2021 Comprehensive Draft Thread

[PartyPosse]

1 hour ago, Burgundy Yoda said:

I'm not saying Jamal Adams isn't a good player, I just feel like he's very overrated. He's good at his job but I feel like his position isn't nearly as important as Minkah's. Minkah Fitzpatrick completely shuts down the deep ball and the pass on his side of the field. He's amazing, and really actually reminds me of ST. He definitely plays on a more stacked defense but that shouldn't take away from his play. 

I’m not gonna tell you you’re wrong because you can make a damn good case for either one of them. But I do know Minkah had the luxury of knowing Haden, who perhaps had his best season as a pro, was his deep ball protection, which made it much easier. Lord knows who the Jets shutdown corner was last year. I don’t think they even have any returning corners. 

[DWinzit]

1 hour ago, PartyPosse said:

I’m not gonna tell you you’re wrong because you can make a damn good case for either one of them. But I do know Minkah had the luxury of knowing Haden, who perhaps had his best season as a pro, was his deep ball protection, which made it much easier. Lord knows who the Jets shutdown corner was last year. I don’t think they even have any returning corners. 

Either one of them elevate the Washington Watchamacallits to a new level.

Just think of one of them with that Dline.

[Going Commando]

14 hours ago, HTTRDynasty said:

@stevemcqueen1 That Steelers defense was getting abused before the Minkah trade.  Largely due to that trade though, they ended up ranked 3rd in defensive DVOA last year, which is their highest ranking since Polamalu was still in his prime in 2010.

 

I still have my doubts about the whole Coverage > Pass Rush argument, but coverage is definitely more important than run defense, which is where Payne excels.

 

PFF recently released an article which ranked positional importance by PFF WAR (quoted the relevant part below): https://www.pff.com/news/draft-why-positional-value-matters-in-the-nfl-draft

 

I'm not saying I agree with their findings, but I welcome anyone who is willing to challenge conventional ways of thinking, and for all we know, it could be a more accurate rankings list than conventional wisdom would have it.

 

PFF WAR is completely useless.  They should have scrapped it based on some of the goofy results and conclusions it was producing, but they're a subscription service who gets most of their value from being broadly used, especially by industry insiders like agents and GMs, and I suspect they're pretty desperate to stake their analytic value on producing a widely referenced and comprehensive stat like WAR (that also is completely reliant on their proprietary grading system/data collection to be produced).

 

The sport utterly defies that kind of analytic approach though.  PFF grading does a fair job of making apples to apples comparisons of performance of guys playing largely similar positions/roles, but it can not produce apples to oranges comparisons that are worth a damn.  Reading through that paper, I think I finally figured out one of the reasons why the player valuations produced by their WAR stat are so ****ed up.  There are systemic flaws in their grading systems that underrate line and run game play when compared to other positions, and lead to crazy fluctuations/variance in DB and WR grades--which they acknowledge but make no real attempt to suss out why it's happening and what problems it creates for their analysis.  They basically just accept it as a given or a product of accurate evaluation and move on. 

 

One really big flaw is that they weight far too much of their grades based on the distance the ball travels on a snap.  Say you have two scoring drives, one of which takes four plays and travels 80 yards and the other of which takes 13 and travels 60.  In PFF's system, that four play drive would massively boost up the grades of offensive skill guys and the QB and kill the back seven defenders while the 13 play drive would do very little to the grades of everyone.  This despite the fact that, in a one or two score game outside of two minute drills (the context in which a majority of NFL football is played), that 13 play drive is indicative of a much more dominant and valuable team performance.  The kind of performance that, if sustained, wins tons of games against every kind of opponent, in every month of the NFL season, and in every kind of environment.  And that kind of performance from a team is so wholly collaborative that you can not properly value the contributions of the people involved in producing it with the kind of grading system they must use out of necessity.

 

There is a reason why conventional wisdom is widely accepted in fooball.  Being iconoclastic and pro-analytics in a moment when analytics have been celebrated in other sports suits PFF's agenda to create buzz for and sell their service, but they've crapped out a bunch of bad analysis recently as a result of trying to push this WAR metric.  But no defensive coach whose job actually ****ing depends on results would underrate the value of line play or the run game like PFF does.  All of them will tell you that the absolute first goal in the design of their scheme is to properly and consistently fit the run and stop it, and that everything is built up from there.

