Pro Football Focus

[Phragle]

Am I the only one that both loves PFF for what it makes available, but also hates it for all the stuff it doesn't make available? For example, I love that I can see that Brady's rating is 20 points higher on play action than non play action passes, but why can't I see how many times he's been sacked on PA when they record those numbers? And why do I have to put their data into an excel file anytime I actually want to figure anything out? Where are the splits and streak finders? 
 
Sorry - rant over.

 

[Super Nomario]

Am I the only one that both loves PFF for what it makes available, but also hates it for all the stuff it doesn't make available? For example, I love that I can see that Brady's rating is 20 points higher on play action than non play action passes, but why can't I see how many times he's been sacked on PA when they record those numbers? And why do I have to put their data into an excel file anytime I actually want to figure anything out? Where are the splits and streak finders? 
 
Sorry - rant over.

No, I have the same criticisms. I wish they split out their grades - I'd love to know Chandler Jones' pass rushing numbers split out by him lining up at DT vs DE. I'd love to see DB grades in man vs zone coverage.

 

[Jnai]

If they have it, why not just email and ask for the raw data? Are they that large a company?

 

[Phragle]

No, I have the same criticisms. I wish they split out their grades - I'd love to know Chandler Jones' pass rushing numbers split out by him lining up at DT vs DE. I'd love to see DB grades in man vs zone coverage.

 
Funny, I asked about that today on twitter. Nothing yet.
 

If they have it, why not just email and ask for the raw data? Are they that large a company?

 
Why would they give it to me but not make it available to all of their paying customers? And they do have it. There are plenty of things the staff talks about on twitter that aren't available on the site.
 
I don't know anything about their business.

 

[Jnai]

 
Funny, I asked about that today on twitter. Nothing yet.
 
 
Why would they give it to me but not make it available to all of their paying customers? And they do have it. There are plenty of things the staff talks about on twitter that aren't available on the site.
 
I don't know anything about their business.

 
Because making data available to people is more complicated than simply looking at the computer and saying "Go, Data!" It's hard to do right. Sometimes sports statistics websites don't want to create complex interfaces for downloading data, but would be happy to answer your specific questions about it or send you more granular statistics with the caveat that they're not interested in helping you analyze the stuff.
 
I do it all the time, especially for academic use.

 

[Phragle]

 
Because making data available to people is more complicated than simply looking at the computer and saying "Go, Data!" It's hard to do right. Sometimes sports statistics websites don't want to create complex interfaces for downloading data, but would be happy to answer your specific questions about it or send you more granular statistics with the caveat that they're not interested in helping you analyze the stuff.
 
I do it all the time, especially for academic use.

 
How hard can it be to add a few more columns to a spread sheet? That can't be hard.
 
I just don't get why they do all the work they do - and it's a lot - and then leave a bunch of stuff out. It's maddening. Do you have PFF, Jnai? It's already a pretty complex interface. Anyone that is looking at the data that I am isn't worried about getting overwhelmed by more data.
 
Maybe I'm spoiled by baseball sites.

 

[SeoulSoxFan]

You have to figure the more detailed information & data sets are being sold to NFL teams (or casinos?) at a much higher premium.
 
If I were part of PFF I'd have "enterprise" level data with enterprise-level $$$ and "consumer" level data at $30 a pop for the rest of us.

 

[Phragle]

No, I have the same criticisms. I wish they split out their grades - I'd love to know Chandler Jones' pass rushing numbers split out by him lining up at DT vs DE. I'd love to see DB grades in man vs zone coverage.

 
Just talked to Steve about this.
 
 
Phragle ‏@Phragle29m
@StevePalazzolo Any luck with those Chan Jones splits?
 
Steve Palazzolo ‏@StevePalazzolo19m
@Phragle He's rushed from DT on 4-man line 72 times with 12 total pressures: 1 sk, 5 hits, 5 hurries. 87 rushes from DE on 3-man line, 1/1/3
 
Steve Palazzolo ‏@StevePalazzolo19m
@Phragle at DE on 4-man line he's rushed 309 times with 7 sacks, 3 hits, 15 hurries. At OLB, 104 rushes with 1 sack, 4 hits, 4 hurries
 
 
Phragle ‏@Phragle3m
@StevePalazzolo Re. Chandler Jones splits: That's only 10 total sacks. PFF lists him with 13 total. What's going on there? Half sacks?
 
