PatsFans.com Menu
PatsFans.com - The Hub For New England Patriots Fans

Dead time thread debate: Pro Football Focus


Status
Not open for further replies.
PFF has come a long way and offers some useful data. Some of their stuff is garbage, like any other stat or trend. For me, it’s as simple as that.
 
PFF has a good concept. Teams grade players on film all the time.
But from there PFF has many issues, including:
1) historically they aren’t very good at grading football players
2) they have a flawed method such as giving a QB more credit for thriving into heavy coverage and getting lucky with a completion than making the proper read and hitting a wide open receiver
3) they try to put a number on things that there isnt a number for, and use it to compare players wholly, such as the horribly flawed passer rating against
4) they do it know the play call of responsibilities yet assign credit or blame any way by guessing.
5) these and other flaws make their ranking inaccurate and sometimes flat out silly.

The concept has merit but the approach they use invalidates most of their results.
 
That’s odd, the site I took it from said the article was 6 months old. My bad.
Just clarifying, that’s all. Not meaning to nitpick. I can remember 4 off the top of my head:

2004 Halloween (broke our long streak)
2008 Brady injury/Slater muffs the kickoff
2011 Polamalu intentionally batting the ball for safety
2018 Gordon’s last game

3 were in Pittsburgh, so Ben obviously does much better at home and struggles in Foxborough.
 
Love the site, but it’s not the end all as well as it’s not complete trash.

It’s come a long way and I really started to take notice when they started tracking qb pressures. In 2015, they had Trey Flowers as the 2nd highest in qb pressures in our the nation, which can help get you excited about a 3rd of 4th round pick. Cool stuff really, cb stats, percentages thrown at etc. Definitely has its uses.

However, I loathe to death that some
use the site to pump up their player but never mention the negative focus stats of same player.

It’s interesting for sure, just very subjective when they think they can grade an entire oline without taking note of how good the qb is thats running it. A high grouping grade is great, but how much does that qb understand pressure and reads etc? That goes a long way and it’s not really graded correctly imo, nor can it be. But the numbers are still fun all the same.
 
I read this article/ad this morning and all I really focused on was how this company was able to fund salaries for 500 employees plus fixed costs. It has become a massive venture that apparently is still spreading its wings. Reminds me of the old scouting service shared by most pro teams back in the day in order to minimize costs (and wins).
Clearly gambling entities are eager to pay for this data but I wonder how reliant the more independent teams are for this shared info.
 
That’s odd, the site I took it from said the article was 6 months old. My bad.
Come on, man. He threw more potential touchdowns than potential interceptions according to PFF. The **** is wrong with you?
 
At this point, you really have to consider them two separate services. The charting is very detailed, and every NFL team should (and apparently does) pay for it because it’s easier than hiring interns to do it.

The grades, rankings, and analysis, which they push much more aggressively to the general public, seem to be ignored by many teams. The Jets, for example, paid a ton of money to both Trumaine Johnson and CJ Mosley recently despite PFF grades far below their generally perceived value.
 
I remember when PFF started in 2003, I believe in Boston, after they contested dribble written by Ron Borges :D:D:D and somehow turned into somewhat of a NFL Stat powerhouse..

PFF is one of those "mezza mezza" things, some stuff they do is revealing and interesting.. others less so, the while DVOA stuff makes me scratch my head.. then there are the other acronyms which are confusing as well "DYAR", "ALEX" & "Pythagorean Projection" ...
 
Last edited:
When I want accurate and impartial analysis, I head straight to the GDT.

Anyone can handpick certain examples and deride the whole package.

It’s interesting analysis based on real numbers, it’s not supposed to be the final word in player evaluation.
 
PFF has a good concept. Teams grade players on film all the time.
But from there PFF has many issues, including:
1) historically they aren’t very good at grading football players
2) they have a flawed method such as giving a QB more credit for thriving into heavy coverage and getting lucky with a completion than making the proper read and hitting a wide open receiver
3) they try to put a number on things that there isnt a number for, and use it to compare players wholly, such as the horribly flawed passer rating against
4) they do it know the play call of responsibilities yet assign credit or blame any way by guessing.
5) these and other flaws make their ranking inaccurate and sometimes flat out silly.