[carex]

I wonder no college season would mean no one could declare early, with only people who were going to be seniors this year eligible

[HTTRDynasty]

22 minutes ago, stevemcqueen1 said:

 

PFF WAR is completely useless. 

 

Perhaps.  It may be useless, or it may be ahead of its time.  I know PFF grades as a whole were clowned on for years before it became so widely accepted that every NFL team now has a subscription and players and teams use it in contract negotiations.  It may be the same with PFF WAR, which has already been studied and refined by those that coach, play and draft in the NFL.

 

Like I said though, I personally don't agree with all of their conclusions, but I do agree a great safety has more of an impact than a great 1-tech in today's league.

[Always A Commander Never A Captain]

Are PFF grades still not clowned on?  Their grades are subjective and not objective.

 

Some of their stats that don't tie into grades are quality.  But the grades themselves kind of blow.  That's the nature of football though.  There's hidden factors at play that aren't there for baseball.

 

How would anybody, PFF included, know the exact assignments/technique the coaches put a game plan in to account for this formation on this down/distance.

 

You've got a player dinged for something that wasn't his responsibility or role at all, because he's doing what the coaches told him to do, as the coaches thought it would work.

 

Remember when we traded for HaHa Clinton-Dix and he was graded as a top 5 safety but played like ass for us?  Then someone went back and looked at his pre-trade film and it was just about the same thing.  Difference was a few lucky breaks went his way.  Poor throws and the like.

 

[PartyPosse]

Everyone acts like the Steelers were going to be the worst defensive team in the history of football before Minkah but they played all of two games before the trade against two pretty good teams. They were eventually going to figure it out with or without Minkah.

[HTTRDynasty]

36 minutes ago, Alcoholic Zebra said:

Are PFF grades still not clowned on? 

 

Nowadays?  Mostly by fans when it doesn't suit their argument.  I haven't seen them clowned on anywhere close to the degree they were several years ago though by players/coaches.  Likely due to the fact that PFF has proven their grades are more explanatory and predictive than most other analytics out there.  For example, a QB who throws a lot of interceptable passes without throwing a lot of INTs will be lauded by the average fan, who relies on traditional stats to evaluate players.  Whereas PFF's grade will more accurately reflect how the QB actually performed by accounting for the dropped INT.

 

 

 

Do they miss on occasion?  Of course.  But by and large, their grades make more sense than other rankings and evaluations based on traditional stats or purely objective measurements.

 

58 minutes ago, PartyPosse said:

They were eventually going to figure it out with or without Minkah.

 

If only someone would have told the Pittsburgh braintrust that before making the trade.  They could have saved themselves a first round pick.

[PartyPosse]

32 minutes ago, HTTRDynasty said:

If only someone would have told the Pittsburgh braintrust that before making the trade.  They could have saved themselves a first round pick.

He made them better in the long term but Ben was already ruled out for the season so I doubt they had serious aspirations with Mason Rudolph or Duck Hodges under center. Does Minkah make them a better team? Sure. No doubt. Is Minkah the main reason the defense ended up balling out all year? That’s debatable. Again, good player, but there’s also an argument to be made he was the third best player on the defense. 

[HTTRDynasty]

34 minutes ago, PartyPosse said:

He made them better in the long term but Ben was already ruled out for the season so I doubt they had serious aspirations with Mason Rudolph or Duck Hodges under center. Does Minkah make them a better team? Sure. No doubt. Is Minkah the main reason the defense ended up balling out all year? That’s debatable. Again, good player, but there’s also an argument to be made he was the third best player on the defense. 

 

An argument could also be made Earl Thomas was the 3rd best defender when the Seahawks were so dominant in the 2010's.  That doesn't take away from the fact that the defense wouldn't have been elite without him.

 

As I said before, the last time the Steelers defense ranked top 3 in defensive DVOA was when Troy Polamalu was in his prime.  Their best ranking was 9th since then.  That's not a coincidence.  Cam Heyward has been there since 2011.

[PartyPosse]

50 minutes ago, HTTRDynasty said:

 

An argument could also be made Earl Thomas was the 3rd best defender when the Seahawks were so dominant in the 2010's.  That doesn't take away from the fact that the defense wouldn't have been elite without him.

