PFF Game Grades

[Krangodnzr]

I learned to not take PFF too seriously when I saw them give David Johnson like an 80 grade for his 200+ yard game against the Eagles. It was probably the single best rushing performance I've seen from a Cardinals back in many years and PFF acted like it was a decent game.

What I have learned about PFF is that even a 60 grade isn't terrible. A 70 grade is a good player. 80 grade is a Pro Bowler. 90 is an elite, game changer.

 

[QuebecCard]

PFF are grading performance. If you get beat in coverage and the QB didn't see the open guy you still performed poorly.

Performance... So like figure skating, rhythmic gymnastics?

Sorry... too many subjective variables for my taste.

 

[Krangodnzr]

I personally think 63 is good.

Its game 1. He played nearly every snap. It's a major leap forward.

And he didn't play all snaps at ILB. He played some edge which may have dragged him down.

If this is 63 imagine him by game 6.

Yeah a 63 grade isn't bad. I do agree that their neutrality on every play being as meaningful is problematic.

I really like ESPN's QBR. Look at Jared Goff as an example. QBR exposes him as a below average to league average QB, but for a bit he was putting up big numbers. QBR takes into account whether a league average QB could step in and make the same play. Goff consistently got a 50 QBR, which means he is an average player, no better.

FYI Murray has a 79.6 QBR to start the year, which ranks 6th.

 

[Dr. Jones]

PFF Is stupid

PFF Is a parody account. Always poor.

Give me FBO please.

 

[Cbus cardsfan]

Pff literally has no idea what defensive scheme was called or what a guys role was on an individual play.
They can guess, but so can you or I.

 

[QuebecCard]

based on the fact we don’t even have scouts that read film at a high level, you pose an excellent question

Fact...?

OK... fine.

 

[BritCard]

Yeah a 63 grade isn't bad. I do agree that their neutrality on every play being as meaningful is problematic.

I really like ESPN's QBR. Look at Jared Goff as an example. QBR exposes him as a below average to league average QB, but for a bit he was putting up big numbers. QBR takes into account whether a league average QB could step in and make the same play. Goff consistently got a 50 QBR, which means he is an average player, no better.

FYI Murray has a 79.6 QBR to start the year, which ranks 6th.

QBR has its issues too. I feel it unfairly weights running QBs.

 

[Krangodnzr]

QBR has its issues too. I feel it unfairly weights running QBs.

Unfairly? Non-running QBs benefit from QB Rating. Why is it unfair to NOT consider rushing?

 

[BritCard]

Unfairly? Non-running QBs benefit from QB Rating. Why is it unfair to NOT consider rushing?

I'm not saying it isn't. I just think it's over weighted. The top of the rankings often seem dominated by runners.

 

[Krangodnzr]

I'm not saying it isn't. I just think it's over weighted. The top of the rankings often seem dominated by runners.

Dominated by runners who pass well lol.

Bridgewater isn't exactly a runner and he is top.

Second is Winston. First "runner" to pop up is either Mahomes if you consider him one (I don't) or Wilson at #5.

 

[kerouac9]

Yeah a 63 grade isn't bad. I do agree that their neutrality on every play being as meaningful is problematic.

I really like ESPN's QBR. Look at Jared Goff as an example. QBR exposes him as a below average to league average QB, but for a bit he was putting up big numbers. QBR takes into account whether a league average QB could step in and make the same play. Goff consistently got a 50 QBR, which means he is an average player, no better.

FYI Murray has a 79.6 QBR to start the year, which ranks 6th.

Britcard is right that all these systems should be evaluated on their merits and there's no magic bullet that reveals the "true" value of a player.

Britcard is wrong in seemingly always pointing exclusively to PFF's grades, which are often comically bad, and then pointing to their marketing copy for support whenever something is odd about their grading. NFL teams subscribe to PFF because their team does the grunt work of play charting that 32 teams used to all hire a $25k/year coaching intern or quality control assistant to do.

I like QBR some. I like that it does a better job of capturing rushing performance and weighting situation better than passer rating. It doesn't account for strength of opponent, so players that face a bunch of good defenses look naturally worse and players who face bad defenses get a boost. Also, last season 23 of 33 qualifying QBs had QBRs over 50. All the children are above average.

People love DVOA when it agrees with them and hate it when it doesn't. DVOA weights for situation and for strength of opponent. It works on a per-play basis, so a 12-yard completion on 3rd-and-22 doesn't mean as much (they call it a failed completion; Kyler lead the league in these last year).