The concept has merit but the approach they use invalidates most of their results.
1. That was probably true when the grading was done by a group of hobbyists, but think about it Andy. They have managed to retain all 32 NFL teams. Do you think think those teams would have signed up if they DIDN'T know the methodology PFF NOW uses to grade individual players AND agreed with it.

2. That is you opinion and actually may be the case. But if it's common knowledge and it must be if you think it, teams and individuals are free to use or ignore it.

3. That is an issue for ALL math driven data. They are all looking to create a "number" to describe an "action", and given there are 22 different interactions going on every play the complexity is enormous. Those numbers are an ATTEMPTS to explain that vast complexity. So by definition they are ALL going to be suspect.

4. That's a reality we have all recognized as a central problem for these kinds of services. But for a knowledgeable grader you can often infer responsibilities based on how you see the play unfolding. You certainly don't get it right 100% of the time, but that's not what teams are looking for. Those grades are merely indications or trends, or tendencies depending on what the grades are for.

Like I said in the OP, going in, there are lots of areas where you can criticized their operations. But their JOB is to be right more than they are wrong. Ultimately, the success of what they bring to the table depends greatly on how the user interprets the data that is given.

5. You last statement is just bluster and hubris, otherwise how do you explain that 32 NFL teams as well as 6o college teams use their services. Are they all idiots. If what your last statement is true....then that's what they are.
 
1. That was probably true when the grading was done by a group of hobbyists, but think about it Andy. They have managed to retain all 32 NFL teams. Do you think think those teams would have signed up if they DIDN'T know the methodology PFF NOW uses to grade individual players AND agreed with it.

2. That is you opinion and actually may be the case. But if it's common knowledge and it must be if you think it, teams and individuals are free to use or ignore it.

3. That is an issue for ALL math driven data. They are all looking to create a "number" to describe an "action", and given there are 22 different interactions going on every play the complexity is enormous. Those numbers are an ATTEMPTS to explain that vast complexity. So by definition they are ALL going to be suspect.

4. That's a reality we have all recognized as a central problem for these kinds of services. But for a knowledgeable grader you can often infer responsibilities based on how you see the play unfolding. You certainly don't get it right 100% of the time, but that's not what teams are looking for. Those grades are merely indications or trends, or tendencies depending on what the grades are for.
Like I said in the OP, going in, there are lots of areas where you can criticized their operations. But their JOB is to be right more than they are wrong. Ultimately, the success of what they bring to the table depends greatly on how the user interprets the data that is given.

5. You last statement is just bluster and hubris, otherwise how do you explain that 32 NFL teams as well as 6o college teams use their services. Are they all idiots. If what your last statement is true....then that's what they are.
1. You do not know that they signed up to get the methodology. In fact you don’t know if PFF charges them next to nothing and gives them perks to sign up just to be able to say all 32 teams subscribe. You also don’t know they are signing up for evaluation rather than data.
2. Of you are free to ignore their flawed evaluations then that makes them useless. You can’t argue something isnt flawed by saying you are free to ignore it because then it has no value. In that regard at least.
3. Exactly. So what value does a made up number purported to express something it doesn’t really have?
4. Huh? Are you telling me that you think teams are abandoning scouting and evaluation and turning it over to PFF? What good is rating a player if you don’t know his assignment? It’s a wild guess.
You seem to be arguing PFF is good because while its methodology is extremely flawed what they are trying to do has inherent flaws.
Its like saying throwing darts at a stock sheet has value because while it’s inherently flawed you can just be free to ignore the results or recognizes that it may be based upon guessing at data.
Now I’m sure if you studied and evaluated their evaluations you would find it better than throwing a dart, but to accept the end result as accurate would be throwing a dart, given multiple levels of flaws.
5.
Have you looked at their rankings?
Tom Brady was a unanimous mvp ranked 20th by pff. The argument was he didn’t complete enough passes into tight coverage. That is not bluster or hubris it’s a flawed rubric. There are many other examples.
You seem to think that teams subscribing to their service is an endorsement of the quality if their ratings. It’s far from that. I listed many reasons above that it may not be that at all.
Like I said it’s a noble concept but it’s not being executed very well (and maybe it just can’t be)
 
Last edited by a moderator:
Analyses on small sample sizes are difficult. News at 11.

