PFF - where do you fall?

[nextsuperstar1]

19 hours ago, jetjuice said:

I interned for them for a brief period of time... it's extremely complex what they do, definitely holds some merit because as far as film watching goes, they are paying attention to every detail of every player on every play. However I don't fully agree with their they use verbiage for positional assignment, it's overly complicated. Also it's pretty hard to evaluate without all-22 film which I was not given. 

My biggest question about their process would be how they lack insight about teams' playbooks. It's not always easy to determine a player's responsibility on a given play. Still, it's a great model, not a flawless one but it's the best out currently. I think they are just scratching the surface too, I feel that they need to partner with the NFL/NCAA to improve the quality of the model with more insight to the teams playbook and scheme and it will get better as the years go on.

Went with option 2 and that seems to be the consensus.

If you can say, how would PFF handle a linebacker who abandons his assignment, free styles, and gets to the QB resulting in a pressure or a sack? That player on that play had an objective positive impact, but also abandoned his assignment which is a negative overall for the team in a macro “do your job” level. 

[BroncoSojia]

PFF had Kareem Jackson as the 5th rated cornerback overall last year, but if you ask Texan fans about him, they'll tell that he was not good as a boundary CB for them at all. The few games he played at safety must have boosted his grade up a bit

Edited May 23, 2019 by BroncoSojia

[Dome]

Went with #2, but I'm between 2-3

[Dome]

5 minutes ago, BroncoSojia said:

PFF had Kareem Jackson as the 5th rated cornerback overall last year, but if you ask Texan fans about him, they'll tell that he was not good as a boundary CB for them at all. The few games he played at safety must have boosted his grade up a bit

in 2017 they had Kenny Vaccaro rated as like the 3rd best CB in the league for the first half of the season, because of his good work in the slot (Lattimore was ranked #1 IIRC) Then he completely fell off back to how he usually plays. He wasn't ever playing like the 3rd best slot guy in the league I don't think, but it was definitely the best 6-8 game stretch of his career.

 Their grades are usually good but a random hot streak can really throw things off. I think sometimes they'll list the player who's playing better (better current grade) as the "better player" on their ranking and that's what people don't like to see. I get that their grades are their grades and maybe player X really did play better than player Y, but the longevity and consistency isn't something PFF really points out unless you go looking for it specifically. 

Edited May 23, 2019 by Dome

[Dome]

This also reminds me that PFF had Sheldon Rankins ranked the 22nd best Defensive Interior. That seems very low considering his 8 sacks and strong ability to play the run as well. He's a really balanced DT who creates a ton of inside pressure. 

It seems even worse when you consider David Onyemata, the Saints #2 DT, was ranked at #24. He's a fine starter but he's not just a notch below Rankins, not by any measure except PFF I don't think. Just look at his marijuana arrest thread, people didn't even know who he was. There's a reason for that lol. 

[ET80]

45 minutes ago, BroncoSojia said:

PFF had Kareem Jackson as the 5th rated cornerback overall last year, but if you ask Texan fans about him, they'll tell that he was not good as a boundary CB for them at all. The few games he played at safety must have boosted his grade up a bit

Yeah, I was one of those who questioned this - his effectiveness as well as his position. Boundary CB was maybe a distant 3rd on his playcard, he was more of a slot CB or joker type S from what I observed. (Jonathan Joseph, Aaron Colvin and Sharice Wright playing more on the boundaries; Wright was basically the starter opposite Joseph from week four on, IIRC).

Jackson played well at times - really patrolled the short stuff well, made some big hits and prevented some plays - but he wasn't exactly doing this as a boundary CB.

[tom cody]

I went with Option #2. 

[jetjuice]

7 hours ago, Hunter2_1 said:

I listen to their podcast sometimes, and they certainly use the All-22. I wouldn't trust any analysis that didn't use that view. What did they get you doing?

Oh the employees def have it, I just didn't get to see it lol.

They didn't have me doing very much. I honestly dropped out before I could make it to the final round because it was too much of a time commitment with my job and training, but I was watching Big 10 college games and analyzing positional assignments. I didn't get to grade anyone, just basically identified who is on the field, where they line up, and what their assignment was. Basically everything that comes before the grading.