 
Steve Palazzolo ‏@StevePalazzolo2m
@Phragle we have him for 3 sacks at "other" positions
 
 
Phragle ‏@Phragle1m
@StevePalazzolo Weird. That's kind of a lot.
 
 
Steve Palazzolo ‏@StevePalazzolo1m
@Phragle assuming they're from more of an OLB position, but it's more a limitation with the way we spit out the info, so I can't easily look
 
 
Phragle ‏@Phragle3m
@StevePalazzolo Might be from plays where he stunts?
 
 
Steve Palazzolo ‏@StevePalazzolo2m
@Phragle no, probably just RLB (3-man LB corps). Our sheet only shows ROLB (4-man LB corps)

 

[Super Nomario]

Just talked to Steve about this.
 
Phragle ‏@Phragle29m
@StevePalazzolo Any luck with those Chan Jones splits?
 
Steve Palazzolo ‏@StevePalazzolo19m
@Phragle He's rushed from DT on 4-man line 72 times with 12 total pressures: 1 sk, 5 hits, 5 hurries. 87 rushes from DE on 3-man line, 1/1/3
 
Steve Palazzolo ‏@StevePalazzolo19m
@Phragle at DE on 4-man line he's rushed 309 times with 7 sacks, 3 hits, 15 hurries. At OLB, 104 rushes with 1 sack, 4 hits, 4 hurries

 
Cool, here it is in table form.
[table Chandler Jones PRP] Rush Sack Hits Hurries TP PRP Rnk 4-3 DT 72 1 5 5 11 11.8 1st/34 3-4 DE 87 1 1 3 5 4.6 20th/23 4-3 DE 309 7 3 15 25 6.6 37th/39 3-4 OLB 104 1 4 4 9 6.7 26th/27 Other LB 65 3 1 12 16 19.6 SUM 637 13 14 39 66 8.3 [/table]
 
The number of hurries from the "other" LB spot looks wrong, but they credit him with 39 hurries in total so I'm just taking the difference. Also, the overall PRP grade (8.3) doesn't match PFF's reported 8.7. His DT performance would be the best of any regular DT in the league, but all his other performance numbers (except the "other") are bottom-of-the-barrel.

 

[Phragle]

 
Cool, here it is in table form.
[table Chandler Jones PRP] Rush Sack Hits Hurries TP PRP Rnk 4-3 DT 72 1 5 5 11 11.8 1st/34 3-4 DE 87 1 1 3 5 4.6 20th/23 4-3 DE 309 7 3 15 25 6.6 37th/39 3-4 OLB 104 1 4 4 9 6.7 26th/27 Other LB 67 3 1 12 16 19.0 SUM 639 13 14 39 66 8.3 [/table]
 
The number of hurries from the "other" LB spot looks wrong, but they credit him with 39 hurries in total so I'm just taking the difference. Also, the overall PRP grade (8.3) doesn't match PFF's reported 8.7. His DT performance would be the best of any regular DT in the league, but all his other performance numbers (except the "other") are bottom-of-the-barrel.

 
Cool. Yeah there is definitely some noise there, and I don't know what to think about the Other position. I don't remember seeing him play 4-3 OLB at all, never mind for at least 67 snaps.
 
Also from his Defensive Snap Summary page it says he had 637 PR snaps rather than 639.
 
I remember from last year while looking at splits that you can't calculate PRP that easily. I had to do a lot of work get CJ's PRP before and after the ankle injury. Unfortunately I don't remember what I had to do or why.
 
This doesn't make it right, but ry multiplying his PRP positional splits by 1.048193 and see how that affects his rankings. 1.048193 will at least make your total PRP = PFF's.
 
Despite the numbers being off, the one thing I'd take away from it is that he's very effective from inside, and I like seeing that. That's a value-adding quality to have in a DE. The reason might not be that he's better inside, it might be because he's allowed to be more aggressive when he's inside. When he's at DT by nature the defense will be playing the pass more aggressively. When he's at DE sometimes he has to play containment.

 

[Eric Ampersand]

Somewhat off topic: does anyone have data for non-QB rushing yards allowed? QB rushing yards often come from broken plays. These plays are important during a given game but may not be predictive. I'd like to know if there is a ranking of rushing yards allowed for designed runs only (to be more specific). Phragle's comment about DEs playing contain gave me flashbacks to the Carolina game. 

