PFF gives Aaron Rodgers negative grade after 5 TD game

[Recon_Hawk]

Pro Football Focus is taking some heat after first giving Aaron Rodgers a -2.3 score for his performance against the Chiefs (which has since been changed to a -0.8 because of an "input error") in a game where Rodgers threw for 5 TDs and scored a 138.5 QB rating.

Here is their reasoning:

Why Aaron Rodgers’ grade was just average versus Chiefs

Packers quarterback Aaron Rodgers ended last night’s game with a -0.8 grade overall. This isn’t a bad game, just because the number begins with a minus, but it is an average grade very close to zero for a player who threw five touchdown passes, which seems crazy on the face of it. It’s not.

On the surface, Rodgers’ raw statistics paint the picture of one of the best games of the season. 333 passing yards, five touchdown passes, no interceptions, a 138.5 passer rating; Rodgers’ should be supplanting Carson Palmer in our team of the week as the top quarterback, not earning a grade with a minus in front of it, right?

Well, not if you dig a little deeper into Rodgers’ performance on a play-by-play basis.
https://www.profootballfocus.com/blog/2 ... ive-grade/

My question is, should this raise huge doubts about the "model" that PFF uses to grade players or do you still have confidence in their methods, overall, and feel it's still the best objective analysis out there?

 

[SeahawksFanForever]

They basically don't want to give Rodgers credit for run-after-catch TD's by Cobb. And, they are also talking about the "almost turnover" (fumble by Rodgers). System is flawed, I can maybe understand the fumble part even though GB recovered but Cobb TD's? C'mon man!

 

[Alexander]

They basically don't want to give Rodgers credit for run-after-catch TD's by Cobb. And, they are also talking about the "almost turnover" (fumble by Rodgers). System is flawed, I can maybe understand the fumble part even though GB recovered but Cobb TD's? C'mon man!

He also had a dropped pick six. Practically threw it right to the defender.

PFF will get a lot of heat for this, but I thought their explanation made sense. They admit that they can't quantify the intangibles that Rodgers brings to the offense, which is obviously a big part of his, or any QB's, game. I don't think he got plus points for drawing the defense offsides multiple times, but those plays had a major positive impact for GB.

Basically, I think PFF grades, like many stats, are limited in what they can tell you. They're best when complemented by other stats, or the keen eye of an observer.

 

[hawknation2015]

There are several major flaws in the way they grade QBs. The most glaring to me is the fact that they admit they are not giving the QB credit when he throws to WRs who are in a position to gain yards after the catch, as they did with three of Rodgers' TD passes. If a QB successfully manipulates a defense to put his receiver in a position to gain major yards after the catch, and then throws an incredibly safe pass to that receiver, how is that not a positive play for the QB? On such plays, he has accomplished everything you could possibly ask from a QB: exploiting a weakness in the defense, emphasizing a strength in his offense, scoring a TD, and taking care of the football.

Needless to say, PFF grades alone are virtually worthless. Their premium stats are the only thing they have worth a damn, and those are no longer being sold to new subscribers.

 

[theENGLISHseahawk]

This is the same PFF that had Earl Thomas as the 29th (or something like that) best safety in football in 2012.

It's shocking how it's developed the reputation it has peddling nonsense like this. Rodgers looked like the best player in the NFL last night. Average my arse.

 

[Recon_Hawk]

They basically don't want to give Rodgers credit for run-after-catch TD's by Cobb. And, they are also talking about the "almost turnover" (fumble by Rodgers). System is flawed, I can maybe understand the fumble part even though GB recovered but Cobb TD's? C'mon man!

I agree, it appears seriously flawed. What drives me a bit mad is there approach to explaining their grades. Yeah, they might point out a few plays, but why not every single play in detail? Why not show how and why they got the score they got instead of the "trust us" attitude.

I don't even understand their reasoning for the negative grade on the fumble play. Now, I did not watch the entire game, but from what I've read there was a defensive penalty that saved the packers from the turnover. Assuming that's correct, how can they say with certainty that this penalty did not affect Rodgers from running the play as called? It seems a bit contrived to grade him negatively on a play where the defense is playing outside of the rules.

 

[Jville]

Statistical formulas are tools that reflect its author's subjective model and preferences ..... and nothing more.

32 nfl teams construct their own unique formulas as tools to reflect their unique preferences and team model.

 

[Alexander]

They basically don't want to give Rodgers credit for run-after-catch TD's by Cobb. And, they are also talking about the "almost turnover" (fumble by Rodgers). System is flawed, I can maybe understand the fumble part even though GB recovered but Cobb TD's? C'mon man!

I agree, it appears seriously flawed. What drives me a bit mad is there approach to explaining their grades. Yeah, they might point out a few plays, but why not every single play in detail? Why not show how and why they got the score they got instead of the "trust us" attitude.

I don't even understand their reasoning for the negative grade on the fumble play. Now, I did not watch the entire game, but from what I've read there was a defensive penalty that saved the packers from the turnover. Assuming that's correct, how can they say with certainty that this penalty did not affect Rodgers from running the play as called? It seems a bit contrived to grade him negatively on a play where the defense is playing outside of the rules.

The defensive holding may well have affected what Rodgers did in the pocket, but in watching the play unfold, I think they were right to grade him negatively on the play. He held on to the ball for way too long with a defender in striking distance. I was frankly a little surprised he held it that long without trying to at least bail out of the pocket. It was a rare instance where his pocket presence failed him.