Regards,
Chris
 
1. You do not know that they signed up to get the methodology. In fact you don’t know if PFF charges them next to nothing and gives them perks to sign up just to be able to say all 32 teams subscribe. You also don’t know they are signing up for evaluation rather than data.
2. Of you are free to ignore their flawed evaluations then that makes them useless. You can’t argue something isnt flawed by saying you are free to ignore it because then it has no value. In that regard at least.
3. Exactly. So what value does a made up number purported to express something it doesn’t really have?
4. Huh? Are you telling me that you think teams are abandoning scouting and evaluation and turning it over to PFF? What good is rating a player if you don’t know his assignment? It’s a wild guess.
You seem to be arguing PFF is good because while its methodology is extremely flawed what they are trying to do has inherent flaws.
Its like saying throwing darts at a stock sheet has value because while it’s inherently flawed you can just be free to ignore the results or recognizes that it may be based upon guessing at data.
Now I’m sure if you studied and evaluated their evaluations you would find it better than throwing a dart, but to accept the end result as accurate would be throwing a dart, given multiple levels of flaws.
5.
Have you looked at their rankings?
Tom Brady was a unanimous mvp ranked 20th by pff. The argument was he didn’t complete enough passes into tight coverage. That is not bluster or hubris it’s a flawed rubric. There are many other examples.
You seem to think that teams subscribing to their service is an endorsement of the quality if their ratings. It’s far from that. I listed many reasons above that it may not be that at all.
Like I said it’s a noble concept but it’s not being executed very well (and maybe it just can’t be)
I'm going to do my best to answer some of your responses to my approach. But be warned I am NOT going to get involved in one of your circular logic attempts to prove you are smarter than everyone else. But I'l make an honest attempt to respond here even though I know you aren't really intellectually honest enough to care whatever I write, because you are only geared to challenge and twist whatever I say to your own narrative. But here I go anyways, which probably denotes that there is a stroke in emanating in my near future. ;)

1. No I do not know the the financial arrangements that PFF has with the NFL, but NEITHER do you. I can INFER that they are they are making enough to pay 500 full time employees, most of whom have college degrees. They aren't hobbyists anymore. And your comment that they give their info to the NFL for little or free is an OPINION worthy of Ben Volin (sorry for the insult) Because it doesn't match the facts. If the NFL isn't paying for the services where are they coming up with the MILLIONS of dollars of fixed costs alone.

The second issue you didn't really address in #1 was about the methodology used to grade players. This was probably why I responded in the first place. Every team uses SOME kind of perimeters to grade an individual on any play. It is only common sense that ANY team interested in that grading service would NEED to know how they come up with the grade . And while it might not be how the individual team would do it, they CLEARLY must have found it reasonable enough or they wouldn't have bought the service.

2. If the data is as obviously flawed as you profess, why would anyone PAY to get it. Again how do you pay 500 white collar workers every month for YEARS, if what you say is true. SOMEBODY must think its worth something.

3. Finally something interesting to talk about. I guess we are talking about the very difficult ATTEMPT to quantify the sometimes random events of a football play NUMERICALLY. But often that's what mathematicians do. They make the ATTEMPT with the hope that what they produce will have enough truth that it aids in their clients' decision making.

We have all snickered on one hand at the varied incarnations of the QB rating to make a point, while on the other hand use it like a hammer to defend or attack a position. The point is that these numbers are just tools. The often carry a ring of truth in them, but are subject to vast ranges of interpretations.

For example, the instance you gave about Brady being marked down because he wasn't throwing into "tight windows". It's a great example of a flaw in the process that PARTICULAR year, btw what year was that? IIRC it was in '12, 13 or 14. But back a while. 2 points. While the number is clearly not the final arbiter of how a QB performs. For the most part guys who have good days have higher numbers than guys who don't.

But in a sense that example kind of makes my point. Do you want to bet that in 2019 QB's are NOT being punished for throwing to open receivers. You see, it seems you are trying to make your point by pointing out a flaw that may have occurred a while ago and deemed it to be the final picture of what will ALWAYS happen in the future. I;m sure that over the years PFF (and others) are constantly adjusting their numbers and parameters.