 

[Super Nomario]

Cool. Yeah there is definitely some noise there, and I don't know what to think about the Other position. I don't remember seeing him play 4-3 OLB at all, never mind for at least 67 snaps.

It's probably not 4-3; it's probably more like 3-3 (or 2-3 or 3-2). Basically, any time he's in a two-point stance, he's classified as a LB, and if there are only two other LBs he's getting this RLB (or LLB) designation. So, most of the time it's probably a 3-3 where he's standing up and Ninkovich has his hand in the dirt.
 

Also from his Defensive Snap Summary page it says he had 637 PR snaps rather than 639.

That's just me sucking at math. I'll fix it.
 

 I remember from last year while looking at splits that you can't calculate PRP that easily. I had to do a lot of work get CJ's PRP before and after the ankle injury. Unfortunately I don't remember what I had to do or why.

I went by the formula I found here: https://www.profootballfocus.com/blog/2013/10/02/signature-stats-pass-rushing-productivity/ - sacks rated 1, hurries and hits .75.
 

Despite the numbers being off, the one thing I'd take away from it is that he's very effective from inside, and I like seeing that. That's a value-adding quality to have in a DE. The reason might not be that he's better inside, it might be because he's allowed to be more aggressive when he's inside. When he's at DT by nature the defense will be playing the pass more aggressively. When he's at DE sometimes he has to play containment.

I think there's something to this. On the other hand, Ninkovich (presumably with the same coaching / responsibilities) has better PRP numbers, Jones isn't always good at containment (see the Carolina game), and the last two weeks, the Pats have been putting Carter in at DT and leaving Jones at DE rather than kicking Jones inside.
 
I fall somewhere in the middle here; I have a hard time believing Jones is only the 41st-best pass rushing DE (as PFF regards him) but I do think he's overrated here - the sacks are great plays, but he's not the consistent disruptor the best pass rushers are, IMO.

 

[Phragle]

It's probably not 4-3; it's probably more like 3-3 (or 2-3 or 3-2). Basically, any time he's in a two-point stance, he's classified as a LB, and if there are only two other LBs he's getting this RLB (or LLB) designation. So, most of the time it's probably a 3-3 where he's standing up and Ninkovich has his hand in the dirt.

 
Oh alright. Well to me - when that happens - he's still a DE in a 4-man front, or at least a pass-rusher in some sort of nickel/dime 3-4 defense. I think it's safe to say the numbers and ranking we crunched underrate CJ quite a bit. Both because of the wrong inputs, and the 20 PRP Other position.
 

I went by the formula I found here: https://www.profootballfocus.com/blog/2013/10/02/signature-stats-pass-rushing-productivity/ - sacks rated 1, hurries and hits .75.

 
Yeah that's the correct formula, but the inputs aren't correct. If they were you'd get 8.7.
 
Fake edit, I may have just figured it out. If you go the the PRP page and punch the numbers in yourself you get 8.3, but that's with 639 passing snaps. If you use pass rushing snaps - 609 - you get 8.7.
 

I think there's something to this. On the other hand, Ninkovich (presumably with the same coaching / responsibilities) has better PRP numbers, Jones isn't always good at containment (see the Carolina game), and the last two weeks, the Pats have been putting Carter in at DT and leaving Jones at DE rather than kicking Jones inside.

 
I wouldn't presume that. Each DE has different DTs next to them and different LBs behind them.
 

I fall somewhere in the middle here; I have a hard time believing Jones is only the 41st-best pass rushing DE (as PFF regards him) but I do think he's overrated here - the sacks are great plays, but he's not the consistent disruptor the best pass rushers are, IMO.

 
I don't think he's allowed to be. We know BB loves to stop the run and keep the mobile QBs in the pocket. Also there are DEs that set-up tackles so that when they really need a sack they can get it, and there are DEs that try to get a sack every play.
 
Watch Robert Quinn or Cameron Wake. It's clear their objective is different.

 

[Super Nomario]

Fake edit, I may have just figured it out. If you go the the PRP page and punch the numbers in yourself you get 8.3, but that's with 639 passing snaps. If you use pass rushing snaps - 609 - you get 8.7.

