per coverage snap, 34th in passer rating against), Rashean Mathis (7th, 47th), Darrelle Revis (8th, 13th) and Xavier Rhodes (9th, 24th).

I agree that we shouldn´t take yards allowed per coverage snap as the only measurement of a cornerback´s performance but it´s a nice indicator combined with some other stats.

BTW it also works for wide receivers as well. Last season the top 10 included A.J. Green, Demariyus Thomas, Odell Beckham Jr, Julio Jones, Dez Bryant, Antonio Brown, Jordy Nelson, Emmanuel Sanders, T.Y. Hilton and Calvin Johnson.

9 of those 10 receivers, ranked in the top 10 in yds. per game. The difference was Golden Tate moving to #8 while Green drops to #13, a function of Detroit throwing the ball more than Cincy. It's hard to see a value-add in looking at yards-per-passing-snap vs. this more pedestrian stat.

http://espn.go.com/nfl/statistics/player/_/stat/receiving/sort/receivingYardsPerGame

I've highlighted previously Cobb's outrageously good stealth 2014 productivity-per-target numbers, which I have not seen highlighted anywhere else. He was 12th. in yards per game; he would be similarly ranked in yards per team pass attempt.

Looking at those top 13 yards-per-game leaders (which include the 10 names you listed), Cobb ranked as follows:

catches % per target: #1 (72.2%)
yards per target: #2 (10.21) vs. T.Y. Hilton #1 (10.35)
TDs per target: #2 (0.095) vs. Dez Bryant #1 (0.116)
+20 yard gains per target: #1 (0.190)
yards after catch per target: #1 (4.41)
first downs per target: #1 (0.563)

I've highlighted the last number because when I first commented on Cobb's productivity a few months ago, I was inclined to dig into these numbers based on the simple reflection that the frequency of "complete...Cobb...first down" calls seemed to have been awfully high in my recollection.

There is no doubt in my mind that Cobb was the best receiver in the NFL last season. The disparity in the quality of nickel corners over cover corners around the league does not account the wide disparities. This collection of outstanding possession receiver rankings with outstanding down-the-field receiver rankings are so remarkable as to be perhaps a once every 5 or 10 year phenomenon.

To draw a baseball analogy, Cobb's season is akin to George Brett's in 1980 when he played only 117 games. The gross numbers are not outrageous, but the production per opportunity is off the charts.

 

[Deleted member 6794]

9 of those 10 receivers, ranked in the top 10 in yds. per game. The difference was Golden Tate moving to #8 while Green drops to #13, a function of Detroit throwing the ball more than Cincy. It's hard to see a value-add in looking at yards-per-passing-snap vs. this more pedestrian stat.

http://espn.go.com/nfl/statistics/player/_/stat/receiving/sort/receivingYardsPerGame

I've highlighted previously Cobb's outrageously good stealth 2014 productivity-per-target numbers, which I have not seen highlighted anywhere else. He was 12th. in yards per game; he would be similarly ranked in yards per team pass attempt.

Looking at those top 13 yards-per-game leaders (which include the 10 names you listed), Cobb ranked as follows:

catches % per target: #1 (72.2%)
yards per target: #2 (10.21) vs. T.Y. Hilton #1 (10.35)
TDs per target: #2 (0.095) vs. Dez Bryant #1 (0.116)
+20 yard gains per target: #1 (0.190)
yards after catch per target: #1 (4.41)
first downs per target: #1 (0.563)

I've highlighted the last number because when I first commented on Cobb's productivity a few months ago, I was inclined to dig into these numbers based on the simple reflection that the frequency of "complete...Cobb...first down" calls seemed to have been awfully high in my recollection.

There is no doubt in my mind that Cobb was the best receiver in the NFL last season. The disparity in the quality of nickel corners over cover corners around the league does not account the wide disparities. This collection of outstanding possession receiver rankings with outstanding down-the-field receiver rankings are so remarkable as to be perhaps a once every 5 or 10 year phenomenon.

To draw a baseball analogy, Cobb's season is akin to George Brett's in 1980 season when he played only 117 games. The gross numbers are not outrageous, but the production per opportunity is off the charts.

In addition to the stats you posted Cobb led the league in QB rating when thrown to last season with Jordy Nelson ranking third. IMO Cobb is without a doubt the best slot receiver in the NFL but I wouldn´t call him the best overall WR in the game.

 

[Raptorman]

In addition to the stats you posted Cobb led the league in QB rating when thrown to last season with Jordy Nelson ranking third. IMO Cobb is without a doubt the best slot receiver in the NFL but I wouldn´t call him the best overall WR in the game.

I'd like to see the stats to back that up. Were did you get that info? Just curious.

 

[HardRightEdge]

In addition to the stats you posted Cobb led the league in QB rating when thrown to last season with Jordy Nelson ranking third. IMO Cobb is without a doubt the best slot receiver in the NFL but I wouldn´t call him the best overall WR in the game.

That Cobb would have the best QB rating on passes thrown to him is no surprise given the other per target stats. The only thing missing in the stats I cited would have been INTs on Cobb's targets. Rodgers didn't throw many at all, so it stands to reason that would work in his favor.

I did not and would not call Cobb the best WR in the game, at least not at this juncture, just as I would not have called Brett the best slugger in the game just because he led the majors in slugging percentage in 1980. I'm calling Cobb the best WR in the game in 2014.

