If you like PFF, you will not like this.

[thebandit27]

 

 

i dont think thats unfair commentary. anyone touting PFF as having every grade perfect is silly. im betting the sites creators dont even argue that. id also guess some of the unfair blame gets balanced out with undeserved credit and across the league most are subject to the same troubles. a degree of errors washing out a bit as sample sizes grow and you are comparing to a large group of guys. high scores should generally be good, and low scores bad, and a whole mess in between.

 

theres also certainly potential for a reviewer to have a fundamental misunderstanding and hurt a certain position in a certain scheme too but with widespread internet review and scrutiny id guess over time that would be picked up on.

 

but no stats tell the whole story definitively - so with that in mind im not sure they are worthless, but just another piece to the puzzle.

 

I wouldn't call it worthless; I would say that their metrics are to be taken with a very large grain of salt.

 

There are some positions that are easier to grade than others, like RB for instance. When it gets to trying to grade linemen on either side of the ball, it's much more subjective.

[NoSaint]

Exactly, well said.

 

Like I said I think the grade PFF is currently giving Pears, Urbik, and the rest of the line is far, far to generous compared to what they are actually accomplishing.

 

Look at the numbers for the last three games, and the drive starts yards speak for themselves. I don't need to look at PFF to see the Bills are currently fielding one of the very worst offenses in the NFL.

Then think about exactly why that is

 

we are 27th in yards per play, 30th in first downs per game, 26th in 3rd down conversions, 32nd in red zone TD percentage.... we have bumped up in the fumbles per game and are no longer last, so thats good.

 

but year, its not shocking we are around the bottom on PFF for offense.

[BuffOrange]

Belichick on PFF:

 

http://www.bostonspo...blindly-trusted

 

That article makes sure to say "blindly" 1,100 times. GMAFB. What is the author's alternative for player/team evaluation - the box score? The redzone channel? Parroting idiot announcers' cliches about momentum?

 

Whether its PFF or Football Outsiders or Kenpom in basketball; analytics opponents always hold these impossibly high standards of perfection that no "eyeball test" is ever held to. Removing emotion from the equation is a good thing. Saying a 37yr old all-time great is in decline - which he is if you watch and document his deep throws - should not be controversial, but obviously Boston fans get their panties in a bunch and interpret that as "OMG they're saying he stinks like Geno Smith" which of course nobody said.

Edited November 18, 2014 by BuffOrange

[K-9]

Sites like PFF, and Football Outsiders have a bad rep with most fans here simply because they post what most Bills fans don't want to see. Which is negative stats about their favorite team ...

 

Disagree entirely. It's not about negative stats. It's about stats compiled completely OUT OF CONTEXT and passed off as some sort of insightful absolutes.

 

FearthePFF.

 

GO BILLS!!!

[zonabb]

That article makes sure to say "blindly" 1,100 times. GMAFB. What is the author's alternative for player/team evaluation - the box score? The redzone channel? Parroting idiot announcers' cliches about momentum?

 

Whether its PFF or Football Outsiders or Kenpom in basketball; analytics opponents always hold these impossibly high standards of perfection that no "eyeball test" is ever held to. Removing emotion from the equation is a good thing. Saying a 37yr old all-time great is in decline - which he is if you watch and document his deep throws - should not be controversial, but obviously Boston fans get their panties in a bunch and interpret that as "OMG they're saying he stinks like Geno Smith" which of course nobody said.

 

Hard to take your thoughts on statistics and data seriously when this is your measure for decline... deep throws. So by your logic, because he can't throw deep as well as before, he's in decline. Like it or not as a measure of anything, but it does offer a standard measure of a QB over time, his current rating is a 102.0, fourth highest of his career. And by the way I was taught statistics in grad school, decline is a constant, that is year over year the variable being measure declines in value. Not so in Brady's case. He did however decline, again if we only use the rating, the prior four seasons. But his overall play now is certainly admirable and maybe, just maybe, because throwing the deep ball is not required on every play, we can't and shouldn't consider him in decline because he doesn't do that one aspect of his job well while performing well in all others??