You're right. In that case, the "other" category has just 37 rush attempts, but 3 sacks, 1 hit, and 12 hurries, which doesn't really pass the smell test. One of the sets of numbers you got from Palazzolo must be wrong.
 

I don't think he's allowed to be. We know BB loves to stop the run and keep the mobile QBs in the pocket. Also there are DEs that set-up tackles so that when they really need a sack they can get it, and there are DEs that try to get a sack every play.
 
Watch Robert Quinn or Cameron Wake. It's clear their objective is different.

I see the point you're trying to make, but I don't think it holds up to scrutiny. Only 1 of Jones' sacks was in the 4th quarter (a takedown of Brees with 13:02 left), and only 1 of his sacks was on 3rd down, so if he's saving his bullets, he's not firing them in key situations. It's actually Ninkovich who's had the penchant for clutch sacks - he only has 8 sacks (vs Chandler's 10.5), but 4 were 4th-quarter sacks and 5 were on 3rd down.

 

[Phragle]

You're right. In that case, the "other" category has just 37 rush attempts, but 3 sacks, 1 hit, and 12 hurries, which doesn't really pass the smell test. One of the sets of numbers you got from Palazzolo must be wrong.

 
I'm a little out of it today. You're going to have to explain this for me.
 

I see the point you're trying to make, but I don't think it holds up to scrutiny. Only 1 of Jones' sacks was in the 4th quarter (a takedown of Brees with 13:02 left), and only 1 of his sacks was on 3rd down, so if he's saving his bullets, he's not firing them in key situations. It's actually Ninkovich who's had the penchant for clutch sacks - he only has 8 sacks (vs Chandler's 10.5), but 4 were 4th-quarter sacks and 5 were on 3rd down.

 
That eliminates one theory, but not the other.  You don't think Jones plays with more responsibility than some other DEs?

 

[Super Nomario]

I'm a little out of it today. You're going to have to explain this for me.

I mean the "other" totals (his totals in the categories Palazzolo didn't give you) are absurdly high for the number of "other" pass-rushing snaps. The numbers Palazzolo gave you were short 3 sacks, 1 hit, and 12 hurries, but there are only 37 pass-rushing snaps unaccounted for.
 

That eliminates one theory, but not the other.  You don't think Jones plays with more responsibility than some other DEs?

Maybe? He plays a ton, and they use him in different roles / fronts. I don't know how much the Patriots ask him to contain versus how much other teams do their edge rushers, but if they do, I don't think he's doing a particularly good job of it. The Patriots have allowed 11 runs of 5+ yards off left end (Jones' usual side) to WRs or QBs; they've only allowed 4 such runs off right end. St. Louis has only allowed 3 such runs off left end (Quinn's side); Miami has allowed 6 off right end (Wake's). So if they're asking Jones to do more in containment, it's not really resulting in limiting edge rushes.

 

[Phragle]

I mean the "other" totals (his totals in the categories Palazzolo didn't give you) are absurdly high for the number of "other" pass-rushing snaps. The numbers Palazzolo gave you were short 3 sacks, 1 hit, and 12 hurries, but there are only 37 pass-rushing snaps unaccounted for.

Because it's like a 35 PRP? Sure that's abnormally high, but his PRP at 3-4 OLB and 4-3 DE look abnormally low to me. I think if you totaled up Other, 4-3 DE, and 3-4 OLB you'd get a pretty accurate PRP for Jones as an edge rusher.
 

Maybe? He plays a ton, and they use him in different roles / fronts. I don't know how much the Patriots ask him to contain versus how much other teams do their edge rushers, but if they do, I don't think he's doing a particularly good job of it. The Patriots have allowed 11 runs of 5+ yards off left end (Jones' usual side) to WRs or QBs; they've only allowed 4 such runs off right end. St. Louis has only allowed 3 such runs off left end (Quinn's side); Miami has allowed 6 off right end (Wake's). So if they're asking Jones to do more in containment, it's not really resulting in limiting edge rushes.

Maybe not, but when I watch Jones he's often trying to go through the LT rather than around him, thus controlling both gaps. Other DEs like Quinn don't do that.

 

[Super Nomario]

Because it's like a 35 PRP? Sure that's abnormally high, but his PRP at 3-4 OLB and 4-3 DE look abnormally low to me. I think if you totaled up Other, 4-3 DE, and 3-4 OLB you'd get a pretty accurate PRP for Jones as an edge rusher.