 

As I said before, the last time the Steelers defense ranked top 3 in defensive DVOA was when Troy Polamalu was in his prime.  Their best ranking was 9th since then.  That's not a coincidence.  Cam Heyward has been there since 2011.

Again, i'm not denying he's a good player and definitely contributed to the turnaround, but it sounds like you believe they still would have stunk all year if they didn't make the trade. 

[HTTRDynasty]

10 minutes ago, PartyPosse said:

Again, i'm not denying he's a good player and definitely contributed to the turnaround, but it sounds like you believe they still would have stunk all year if they didn't make the trade. 

 

No, I believe they would've hovered around the top 10-15 again like they had since 2012.  But there's a massive difference in ranking top 3 vs top 10-15, and that jump is mostly due to the addition of Minkah.

[Always A Commander Never A Captain]

2 hours ago, HTTRDynasty said:

 

Nowadays?  Mostly by fans when it doesn't suit their argument.  I haven't seen them clowned on anywhere close to the degree they were several years ago though by players/coaches.  Likely due to the fact that PFF has proven their grades are more explanatory and predictive than most other analytics out there.  For example, a QB who throws a lot of interceptable passes without throwing a lot of INTs will be lauded by the average fan, who relies on traditional stats to evaluate players.  Whereas PFF's grade will more accurately reflect how the QB actually performed by accounting for the dropped INT.

 

 

Do they miss on occasion?  Of course.  But by and large, their grades make more sense than other rankings and evaluations based on traditional stats or purely objective measurements.

 

I guess it's good that they're at the top of the list for correlation between other stats.  But the difference is 0.01, which is nothing.  It starts being something the closer you get to -1/1, but it's in the .4's and only has 11 seasons to compare.  Who knows, after 2020, the correlation value for their grades could fall to 0.4 and be at the bottom or rise to 0.5 and still be at the top.  Right now all I'd say is that none of these stats are that good at predicting future success.  It's basically saying that teams change too much to replicate the same level of success from one season to another.

 

Dropped/Tipped/etc interceptions are whacky.  Jameis Winston threw 30 last season, but if you go by this, it could have risen to 40.  Didn't realize it, but our QB Haskins looks pretty solid here (especially given the context).  Granted, this is a subjective analysis of what constitutes a dropped interception.  You could put those same samples in front of someone else, and the numbers would probably differ by 1 or 2 per QB.

https://www.footballoutsiders.com/stat-analysis/2020/adjusted-interceptions-2019

 

I mean, when talking about traditional stats, they're fun but don't mean a whole lot.  Kirk Cousins threw for 4,900 yards once, but we still missed the playoffs, and nobody would say we had an elite passing attack that you might assume with 4,900 passing yards.

 

DVOA is good because it completely removes the subjective, it's purely objective.  There are issues with it, but that's because it's only objective.  Once you start putting subjective analysis into a stat, it should never be viewed as infallible.  Which is what PFF advocates try and say.  There's also the issue with lack of transparency.  ESPN's QBR suffers the same fate here, although they come across a bit better from a transparency viewpoint, but still not well.

 

Did HaHa Clinton Dix deserve the lofty PFF grades he got before we traded for him?  When looking at the tape of his games with the Packers, no, absolutely not.  There's too much inherent bias in subjective stats, and in PFF's case, lack of knowledge.

 

What about how they view the challenge of the assignment?  Check out Darius Slay in 2019.  PFF said he stunk and was not a starting caliber corner.  Yet that was followed with analysis from other players/bloggers/whoever, sometimes showing tape, that PFF doesn't know to adjust grades.  Since we don't have all of the info from PFF, we can only guess it's some combination of errors based on difficulty of assignments within scheme, difficulty of matchup, difficulty vs throw, difficulty vs down/distance.

 

Josh Norman last season had some completions let up where the fault basically lied pre-snap with his technique and leverage.  Was that his fault deciding the leverage?  Was it the fault of the coaching telling him to play that leverage in this spot/situation?  Was Josh/coaches both actually correct but another player botched their assignment meant to cover the direction of the leverage?  Unless players/coaches come out and say who screwed up, we as fans can't know.  And PFF can't know either, because no team would ever give out that information to external sources.

 

As a subjective grade, it should never be taken alone, but always used in conjunction with objective grades.  Otherwise it can paint a misleading or sometimes flat out wrong picture.  