Rodgers was the NFL MVP last season; few think twice about the fact that Luck threw for more yards and TDs. That's because we are all (not just Packer fans) conditioned to think in terms of QB Rating as the primary measure, which is predicated in all aspects on per throw data, not gross data. It helps that per throw data makes a lot of sense. We recognize the fact that Luck threw 18% more balls than Rodgers requires adjustments.

With receivers, it is far less common to think in terms of production per target.

Antonio Brown led the league in gross yards while adding 13 TDs. He was named first team All Pro with Dez Bryant. So, many folks, I would surmise, think of him as the best receiver in 2014. But for some reason, the fact that Brown had a staggering 44% more targets than Cobb gets little if any consideration. The fact that Roethlisberger threw 15% more passes than Rodgers, elevating gross number opportunities across the board for his receivers, should also be a strong consideration but typically is not.

If receivers were evaluated in a way similar to QBs, Cobb would have been the "most valuable" in 2014.

 

[HardRightEdge]

I'd like to see the stats to back that up. Were did you get that info? Just curious.

Personally, I don't see the need. Cobb's completion % per target, TD% per target and yards gained per target cover 3 of the 4 QB rating factors, and his numbers are so high (as noted above) in those 3 categories as to give little room to question it. The only thing missing is INTs per target. I'd assume that to be very low given that Rodgers threw only 5 all season.

 

[Deleted member 6794]

I'd like to see the stats to back that up. Were did you get that info? Just curious.

The stats are from Pro Football Focus. Rodgers had a 134.3 QB rating when targeting Cobb and 128.2 when throwing to Nelson. Cole Beasley is second at 130.6.

 

[Deleted member 6794]

That Cobb would have the best QB rating on passes thrown to him is no surprise given the other per target stats. The only thing missing in the stats I cited would have been INTs on Cobb's targets. Rodgers didn't throw many at all, so it stands to reason that would work in his favor.

Rodgers threw one interception last year when targeting Cobb.

With receivers, it is far less common to think in terms of production per target.

Well, the ability to get open factors into the total amount of targets. Maybe that's a reason why not a lot of people look at production per target with receivers. The total amount of passes thrown by a team should be considered though.

 

[HardRightEdge]

Rodgers threw one interception last year when targeting Cobb.

Well, the ability to get open factors into the total amount of targets. Maybe that's a reason why not a lot of people look at production per target with receivers. The total amount of passes thrown by a team should be considered though.

Well...if you bump Cobb's gross numbers across the board by 15 - 18%, the differential in passes thrown by Rodgers vs. Roethlisberger or Luck, Cobb's superior value over Brown and Hilton, for example, would be more apparent, if not obvious.

 

[Deleted member 6794]

Well...if you bump Cobb's gross numbers across the board by 15 - 18%, the differential in passes thrown by Rodgers vs. Roethlisberger or Luck, Cobb's superior value over Brown and Hilton, for example, would be more apparent, if not obvious.

In that case Cobb would have had 106 catches for 1,502 yards with 14 TDs. Great numbers but aside of the touchdowns still behind Antonio Brown.

 

[HardRightEdge]

The stats are from Pro Football Focus. Rodgers had a 134.3 QB rating when targeting Cobb and 128.2 when throwing to Nelson. Cole Beasley is second at 130.6.

Citing Cole Beasley diminishes the significance of this stat. Beasely had only 49 targets, tied for 136th. in the league. That would make him something less than a #4 team target in a league average.

There should be a minimum number of opportunities to get to a meaningful number.

There were 49 receivers with 100 or more targets last season; 39 with 110 or more. A reasonable cut-off should be somewhere around this range.

 

[Raptorman]

The stats are from Pro Football Focus. Rodgers had a 134.3 QB rating when targeting Cobb and 128.2 when throwing to Nelson. Cole Beasley is second at 130.6.

Not to be to picky, but I want to see if my numbers are correct. Cobb had 127 targets, 91 catches. 1287 yards and 12 TD's. Flynn threw to him 2 times for 18 yards. Now no matter how I put the number in the QB Rating calculator I cannot come up with that number. Even if I put in the post season numbers it only comes up to 132.9. Interesting. Still damn good.

Anyway, it's close enough. I take it this is from the "pay" portion of PFF.

 

[Deleted member 6794]

Citing Cole Beasley diminishes the significance of this stat. Beasely had only 49 targets, tied for 136th. in the league. That would make him something less than a #4 team target in a league average.

There should be a minimum number of opportunities to get to a meaningful number.

There were 49 receivers with 100 or more targets last season; 39 with 110 or more. A reasonable cut-off should be somewhere around this range.

You're right about that. Taking only the top 50 receivers in targets into consideration Cobb and Nelson are the top ranked WRs in the league.

 

[Deleted member 6794]

Not to be to picky, but I want to see if my numbers are correct. Cobb had 127 targets, 91 catches. 1287 yards and 12 TD's. Flynn threw to him 2 times for 18 yards. Now no matter how I put the number in the QB Rating calculator I cannot come up with that number. Even if I put in the post season numbers it only comes up to 132.9. Interesting. Still damn good.

Anyway, it's close enough. I take it this is from the "pay" portion of PFF.

According to PFF Cobb was targeted a total of 125 times.

 

[Raptorman]

According to PFF Cobb was targeted a total of 125 times.

By Rodgers. He was 2 for 2 from Flynn.

 

[Deleted member 6794]

By Rodgers. He was 2 for 2 from Flynn.

No, their total of 125 targets includes passes by Flynn as well.

 

[Raptorman]

No, their total of 125 targets includes passes by Flynn as well.

Interesting. Everyone else has either 127 or 126. That would do it though. Thanks.