I think we're on the same page. Some of the hurries and sacks that are "other" in my chart (based on the numbers you got from Palazzalo) are probably mis-categorized and should be under 3-4 OLB or 4-3 DE. It's more likely to me that he fat-fingered one of the hurry totals he gave you than that Jones really got 16 TP in 37 "other" snaps after getting just 9 in 104 3-4 OLB snaps.
 

Maybe not, but when I watch Jones he's often trying to go through the LT rather than around him, thus controlling both gaps. Other DEs like Quinn don't do that.

I think containment is probably a part of this, but I think the other piece is that Jones doesn't really have Quinn's elite bend-the-edge speed. It would be interesting to look at the games against mobile guys (Newton, Roethlisberger, Manuel, Smith) and the ones against immobile guys (Manning, Flacco, Ryan) and see if he's using different technique.

 

[Phragle]

I think we're on the same page. Some of the hurries and sacks that are "other" in my chart (based on the numbers you got from Palazzalo) are probably mis-categorized and should be under 3-4 OLB or 4-3 DE.

 
Right. Only you say "3-4 OLB or 4-3 DE" but I think '3-4 OLB and 4-3 DE' is also possible.
 

It's more likely to me that he fat-fingered one of the hurry totals he gave you than that Jones really got 16 TP in 37 "other" snaps after getting just 9 in 104 3-4 OLB snaps.

But here you lose me. The numbers he gave me total up to 66 total pressures, and so does the PRP page on PFF. So that checks out. You think he fat fingered it two or three times and it still added up correctly? That sounds impossible to me.
 

I think containment is probably a part of this, but I think the other piece is that Jones doesn't really have Quinn's elite bend-the-edge speed. It would be interesting to look at the games against mobile guys (Newton, Roethlisberger, Manuel, Smith) and the ones against immobile guys (Manning, Flacco, Ryan) and see if he's using different technique.

 
 
Jones isn't as good as Quinn so he doesn't get the same freedom. I guess so, but I still think Jones is being held back (by design) a little more than the players he's compared to.

 

[Super Nomario]

 
But here you lose me. The numbers he gave me total up to 66 total pressures, and so does the PRP page on PFF. So that checks out. You think he fat fingered it two or three times and it still added up correctly? That sounds impossible to me.
 

It adds up correctly because I made it add up correctly - the "Other LB" line is calculated by subtracting from PFF's known totals.
 
This is what Palazzalo gave you:
@Phragle He's rushed from DT on 4-man line 72 times with 12 total pressures: 1 sk, 5 hits, 5 hurries. 87 rushes from DE on 3-man line, 1/1/3
 
Steve Palazzolo ‏@StevePalazzolo19m
@Phragle at DE on 4-man line he's rushed 309 times with 7 sacks, 3 hits, 15 hurries. At OLB, 104 rushes with 1 sack, 4 hits, 4 hurries
 
That only adds up to 10 sacks, 13 hits, 27 hurries, but PFF credits him with 13 sacks, 14 hits, 39 hurries. So there's missing production, which I just plugged in the "other" column but I'm guessing some of the numbers he gave you are screwed up.

 

[Phragle]

It adds up correctly because I made it add up correctly - the "Other LB" line is calculated by subtracting from PFF's known totals.

 
Yeah of course. I don't now how I messed that up.
 

This is what Palazzalo gave you:
 
@Phragle He's rushed from DT on 4-man line 72 times with 12 total pressures: 1 sk, 5 hits, 5 hurries. 87 rushes from DE on 3-man line, 1/1/3
 
Steve Palazzolo ‏@StevePalazzolo19m
@Phragle at DE on 4-man line he's rushed 309 times with 7 sacks, 3 hits, 15 hurries. At OLB, 104 rushes with 1 sack, 4 hits, 4 hurries
 
 
That only adds up to 10 sacks, 13 hits, 27 hurries, but PFF credits him with 13 sacks, 14 hits, 39 hurries. So there's missing production, which I just plugged in the "other" column but I'm guessing some of the numbers he gave you are screwed up.

 
I don't see how any of that means he messed up relaying the numbers. It's possible but I don't think suggests it. I think it just means Jones was exceptional a bunch of plays and PFF happened to screw up his actual position on those plays.
 