 

Take Darius Slay again.  PFF says not a starting caliber corner.  Let's do an objective stat, QB Rating when thrown at.  That was 81.6, which is clearly starting caliber worthy and probably even above average.  Then you take what you know of the Lions scheme and how they used him and form your own subjective opinion comparing PFF, an objective stat, and what you know of the context.

 

 

Edited July 27, 2020 by Alcoholic Zebra

[PartyPosse]

5 minutes ago, HTTRDynasty said:

 

No, I believe they would've hovered around the top 10-15 again like they had since 2012.  But there's a massive difference in ranking top 3 vs top 10-15, and that jump is mostly due to the addition of Minkah.

I think you're really underestimating what Haden and Watt did. There's a reason Watt was third in DPOY and a reason Devin Bush was third in DROY. That team was stacked on that side of the ball. Dupree had a breakout year with 11.5 sacks and the aforementioned Haden had his best season maybe ever. Minkah probably boasted them a spot or two but as good as he is there is no chance he's that all-worldly that he singlehandedly turned a middle-of-the-pack team into a top three team. If he did he would have been the unanimous DPOY instead of not getting a single vote.

[HTTRDynasty]

1 minute ago, Alcoholic Zebra said:

 

I guess it's good that they're at the top of the list for correlation between other stats.  But the difference is 0.01, which is nothing.  It starts being something the closer you get to -1/1, but it's in the .4's and only has 11 seasons to compare.  Who knows, after 2020, the correlation value for their grades could fall to 0.4 and be at the bottom or rise to 0.5 and still be at the top.  Right now all I'd say is that none of these stats are that good at predicting future success.  It's basically saying that teams change too much to replicate the same level of success from one season to another.

 

Are you saying PFF grades vs. TD/INT ratio isn't a big difference?  That's what I was comparing it to.  The other stats are still inferior, but are much better than traditional stats like TD/INT ratio.

 

2 minutes ago, Alcoholic Zebra said:

Dropped/Tipped/etc interceptions are whacky.  Jameis Winston threw 30 last season, but if you go by this, it could have risen to 40.  Didn't realize it, but our QB Haskins looks pretty solid here (especially given the context).  Granted, this is a subjective analysis of what constitutes a dropped interception.  You could put those same samples in front of someone else, and the numbers would probably differ by 1 or 2 per QB.

https://www.footballoutsiders.com/stat-analysis/2020/adjusted-interceptions-2019

 

Yes, I'm well aware of that adjusted interceptions article.  I've used it many times against all the Daniel Jones stans out there on other sites.  Adjustable interceptions have been proven as more predictive than just looking at interception totals on its own.

 

7 minutes ago, Alcoholic Zebra said:

 

DVOA is good because it completely removes the subjective, it's purely objective.  There are issues with it, but that's because it's only objective.  Once you start putting subjective analysis into a stat, it should never be viewed as infallible.  Which is what PFF advocates try and say.  There's also the issue with lack of transparency.  ESPN's QBR suffers the same fate here, although they come across a bit better from a transparency viewpoint, but still not well.

 

Actually, PFF is much more transparent than QBR.  It's definitely not infallible, but they do a good job of explaining their process for each position.  Where exactly are you seeing a lack of transparency?  Maybe you're looking in the wrong places.

 

9 minutes ago, Alcoholic Zebra said:

Did HaHa Clinton Dix deserve the lofty PFF grades he got before we traded for him?  When looking at the tape of his games with the Packers, no, absolutely not.  There's too much inherent bias in subjective stats, and in PFF's case, lack of knowledge.

 

People love to use HHCD as some "ha, gotcha" example against PFF, but he ranked 19th last year, 17th the year before that and 37th in both of the two years before that.  It's not like they've had him ranked as some super stud.  And note: those rankings don't even include all the players who ranked higher than him that played less than 300 snaps that season.  Is he the 17th best safety in the league?  Probably not.  Has he been somewhere between 17th and 37th best throughout his career?  Yes, I would say so.

 

14 minutes ago, Alcoholic Zebra said:

 

What about how they view the challenge of the assignment?  Check out Darius Slay in 2019.  PFF said he stunk and was not a starting caliber corner.  Yet that was followed with analysis from other players/bloggers/whoever, sometimes showing tape, that PFF doesn't know to adjust grades.  Since we don't have all of the info from PFF, we can only guess it's some combination of errors based on difficulty of assignments within scheme, difficulty of matchup, difficulty vs throw, difficulty vs down/distance.