4. Now that was interesting and worthy of a fuller discussion that I'd be happy to have with you or anyone else. What follows is just BS. Here is an example of your over the top argumentative crap that you stoop to just for the sake of arguing

No I did NOT profess that the NFL is turning over all its "scouting and evaluations and turning over to PFF". I didn't infer it. I didn't even imply it. It was simply a figment of your imagination, and an example of why I mentioned "intellectual honesty" earlier. This is just you wanting to start an argument

In the rest of this number you yammer about how I think PFF Is good while its methodology is so clearly and obviously flawed. Its the central point of your entire rant, which is clearly and obviously flawed. How can that be when people who know a LOT more than you or me are paying the a LOT of money for their services. They aren't doing it because they believe the info they get is going to be right 100% of the time, but because they believe it will help them be right more times than not.

5 The drivel you wrote in #5 (outside of the Brady example which was interesting) is just more of the same crap and I'm getting bored. So I'll end it here and bemoan the time I lost writing it.

BTW- in crafting your response to this just answer 2 simple question with any kind of logic and I'll immediately change my position to yours.

1 Why did 32 NFL teams and 60 college teams purchase their services if they didn't think they'd get useful data.

2. Where is the money coming from to pay a MONTHLY multi-million dollar nut?
 
Last edited:
I'm going to do my best to answer some of your responses to my approach. But be warned I am NOT going to get involved in one of your circular logic attempts to prove you are smarter than everyone else. But I'l make an honest attempt to respond here even though I know you aren't really intellectually honest enough to care whatever I write, because you are only geared to challenge and twist whatever I say to your own narrative. But here I go anyways, which probably denotes that there is a stroke in emanating in my near future. ;)

Load of crap Ken. I’m done
 
Well the actual REASON I bothered to give you the link was something that was actually IN the article. They offered the opinion that all things being equal in today's game, you'd rather have good coverage than a great pass rush and they gave the Pats as the great exemplar on both sides.

I don't trash the site, because they do have decent articles. My concern with that particular article is the stats felt a bit too subjective in determining the final figures (the more subjective, the less reliable/accurate the results generally). Both grading scales and the notion of all things being equal (schedules are not equal, nor are the quality of play of offensive and defensive line as used to determine the plots).

I always like the attempts at analysis, and did like the article's explanation of approach. That feels a bit more scientific and allows criticism of the methodology. I enjoyed reading the article, although not necessarily embracing its methodology. The conclusion actually seemed intuitively correct without the graphs, given the NFL crack down on defenses with current/recent rules favoring offenses.

Still cannot stand Collinsworth, so unfortunately you tainted PFF a bit for me as I did not know he had a major role there. After watching him claim before the Chargers game in 2007 he no longer believed in Belichick after Spygate broke, them babbling incoherently (they actually cut away because he literally was not speaking in coherent phrases) because he could not explain his moronic conclusion after the Pats stomped the Chargers, I would love to see him climb into a closet at PFF and never emerge again on TV.
 
I kinda liked the way Football Outsiders went about providing data that I found to be useful & mostly relevant. PFF, on the other hand, claimed to offer context where none existed and then offered no context when it was an absolute necessity. Are they better than they were a decade ago? Probably, but they should be anyway. Are they the Bible of football research? Doubtful.
 
Status
Not open for further replies.


MORSE: Patriots Draft Needs and Draft Related Info
Friday Patriots Notebook 4/19: News and Notes
TRANSCRIPT: Eliot Wolf’s Pre-Draft Press Conference 4/18/24
Thursday Patriots Notebook 4/18: News and Notes
Wednesday Patriots Notebook 4/17: News and Notes
Tuesday Patriots Notebook 4/16: News and Notes
Monday Patriots Notebook 4/15: News and Notes
Patriots News 4-14, Mock Draft 3.0, Gilmore, Law Rally For Bill 
Potential Patriot: Boston Globe’s Price Talks to Georgia WR McConkey
Friday Patriots Notebook 4/12: News and Notes
Back
Top