Anyway, when I make all the adjustments I think need to be made here are the numbers I got.
 
4-3 DT: 12.4 PRP (1st/34)
3-4 DE: 4.8 PRP (20th/23)
Edge*: 8.8 PRP (20th/37)
All Positions: 8.7 PRP
 
* 3-4 OLB, 4-3 DE, and Other combined.
 
There's certainly some margin of error there but that's the best I can do with the info we have.

 

[Super Nomario]

I don't see how any of that means he messed up relaying the numbers. It's possible but I don't think suggests it. I think it just means Jones was exceptional a bunch of plays and PFF happened to screw up his actual position on those plays.

OK, but either way the numbers he gave you were probably wrong. Whether PFF misclassified his position or Steve relayed them to you incorrectly, they don't make a lot of sense.

 

[Phragle]

OK, but either way the numbers he gave you were probably wrong. Whether PFF misclassified his position or Steve relayed them to you incorrectly, they don't make a lot of sense.

 
What do you think is wrong? I think that when Jones was recorded at Other, he was actually either playing 3-4 OLB or 4-3 DE. Would that explain what you think doesn't make sense? 
 
So if I'm right about that I can combine the numbers and get some accuracy. They look about right to me.
 
I kind of want to invite him in here to clear everything up but the first post is me just shitting all over PFF.

 

[Super Nomario]

 
What do you think is wrong? I think that when Jones was recorded at Other, he was actually either playing 3-4 OLB or 4-3 DE. Would that explain what you think doesn't make sense? 

2 things:
1) If your theory is right, and they just accidentally miscategorized his position, it would be a pretty big coincidence that he had crazy production on plays where they miscategorized. I'd expect it to just be a random subsection of play.
2) I have some insight into how PFF charts position (I applied for a job with them last offseason and ran through participation charting for a couple games before withdrawing my application) and they put this stuff in spreadsheets with elaborate field codes to make sure people don't screw this stuff up. If you put Jones at RLB (other) instead of ROLB and it doesn't match the number of LBs, it's going to throw an error. They also check this stuff later; you see them retroactively fix stuff all the time. So while I could see a couple plays off here or there, I doubt it's very many - and it still doesn't explain why those plays are disproportionately productive.
 
New theory: Jones lines up on the right side ~90% of the time, but he does line up on the left side some. He had 2 sacks and 9 hurries from the left side. I bet Palazzolo forgot to count those, just running checks for RE, RDE, RDT, and ROLB, but he included them in the total number of pass rushes. That would explain most of the "missing" production, and leave the "other" production - 1 sack, 1 hit, 3 hurries - way more in line with the other stuff.

 

[Phragle]

2 things:
1) If your theory is right, and they just accidentally miscategorized his position, it would be a pretty big coincidence that he had crazy production on plays where they miscategorized. I'd expect it to just be a random subsection of play.
2) I have some insight into how PFF charts position (I applied for a job with them last offseason and ran through participation charting for a couple games before withdrawing my application) and they put this stuff in spreadsheets with elaborate field codes to make sure people don't screw this stuff up. If you put Jones at RLB (other) instead of ROLB and it doesn't match the number of LBs, it's going to throw an error. They also check this stuff later; you see them retroactively fix stuff all the time. So while I could see a couple plays off here or there, I doubt it's very many - and it still doesn't explain why those plays are disproportionately productive.

 
1. Coincidences happen. Especially so in SSS.
 
2. How do the nickel an dime defenses affect that? I think it would be hard to misidentify him in a base 4-3, but if he was standing up at DE in the 4-2 or 4-1 I think one could easily misidentify him as an OLB and call the defense a 3-3 or 3-2.
 

New theory: Jones lines up on the right side ~90% of the time, but he does line up on the left side some. He had 2 sacks and 9 hurries from the left side. I bet Palazzolo forgot to count those, just running checks for RE, RDE, RDT, and ROLB, but he included them in the total number of pass rushes. That would explain most of the "missing" production, and leave the "other" production - 1 sack, 1 hit, 3 hurries - way more in line with the other stuff.

He said "we have him for 3 sacks at 'other' positions" not 1.
 
Also if this was true I think when he's on the left it's usually at DE in a 4 man front so the numbers I came up with would still be close.