Having said that, I think you could argue for the play not counting at all towards the grade, since it was nullified by a penalty.

The main issue I have with PFF is that they're not transparent. As you pointed out, they only highlight a few plays to make their point. If they really have faith in their grading, making their play-by-play grades available would go a long way to building their credibility.

 

[hawknation2015]

The main issue I have with PFF is that they're not transparent. As you pointed out, they only highlight a few plays to make their point. If they really have faith in their grading, making their play-by-play grades available would go a long way to building their credibility.

Instead, they are going in the opposite direction in eliminating all their premium stats, which was the only useful thing about that site. There will now be way less transparency, with nothing but naked grades for each player and no supporting statistics.

 

[ringless]

I agree that their stats while useful are still lacking the entirety of the picture.

Here is an example. Anquan Boldin has a higher rating than Larry Fitzgerald after 3 weeks this season. I guess I just don't understand everything about the site or it's grading and never really have. Why Fitzgerald has a lower grade I just don't know.

Anquan Boldin
12 Receptions
112 Yards
37.3 Yds per game
1 TD

Larry Fitzgerald
23 Receptions
333 yards
111.0 Yds per game
5 Td's

Plus he has been one of our best blockers on runs.

From PFF
When Carson Palmer has thrown to Larry Fitzgerald, he’s had an NFL passer rating of 155.8 which is the best for a quarterback receiver duo with at least 15 targets. Fitzgerald is also the receiver with the most receptions (23) with no drops.

 

[RolandDeschain]

It would seem that PFF is actually starting to follow in the footsteps of Football Outsiders more with their ratings, based on this.

If true, it's only a good thing. Everyone here agrees that it's not a QB's fault when a receiver bounces a well-thrown pass off his hands and into the arms of a defender, so why shouldn't we hold it against a QB when a defender jumps what should have been a surefire interception due to a poor throw (or poor decision to throw in the circumstances) by a QB?

 

[kearly]

It's hypocrisy to not give credit for YAC while simultaneously assuming that a dropped interception should have been returned for six. If you aren't going to account for YAC, then you shouldn't account for interception yards either. And accounting for 'should have been' interceptions is the realm of subjectivity anyway.

They should also give the final statline at least some weight so as to help iron out some of the flaws in their grading system.

Assigning a PFF grade for a QB is kinda uneccessary anyway. QB is one of the easier positions to evaluate with simple statistics like adjusted net YPA, etc.

 

[hawknation2015]

It would seem that PFF is actually starting to follow in the footsteps of Football Outsiders more with their ratings, based on this.

If true, it's only a good thing. Everyone here agrees that it's not a QB's fault when a receiver bounces a well-thrown pass off his hands and into the arms of a defender, so why shouldn't we hold it against a QB when a defender jumps what should have been a surefire interception due to a poor throw (or poor decision to throw in the circumstances) by a QB?

PFF has always done this.

 

[RolandDeschain]

PFF has always done this.

Nowhere near to this extent that I can recall seeing. I think they're weighting "got lucky that wasn't a disaster" situations a lot more highly now. Do you have evidence to the contrary?

 

[hawknation2015]

PFF has always done this.

Nowhere near to this extent that I can recall seeing. I think they're weighting "got lucky that wasn't a disaster" situations a lot more highly now. Do you have evidence to the contrary?

Ever since I have followed them, for the last 2+ years at least, they have given a negative grade to throws that are nearly intercepted and not positive grade to yards after the catch by the receiver.

The problem with the 2nd category has always been those plays where the QB anticipates a running lane for the receiver. Of course, there are fluke plays where the receiver breaks five tackles, but that is not what happened on those three Rodgers' TDs. They need to do a better job differentiating the "expected" YAC, i.e. Cobb to the outside, vs. the "fluke" YAC.

 

[RolandDeschain]

I know they've done that, the question is how much of a negative grade for stuff like that. It seems like it has been increased considerably, though they never gave hard numbers behind their ratings before and won't now, though they did confirm they have reworked their rating system in some ways.

Just a guessing game at this point, I suppose.

 

[Smellyman]

You could always use your eyballs instead.

 

[hawknation2015]

I know they've done that, the question is how much of a negative grade for stuff like that. It seems like it has been increased considerably, though they never gave hard numbers behind their ratings before and won't now, though they did confirm they have reworked their rating system in some ways.

Just a guessing game at this point, I suppose.

Their rating system has been pretty much the same for the last few years with a "near interception" receiving around a -1.5 per play. That was important for their purposes because they often compared player ratings to years past.

Soon, they will completely change the way they grade players . . . moving to what I guess is more of a "Madden style" 100-point system. That corresponds with their elimination of all their signature stats, i.e. pressures allowed, snaps per reception, run-blocking grades, etc. Their site is headed in the totally wrong direction.

 

[RolandDeschain]

I've never been a fan of their ratings, but I last really followed them like 3+ years ago. I'll wait to judge until the new system's out and "tweaked", lol.

Either way, the only ratings system I put any real faith in is DVOA.

 

[hawknation2015]

I've never been a fan of their ratings, but I last really followed them like 3+ years ago. I'll wait to judge until the new system's out and "tweaked", lol.

Either way, the only ratings system I put any real faith in is DVOA.

Saying someone is an "87.5" means nothing. Just like saying someone was a +4.4 meant very little on its own. It was the premium stats that gave the site value. Without those, I certainly won't be subscribing to them.