What's a bummer about DVOA is that all of the models assume some real-world data at the front end, so for the first half of the season they're relying on their own preseason predictions to weight opponents. They're also super-upfront about what they can't measure -- like individual offensive lineman performance as contributions to the overall performance of the team. I like their "blown blocks" metric, though.

 

[BritCard]

Britcard is right that all these systems should be evaluated on their merits and there's no magic bullet that reveals the "true" value of a player.

Britcard is wrong in seemingly always pointing exclusively to PFF's grades, which are often comically bad, and then pointing to their marketing copy for support whenever something is odd about their grading. NFL teams subscribe to PFF because their team does the grunt work of play charting that 32 teams used to all hire a $25k/year coaching intern or quality control assistant to do.

I like QBR some. I like that it does a better job of capturing rushing performance and weighting situation better than passer rating. It doesn't account for strength of opponent, so players that face a bunch of good defenses look naturally worse and players who face bad defenses get a boost. Also, last season 23 of 33 qualifying QBs had QBRs over 50. All the children are above average.

People love DVOA when it agrees with them and hate it when it doesn't. DVOA weights for situation and for strength of opponent. It works on a per-play basis, so a 12-yard completion on 3rd-and-22 doesn't mean as much (they call it a failed completion; Kyler lead the league in these last year).

What's a bummer about DVOA is that all of the models assume some real-world data at the front end, so for the first half of the season they're relying on their own preseason predictions to weight opponents. They're also super-upfront about what they can't measure -- like individual offensive lineman performance as contributions to the overall performance of the team. I like their "blown blocks" metric, though.

I don't continually point to them. They are simply the only publicly available grading system for all round performance on every play.

While they aren't perfect they are at least consistent. If I want to know if X WR is better than Y WR it should be fair.

 

[kerouac9]

I don't continually point to them. They are simply the only publicly available grading system for all round performance on every play.

While they aren't perfect they are at least consistent. If I want to know if X WR is better than Y WR it should be fair.

It's your first reference to player quality almost every time. You point to PFF grades multiple times a day on here. Just be honest about it.

What does "fair" mean in this context? Because by no means is PFF "objective", which is it's biggest complication compared to actual analytics.

 

[Proximo]

What's the relevancy of any of Simmons plays that don't have a material impact on a given outcome?

I can see whether he made a tackle, broke up a pass, or failed at either. What else is there?

Blowing coverage leaving a guy open - you would only notice if the ball was thrown to that guy

The significance is even if he gets away with the mistake in this game, it is liable to be a problem in a future game, and it shows his overall mental command of the game.

 

[DVontel]

I don't continually point to them. They are simply the only publicly available grading system for all round performance on every play.

While they aren't perfect they are at least consistent. If I want to know if X WR is better than Y WR it should be fair.

You know you don’t even believe this.

 

[Proximo]

I'll try remember to do everyone

Defense

Marco Wilson - 62
Byron Murphy - 76.3
Robert Alford - 76.1
Budda Baker - 85.4
Jalen Thompson - 62.8

Isaiah Simmons - 63 (PFF on crack)
Jordan Hicks - 52
Zaven Collins - 72.5 (Run defense 90, coverage 41. This might explain why he vanished during passing downs)

Chandler Jones - 93.8
JJ Watt - 67.5
Markus Golden - 68
Zach Allen - 34.8 - Yikes
Leki Fotu - 42.2
Rashard Lawrence - 56.9
Michael Dogbe - 62.7
Corey Peters - 62.9

Those young D line guys really have a long way to go.

Offense coming up

No matter what the haters say about PFF, please keep posting these scores every week. I very much appreciate it.

 

[jf-08]

No matter what the haters say about PFF, please keep posting these scores every week. I very much appreciate it.

Same.

 

[BritCard]

It's your first reference to player quality almost every time. You point to PFF grades multiple times a day on here. Just be honest about it.

What does "fair" mean in this context? Because by no means is PFF "objective", which is it's biggest complication compared to actual analytics.

Ok, what other metric would you like me to reference when comparing the all round performance of a player against another player?

What makes me laugh with PFF haters is they always point to that one outlier grade as proof that PFF is garbage while ignoring the remaining 95% they agree with.

Look at our WR grades. Nobody has issue with those. Nuk, Kirk, Moore all over 80. Our CB and safety grades all seem on the money. O line grades all seem on the money.

Poor as some of them are even D line grades look right because the players with poor grades were invisible on Sunday.

The one grade everyone is complaining about is Simmons because between grabbing chips and posting in the gameday thread we saw Simmons make 5 or 6 nice plays out of 60+.

Do you not think as all the other grades look right they might be right on Simmons too? (Which by the way was a decent grade considering the number of snaps he played and on all 3 downs and a step in the right direction).