 

[Super Nomario]

1. Coincidences happen. Especially so in SSS.

You're talking about 23% of his sacks and 31% of his hurries happening in 6% of his snaps. That's hugely out of proportion. He's got three times as many sacks and hurries in "other" vs OLB in about 1/3 of the snaps; that doesn't seem right.
 

2. How do the nickel an dime defenses affect that? I think it would be hard to misidentify him in a base 4-3, but if he was standing up at DE in the 4-2 or 4-1 I think one could easily misidentify him as an OLB and call the defense a 3-3 or 3-2.

Standing up vs 3-point stance is how PFF distinguishes between DE and OLB. If he was standing up in a 3-3 or 3-2 he would get classified as RLB or LLB and presumably reported in the "other" group.
 

He said "we have him for 3 sacks at 'other' positions" not 1.

I don't know what that means. There's no "other" designation in charting. It sounds like he's looking at some sort of proprietary report but I don't know what it is. Forgetting about the left side snaps still seems like the likeliest issue to me.
 

 
Also if this was true I think when he's on the left it's usually at DE in a 4 man front so the numbers I came up with would still be close.

I suspect your numbers are still close.

 

[Phragle]

I feel like we're talking in circles here.
 

You're talking about 23% of his sacks and 31% of his hurries happening in 6% of his snaps. That's hugely out of proportion. He's got three times as many sacks and hurries in "other" vs OLB in about 1/3 of the snaps; that doesn't seem right.

 
Maybe he should play standing up more, huh?
 
That's a joke.
 

Standing up vs 3-point stance is how PFF distinguishes between DE and OLB. If he was standing up in a 3-3 or 3-2 he would get classified as RLB or LLB and presumably reported in the "other" group.

 
That's probably the easiest way to do it, but it's not right and it's why we're talking in circles.
 

I don't know what that means. There's no "other" designation in charting. It sounds like he's looking at some sort of proprietary report but I don't know what it is. Forgetting about the left side snaps still seems like the likeliest issue to me.

 
Other is 4-3 OLB. It has to be.
 
Feel free to bring it up with Steve. 

 

[Super Nomario]

That's probably the easiest way to do it, but it's not right and it's why we're talking in circles.

It is how PFF does it. 2-point stance = LB, 3-point stance = DL.

 

[Kenny F'ing Powers]

It is how PFF does it. 2-point stance = LB, 3-point stance = DL.

 
By PFF standards, the BB teams of the mid 2000's would have 11 LB's on a few plays just about every game.
 
Sick.

 

[Super Nomario]

I noticed that Hoomanawanui is a -12.3 in run blocking. Then I took a closer look and noticed that tons of TE are negative in run blocking. Of the 64 TEs who've played 25% or more snaps, only 12 are +1 or more in run blocking and 47 are -1 or worse. Cumulatively all TE add up to -409.8 in run blocking. That seems crazy. I asked PFF:
 
Pro Football Focus ‏@PFF  4h
@davearchie Part of that is us redistributing some penalty grades so if you commit offensive holding in run blocking, goes in RB grade now
 
Pro Football Focus ‏@PFF  4h
@davearchie Will work to make all years the same in the offseason

 

[Super Nomario]

Division Game Weekend curiosities:
 
- Brady graded out negative for Saturday
- Despite Andrew Luck dropping back 48 times and not getting sacked, every member of the Colts OL graded negatively in pass blocking

 

[Devizier]

Division Game Weekend curiosities:
 
- Brady graded out negative for Saturday

 
Flacco, too. Does not compute.

 

[semsox]

That's not a curiosity, it's a bad model

 

[lexrageorge]

Truly indefensible. Garbage in, garbage out.

 

[DJnVa]

Have they tried to defend those numbers anywhere?

 

[Jnai]

Tom Brady, QB: -0.4
Breakdown: Connected on several extremely impressive throws, including on the go-ahead touchdown, but mostly took what the Ravens’ defense offered with 28 of his 33 completions coming on passes thrown within 10 yards of the line of scrimmage. Brady repeatedly picked on corner Rashaan Melvin, taking advantage of soft coverage and sending more than 15 targets into his coverage on the left side of the field.

 
What on earth is this shit.
 
The paragraph should read: "Quarterbacked his team to 35 points in a game in which the Patriots had 14 total rushing yards and no defensive scores."

