Jump to content

Recommended Posts

Posted (edited)
1 hour ago, GunnerBill said:

 

With Tre it is that to earn their really high grades you need to be making plays on the ball. Tre could play every coverage perfectly but if the ball is never thrown his way (and it rarely was) then he isn't gonna get the top marks. 

 

That is why I say their grading system is far from perfect, but it isn't useless either.

You're right, it's worse than useless.  It's downright deceptive and the incarnation of "there's lies, damn lies, and statistics".  It blows my mind that people still defend them as being hardcore data when they are actually the opposite.  Their whole system is literally based on arbitrary and cherrypicked SUBJECTIVITY applied to ACTUAL data.

 

In their system, Allen having faith in his receivers to make contested catches, sideline snags, and dives to catch balls placed around tight coverage for a TD is actually a bad thing!  He could go down as the worst rated MVP in history because their contextual rating system either doesn't account for everything, or incorrectly categorizes some things as negative.

Edited by 1ManRaid
Posted
23 minutes ago, CNY315 said:

 

I saw them respond to demands that they justify having Darnold ranked ahead of Allen.  His response?  I crap you not, was (after some nervous stuttering clearly trying to come up with plausible BS) "cuz reasons".  No actual data, literally just vague assertions that there WERE reasons.

Posted
4 minutes ago, Lieutenant Aldo Raine said:

 

Exactly; amazes me how these sites have any credibility when they insert their own bias.  

They were somehow able to rebrand their subjective bias as objective data and gain credibility as the "hardcore data" guys.  They pander "data based" narratives to big markets and trending popular teams.

 

Well until they lost what credibility they had left when they exposed themselves with this Darnold>Allen fiasco.

Posted
4 hours ago, HappyDays said:

 

 

LOL I put PFF  on my "do not respond, Ever" list when they admitted they trolled Bills Mafia by saying Devlin "Duck" Hodges was a better QB (Momson said, "well yeah we said it but I don't think anyone here really believed it".  Oh, so why should we think you believe anything you say and aren't just going for clicks?).

 

PFF has all kinds of phony baloney in their system like "interceptable balls" (balls that could have been intercepted, but weren't).  They give them a lot of weight.  That's how a QB like Tyrod Taylor was their 13th highest- ranked QB in 2015 and their 11th highest ranked QB in 2016.  He never took chances and threw "interceptable balls" much less interceptions.  Of course, that meant he left all kinds of passing yardage on the field, but hey.

That's a silly list, right there.

  • Like (+1) 1
Posted
32 minutes ago, Lieutenant Aldo Raine said:

 

Exactly; amazes me how these sites have any credibility when they insert their own bias.  

 

Just go back to their rankings of 2015 and 2016 and look at the QB they had ranked above and below TT and ask if that ranking makes sense.

 

It's an interesting question because at heart, I am very, very much a "data geek".  I believe in data. I believe in statistics.  But, not all data and statistics are equal.  Fundamentally, for a team game like football, the key question is really "how closely does that statistic correlate with what counts - winning the game"?  I like to look at statistics that are closest to objectively measurable data and that correlate with winning.  Completion percentage matters.  TD/INT matter.  YPA matter.

 

PFF and some of these other analytics sites get into a very very grey area where they assert that they have objective statistics - the best, most objective statistics - without apparently being aware of how they can be striving so hard to be objective, that they become subjective.  When it comes to grading ball placement and creating a statistic like "interceptable balls", there are a lot of nuances.

 

Anyway, here's an article from PFF describing how they grade QB.  Sounds real good, doesn't it?

https://www.pff.com/news/pro-pff-qb-grading-most-effective-tool-there-is

  • Like (+1) 1
Posted (edited)

I thought the Tre crap last year was bad, but this is just idiocracy on every level imaginable. 

 

100% confirmed Clowns and there is really nothing else to say. 

 

giphy.gif

 

Edited by Real McNasty
Posted
2 hours ago, WideNine said:

 

The Josh Allen is PFF's  Schrödinger's cat analogy?

 

He may be good QB or he may suck, but you have to actually open the box and watch the game to know.... something like that?

 

 

 


Schrodingers Cat?  Makes me think of the Big Bang Theory.  That was funny and yes I knew the reference out of that example.  Why worry not you but in general to my TSW friends what these ding dongs say?  They have an agenda and just can’t let it go.  As I used to say to my kids I coached or when I played, “just kick their ass (sorry mods) and that will speak for itself”.  When I played, I loved bulletin board material.   As a coach, it was great.  My concern as a coach is they would get on out of their minds trying to take a guys head off and make a mistake.  As much as I hate Belicheck, just do you’re job.