It's the same over a season. The Mahomes's, Kelce's, Nuk's, Hill's and Kamara's of the league always have grades you would expect for their talent levels. If the system was flawed that wouldn't be the case.

 

[kerouac9]

Ok, what other metric would you like me to reference when comparing the all round performance of a player against another player?

What makes me laugh with PFF haters is they always point to that one outlier grade as proof that PFF is garbage while ignoring the remaining 95% they agree with.

Look at our WR grades. Nobody has issue with those. Nuk, Kirk, Moore all over 80. Our CB and safety grades all seem on the money. O line grades all seem on the money.

Poor as some of them are even D line grades look right because the players with poor grades were invisible on Sunday.

The one grade everyone is complaining about is Simmons because between grabbing chips and posting in the gameday thread we saw Simmons make 5 or 6 nice plays out of 60+.

Do you not think as all the other grades look right they might be right on Simmons too? (Which by the way was a decent grade considering the number of snaps he played and on all 3 downs and a step in the right direction).

It's the same over a season. The Mahomes's, Kelce's, Nuk's, Hill's and Kamara's of the league always have grades you would expect for their talent levels. If the system was flawed that wouldn't be the case.

Dude, you’ve just defined confirmation bias. It it tells you want you expect to see, then what’s the point?

I disagree with the premise of PFF grades and their process. They’re not “metrics” — they’re the averages opinion of a bunch of people. No different than this board.

 

[Krangodnzr]

By the KTBGS (Krangthebrain grading system), Simmons got like an 80.

My grading system is similar to PFF except I give big weight to splash plays and Simmons made a few of those.

The problem with PFF is that a yeoman guy who is always in the right position could score a 80, while a guy who wasn't in the right position all of the time but made 2 big plays could score a 65. Thats why you'll routinely see guys that are only ok score high. I remember one game where Zach Allen scored really high and he didn't really do squat in a loss.

 

[BritCard]

By the KTBGS (Krangthebrain grading system), Simmons got like an 80.

My grading system is similar to PFF except I give big weight to splash plays and Simmons made a few of those.

The problem with PFF is that a yeoman guy who is always in the right position could score a 80, while a guy who wasn't in the right position all of the time but made 2 big plays could score a 65. Thats why you'll routinely see guys that are only ok score high. I remember one game where Zach Allen scored really high and he didn't really do squat in a loss.

Wait, what?

The guy that was always in the right position and did the right things should score high, no matter who he is. And the guy that made 2 big plays but screwed up the other 60 should score low.

 

[SoonerLou]

Offense

Kyler Murray - 84.5

Chase Edmonds - 63.8
James Conner - 63.3

Maxx Williams - 54.7
Demetrius Harris - 55.3
Darrell Daniels - No grade

Hump - 78.7
Pugh - 67.8
Hudson - 52.7 (Only 28.5 pass blocking)
Jones - 42.5 (65 pass blocking)
Beachum - 67.4
Murray - 66.4

Hopkins - 81.8
Green - 55.9 (Nothing to see here, quickly move on)
Kirk - 85.9
Moore - 80.9

When did we last have a QB and 3 WR's over 80 grades?

Obviously not gospel, but I really hope Hudson works out. @Chopper0080 I really believe Creed Humphrey would have fell to our pick in the 3rd round since Hudson was likely to sign with the Chiefs.

 

[SoonerLou]

I think if a site gains access to actually game day grades from teams that would be huge.

Surprised the NFL hasnt set that up actually. Imagine if the Cardinals delievered their own grades to a site every wednesday.

 

[Krangodnzr]

Wait, what?

The guy that was always in the right position and did the right things should score high, no matter who he is. And the guy that made 2 big plays but screwed up the other 60 should score low.

Thats not exactly what I'm saying.

I've seen guys who did almost nothing in a game get 80 grades.

PFF does not, even by your admission, give grade bumps based on context. Context matters in sports! It matters a lot.

Thats why you have some fans who think of NBA player Robert Horry as some great player. He wasnt. He was a role player who hit a few big shots and some people think he is HOF worthy.

Simmons made a handful of ordinary big plays (his tackle one on one in the hole on Henry) and had a big interception.

I've seen this a lot from PFF so I take them with a massive grain of salt. I'm not saying to fully disregard, but they aren't perfect at all.

 

[football karma]

i will plead guilty to using/ posting PFF grades many, many times

its a handy source for objective analysis: "objective" defined as a source that doesn't care about the Cardinals or any other NFL team.

At their worst, PFF grades are subjective evaluations dressed up as objective, scientific ones.

At the same time, they are something and generally track the eye test. They are a data point that can make you question your priors on a given player.