 

[mascho]

IOW, Brady did his job.

And did it better than 99.9% of the humans on the planet ever could.

Totally deserving of a negative grade.

 

[rodderick]

I love how "took what the Ravens defense offered" is somehow a negative. And when will people realize that you shouldn't devalue a quarterback if he can get the job done with short throws? 

 

[Shelterdog]

I love how "took what the Ravens defense offered" is somehow a negative. And when will people realize that you shouldn't devalue a quarterback if he can get the job done with short throws? 

 
What they also don't seem to realize is that it's hard to take what the Ravens defense offers.  It's not like the Ravens tell you "yeah, we're not covering the slot receiver on a 5 yard out."  Brady has to diagnose the defense and make all sorts of adjustments to make sure your guy is going to be in the right place to get that completion, but apparently all that PFF notices is a 5 yard throw that's not worthy of comment.

 

[mascho]

I love how "took what the Ravens defense offered" is somehow a negative. And when will people realize that you shouldn't devalue a quarterback if he can get the job done with short throws? 

 
It makes you wonder. If Brady forced a bunch of throws into coverage and threw two picks, would he have received a better grade?

 

[Devizier]

 
It makes you wonder. If Brady forced a bunch of throws into coverage and threw two picks, would he have received a better grade?

 
Probably not; Flacco was -3 point something in their grading. Honestly, it seems like 0 is the best score a QB can get according to PFF.

 

[Super Nomario]

 
It makes you wonder. If Brady forced a bunch of throws into coverage and threw two picks, would he have received a better grade?

It worked for Andrew Luck!

 

[DJnVa]

Yes, Peyton Manning had a higher grade than Joe Flacco.

 

[Shelterdog]

Yes, Peyton Manning had a higher grade than Joe Flacco.

 
That's because they don't bow down to the conventional wisdom dipshit.

 

[Smiling Joe Hesketh]

JFC people, why is this still a topic? PFF is JUST MAKING THINGS UP. They have no expertise, no system, nothing but a bunch of amateurs sitting around and guessing. They don't know the player responsibilities on any play and thus cannot assign blame or credit intelligently.
 
They graded both Brady and Flacco negatively while Manning got a positive score. Just read that again. That's moronic, and their whole system should be thrown into the trash.
 
They suck, and it's a terrible shame that people and media still cite to them as if they conduct rigorous analysis. They do not. They are the Hughes 2.50 of football analytics.

 

[Vinho Tinto]

They are the Hughes 2.50 of football analytics.

 
Good callback.

 

[Ed Hillel]

I gave PFF a try this year, even paid their fee, but I am not a repeat customer. It's a load of hot garbage.

 

[Shelterdog]

JFC people, why is this still a topic? PFF is JUST MAKING THINGS UP. They have no expertise, no system, nothing but a bunch of amateurs sitting around and guessing. They don't know the player responsibilities on any play and thus cannot assign blame or credit intelligently.
 
They graded both Brady and Flacco negatively while Manning got a positive score. Just read that again. That's moronic, and their whole system should be thrown into the trash.
 
They suck, and it's a terrible shame that people and media still cite to them as if they conduct rigorous analysis. They do not. They are the Hughes 2.50 of football analytics.

 
They're ratings are obviously stupid (as one example they've said Connolly is one of the worst guards in all of football two years in a row: Belichick isn't spending $4 million and making a captain out of one of the worst players in football).  However I do think the player participation, sacks/hit/pressures, coverage numbers, and passing stats by area of the field are reasonably accurate and quite useful.

 

[Smiling Joe Hesketh]

 
They're ratings are obviously stupid (as one example they've said Connolly is one of the worst guards in all of football two years in a row: Belichick isn't spending $4 million and making a captain out of one of the worst players in football).  However I do think the player participation, sacks/hit/pressures, coverage numbers, and passing stats by area of the field are reasonably accurate and quite useful.

 
I agree with that, because there's no judgment involved there, just a ton of charting. Those are useful indeed. Anything involving football grading and player judgment is well beyond their ken and should be ignored by any fan with a functional cerebellum.

 

[Super Nomario]

They graded both Brady and Flacco negatively while Manning got a positive score. Just read that again. That's moronic, and their whole system should be thrown into the trash.

They did not grade Manning positively, FWIW.