 

And the reason why these measures are, and should be, held to a higher standard than the eyeball test is the reason why statistics we developed in the first place... the eyeball test is worthless and wildly inaccurate and imprecise. In my mind (of two graduate degrees), most measures have very little value other than as simple descriptive stats that are not connected in any statistically significant way to wins and losses. Stats that matter, and maybe they exist but I don't waste my on frivolous indulgence in football stats, would be ones that could model or predict wins and losses and outcomes on given plays based on a set of variables. Show me a method that uses a standardized method for grading all offensive lineman on a play and how their individual scores predict running outcomes and you're on to something. Otherwise, it's basically a bunch of data miners selling "magic" to those who are easily wowed. Give this garbage to any college statistics instructor and it surely gets what it deserves, a good chuckle.

 

I think the worst part is, it's being taken so religiously by so many (including most of the blowhards at WGR55), that it's actually doing a disservice to high-level statistical analysis. The reason why so little exists in academic literature for this type of NFL analysis is because the game is so random with so many moving parts, that it's hard to predict much. Go on Google Scholar (free) and look. There's actually a lot more interesting work on predicting player success.

[filthymcnasty08]

It's all relative……and on offense….we're relatively horrific when held up to the rest of the league

 

I don't need PFF to tell me that...but the numbers are there.

[BuffaloBillsForever]

@MatthewFairburn: The Buffalo Bills now have the worst cumulative score on Pro Football Focus of any team in the NFL. They are -110.9.

 

--->>> offense only

 

Dynamic playmaker. That’s what this game is all about, making plays. This game is about making plays and surrounding our quarterback with playmakers. He’s automatically going to make our quarterback better and us better as a team. I do want to start this by saying this all goes back to Russ Brandon. He allowed us to make this move and he okayed it and was willing to give up the resources and future picks to go up and get a guy that we think will help us get to where we need to be.

 

Well that didn't quite happen this year Doug and Russ.

[BuffOrange]

Hard to take your thoughts on statistics and data seriously when this is your measure for decline... deep throws. So by your logic, because he can't throw deep as well as before, he's in decline. Like it or not as a measure of anything, but it does offer a standard measure of a QB over time, his current rating is a 102.0, fourth highest of his career. And by the way I was taught statistics in grad school, decline is a constant, that is year over year the variable being measure declines in value. Not so in Brady's case. He did however decline, again if we only use the rating, the prior four seasons. But his overall play now is certainly admirable and maybe, just maybe, because throwing the deep ball is not required on every play, we can't and shouldn't consider him in decline because he doesn't do that one aspect of his job well while performing well in all others??

 

And the reason why these measures are, and should be, held to a higher standard than the eyeball test is the reason why statistics we developed in the first place... the eyeball test is worthless and wildly inaccurate and imprecise. In my mind (of two graduate degrees), most measures have very little value other than as simple descriptive stats that are not connected in any statistically significant way to wins and losses. Stats that matter, and maybe they exist but I don't waste my on frivolous indulgence in football stats, would be ones that could model or predict wins and losses and outcomes on given plays based on a set of variables. Show me a method that uses a standardized method for grading all offensive lineman on a play and how their individual scores predict running outcomes and you're on to something. Otherwise, it's basically a bunch of data miners selling "magic" to those who are easily wowed. Give this garbage to any college statistics instructor and it surely gets what it deserves, a good chuckle.

 

I think the worst part is, it's being taken so religiously by so many (including most of the blowhards at WGR55), that it's actually doing a disservice to high-level statistical analysis. The reason why so little exists in academic literature for this type of NFL analysis is because the game is so random with so many moving parts, that it's hard to predict much. Go on Google Scholar (free) and look. There's actually a lot more interesting work on predicting player success.

 

I don't even...

You're all over the place. Some of that rant sounds like Rush Limbaugh yelling at a conservative.

 

The Football Outsiders/Advanced Statheads that you have such venom for are the ones constantly reminding others that randomness exists to a larger degree than most realize - your insistence otherwise is bizarre and illustrates your lack of understanding for their work.

There are absolutely models that predict W/L's - how do you think Vegas Over/Under Win totals are determined before the season? They know who was better or worse than their record last year - fumble recovery rates and record in 1score games are always good places to start.

 

For the Brady part it almost sounds like you're trolling yourself. You're acknowledging the declining rating over multiple years and I guess saying the last month is a more significant sample. Okie Dokie. If he doesn't do anything better than he did years ago and is worse in an important area (an area that's usually the first to decline with father time), then yes, my insane logic concludes from that he's in decline. I'm nutty like that.