[jetjuice]

1 hour ago, nextsuperstar1 said:

If you can say, how would PFF handle a linebacker who abandons his assignment, free styles, and gets to the QB resulting in a pressure or a sack? That player on that play had an objective positive impact, but also abandoned his assignment which is a negative overall for the team in a macro “do your job” level. 

Their whole premise of grading is based on how often they make efficient impact plays. Frequency plays a huge role in their grading, it's all based on an effective outcome on a per play basis, so I'd guess that would give the player a positive mark regardless of assignment. Depending on who you ask that would make the system either flawed or not.

[jetjuice]

1 hour ago, Dome said:

This also reminds me that PFF had Sheldon Rankins ranked the 22nd best Defensive Interior. That seems very low considering his 8 sacks and strong ability to play the run as well. He's a really balanced DT who creates a ton of inside pressure. 

It seems even worse when you consider David Onyemata, the Saints #2 DT, was ranked at #24. He's a fine starter but he's not just a notch below Rankins, not by any measure except PFF I don't think. Just look at his marijuana arrest thread, people didn't even know who he was. There's a reason for that lol. 

Same way they had Mike Pennel graded at 87.something while he wasn't our best DT.

[SDotNova]

23 hours ago, Danger said:

It's more appropriate for positions like OLine and Corner where it's less simple to statistically grade a players performance.

What about WR? Is that reception due to the QB placing the ball on the money or is that reception do to a WR creating massive separation? Same with RB. Did the RB break 10 tackles to gain 5 yards or did he largely go untouched to get 6 yards? 

I think PFF adds a lot to the conversation. Their grading aligns more to what you see on film than traditional stats. 

Also, for QBs they have their accuracy plus statistics. So, they are tracking whether the QB put the ball on the money orif the WR had to stretch for the ball. Both may result in a completion. But one shows more true accuracy (i.e, the one where the ball hit the WR on the numbers).

Edited May 23, 2019 by SDotNova

[SDotNova]

22 hours ago, jetjuice said:

I interned for them for a brief period of time... it's extremely complex what they do, definitely holds some merit because as far as film watching goes, they are paying attention to every detail of every player on every play. However I don't fully agree with their they use verbiage for positional assignment, it's overly complicated. Also it's pretty hard to evaluate without all-22 film which I was not given. 

My biggest question about their process would be how they lack insight about teams' playbooks. It's not always easy to determine a player's responsibility on a given play. Still, it's a great model, not a flawless one but it's the best out currently. I think they are just scratching the surface too, I feel that they need to partner with the NFL/NCAA to improve the quality of the model with more insight to the teams playbook and scheme and it will get better as the years go on.

Went with option 2 and that seems to be the consensus.

Nice post. From a stat tracking standpoint, it's almost as good as you are going to get. As you noted, it's not perfect because only the coaches will ever know what each player was suppose to do. But for what they are trying to do, it's quality work. 

I'm between 1 and 2. I went with 1. 

[Hunter2_1]

2 hours ago, jetjuice said:

Same way they had Mike Pennel graded at 87.something while he wasn't our best DT.

Was he not? Crap. I've been using that stat to justify us signing him 

[BleedTheClock]

4 hours ago, nextsuperstar1 said:

If you can say, how would PFF handle a linebacker who abandons his assignment, free styles, and gets to the QB resulting in a pressure or a sack? That player on that play had an objective positive impact, but also abandoned his assignment which is a negative overall for the team in a macro “do your job” level. 

Considering they had Jamie Collins rated extremely low in past years, I'm guessing that they do consider their gap assignments. Collins made a lot of splash plays, but was fundamentally horrific.

[Hunter2_1]

4 hours ago, nextsuperstar1 said:

If you can say, how would PFF handle a linebacker who abandons his assignment, free styles, and gets to the QB resulting in a pressure or a sack? That player on that play had an objective positive impact, but also abandoned his assignment which is a negative overall for the team in a macro “do your job” level. 

 

They rinsed him

 

 

"Jamie Collins certainly had a larger impact during his time w/ the Patriots; how'd he do it? Collins spent just 11.4% of his snaps in New England along the edge while he spent 31.8% in Cleveland at edge defender. "

Edited May 23, 2019 by Hunter2_1