 

If we get the Boys are back in town (Edmunds and Milano), and beat handedly the Rams, you know we are going to beat Chucky.  The Titans look good, but they can be beat.  Just let’s move onto the Rams.

  • Thank you (+1) 1
Posted
1 hour ago, Joe in Winslow said:

And here we clearly see the problem with subjective vs. objective reality.

 

1 hour ago, Doc said:

Huh?  Who got those stats then?  

 

Oh my gosh. Allen LITERALLY DID have that kind of game. What are these "analysts" talking about?

 

That's the most dominant performance from a Bills QB we have seen in decades.

  • Like (+1) 1
Posted
12 minutes ago, Hapless Bills Fan said:

 

Just go back to their rankings of 2015 and 2016 and look at the QB they had ranked above and below TT and ask if that ranking makes sense.

 

It's an interesting question because at heart, I am very, very much a "data geek".  I believe in data. I believe in statistics.  But, not all data and statistics are equal.  Fundamentally, for a team game like football, the key question is really "how closely does that statistic correlate with what counts - winning the game"?  I like to look at statistics that are closest to objectively measurable data and that correlate with winning.  Completion percentage matters.  TD/INT matter.  YPA matter.

 

PFF and some of these other analytics sites get into a very very grey area where they assert that they have objective statistics - the best, most objective statistics - without apparently being aware of how they can be striving so hard to be objective, that they become subjective.  When it comes to grading ball placement and creating a statistic like "interceptable balls", there are a lot of nuances.

 

Anyway, here's an article from PFF describing how they grade QB.  Sounds real good, doesn't it?

https://www.pff.com/news/pro-pff-qb-grading-most-effective-tool-there-is

 

Well therin lies the problem. PFF grades have NOTHING to do with data. They make the data up. They look at a play and assign a number to it. It's not actually data.

 

I am a data guy too, and my job is in data analysis. What PFF does is something completely different. They have some interesting insight sometimes, but they really screw up the QB position and have some deep flaws. Russel Wilson didn't receive high grades from his performance? Their version of reality does not mesh with what we see on Sundays.

  • Like (+1) 1
  • Thank you (+1) 2
Posted

There is nothing PFF offers that cannot be had elsewhere, why do you fellas take the “click” bait? 
 

Just stop going on their site, they are a half arsed outfit on their best day, 

 

Go Bills!!!

 

Posted
2 minutes ago, Don Otreply said:

There is nothing PFF offers that cannot be had elsewhere, why do you fellas take the “click” bait? 
 

Just stop going on their site, they are a half arsed outfit on their best day, 

 

Go Bills!!!

 

 

There is actually. But that is the stuff you really have to pay for. The player grades are the cheap stuff that to be honest is a good marketing tool. They are not useless but they are far from perfect. However, they are the reason you guys are all talking about PFF again.

Posted
1 minute ago, GunnerBill said:

 

There is actually. But that is the stuff you really have to pay for. The player grades are the cheap stuff that to be honest is a good marketing tool. They are not useless but they are far from perfect. However, they are the reason you guys are all talking about PFF again.

I have not clicked on their site in years they imo offer nothing that a multitude of of venues already make available. 
 

I just try to get folk to realize they are being had, here on this site, but people are gluttons for psychological punishment... I guess that’s why we have been fans of this team for so long...

 

Go Bills!!!

Posted
17 minutes ago, MJS said:

 

Well therin lies the problem. PFF grades have NOTHING to do with data. They make the data up. They look at a play and assign a number to it. It's not actually data.

 

I am a data guy too, and my job is in data analysis. What PFF does is something completely different. They have some interesting insight sometimes, but they really screw up the QB position and have some deep flaws. Russel Wilson didn't receive high grades from his performance? Their version of reality does not mesh with what we see on Sundays.

Obligatory "I am not a data analyst but did take a semester of statistics in university", and the literal first thing they taught us is how data can be skewed to push narratives and deceive people.  

 

Oh and Wilson did get high grades from them, just not as high as Mediocre Brady and One Half Jimmy.

Posted (edited)

Analytics has its place. The problem like with most ideologies is trying to use it in every single facet of sports and being a fundamentalist about it.... 

 

The movie money ball was a great movie but the irony was within analytics itself.  The lesson here is to never dismiss new world views, new ideas and to have a broad view with success. Nothing is always one way...much like Bobby Fisher who would break the cardinal rule in chess giving up his queen, there are multiple ways to win a game. Dogma hurt bill polian in Lamar Jackson's evaluation etc

 

So when you hear a college qb with under 60% comp % has never done x y or z 

 

You should pause before casting complete JUDGMENT and use it as a tool , not the end all be all

Edited by AirAllenPower
×
×
  • Create New...