 

PFF has discussed their view on Slay in several articles.  They still view him as a top-tier CB who had a down year.  He had similar assignments in all the years PFF had him rated highly - so I doubt it's some massive error on their part and more likely players/bloggers displaying an anchoring bias based on how they've viewed Slay throughout his career, rather than just focusing on 2019, when getting angry at Slay's 2019 grade.

 

18 minutes ago, Alcoholic Zebra said:

Josh Norman last season had some completions let up where the fault basically lied pre-snap with his technique and leverage.  Was that his fault deciding the leverage?  Was it the fault of the coaching telling him to play that leverage in this spot/situation?  Was Josh/coaches both actually correct but another player botched their assignment meant to cover the direction of the leverage?  Unless players/coaches come out and say who screwed up, we as fans can't know.  And PFF can't know either, because no team would ever give out that information to external sources.

 

I mean, they had Quinton Dunbar as their #2 ranked CB last year, so I'm inclined to think the fault lied with Norman rather than the coaches - especially considering how often he was benched.  PFF can't know everything.  But I think it's foolish to completely dismiss an analysis just because they don't get something right 100% of the time.  PFF was actually the first to recognize Josh Norman for his elite play well before he had made a name for himself among players/coaches/fans.  They go through an extensive grading process that has multiple redundancies, including former coaches who review their grades.  

 

Are they infallible?  No way.  Neither are NFL GMs.  

[HTTRDynasty]

 

54 minutes ago, PartyPosse said:

I think you're really underestimating what Haden and Watt did. There's a reason Watt was third in DPOY and a reason Devin Bush was third in DROY. That team was stacked on that side of the ball. Dupree had a breakout year with 11.5 sacks and the aforementioned Haden had his best season maybe ever. Minkah probably boasted them a spot or two but as good as he is there is no chance he's that all-worldly that he singlehandedly turned a middle-of-the-pack team into a top three team. If he did he would have been the unanimous DPOY instead of not getting a single vote.

 

I never said he singlehandedly turned them around.  I said he was the biggest factor for their jump in ranking.  Watt and Haden have been with the Steelers since 2017, Dupree since 2015.    Fitzpatrick was the missing piece.  Here's a good article explaining his impact:

 

HOW MINKAH FITZPATRICK IMMEDIATELY IMPACTED STEELERS' DEFENSE

On offense, one star in a group of 11 can make all the difference. Even if it’s not the quarterback, if a team has an elite running back or an elite wide receiver, at times, they can take over and still have success—even if the group around them might not be up to par.

 

It doesn’t exactly work like that on defense. If you have an elite pass rusher or an elite shutdown corner, sometimes they can really impact the game by themselves. But, for the most part, even with a great player at either of those spots, offenses can typically just game plan to neutralize the impact any one player can have.

 

On defense, it really does take all three levels of the field working together to be productive. A one-man difference-maker on defense doesn’t point to stats as the element that proves they’re a difference-maker; it’s that the fact that their presence elevates the other 10 players on the field so much so that the team stats improve across the board. 

That’s exactly what happened when the Pittsburgh Steelers traded for Minkah Fitzpatrick last season. 

 

Fitzpatrick, the former No. 11 overall pick in the 2018 NFL Draft had been butting heads with the Miami Dolphins’ brass for quite a while before the trade. The root of that conflict was Fitzpatrick’s use on the field.

 

Fitzpatrick came into the NFL billed as one of the most versatile defensive players in the 2018 draft class. In Nick Saban’s Alabama defense, Fitzpatrick played free safety, strong safety, outside cornerback, nickel cornerback, and even had some linebacker responsibilities, at times.

 

The Dolphins saw that versatility and fell in love, but you can’t play every position every down. Eventually, you have to commit to a main role. The Dolphins did that, but Fitzpatrick believed they chose the incorrect one. 

 

“[Miami] played me real close to the line of scrimmage, sometimes on the line of scrimmage, in the box playing the linebacker/strong safety type guy”, Fitzpatrick said. “In practice, I was doing the same. I wasn’t really doing coverage; I was just taking on blockers and just learning how to adjust to different run schemes and stuff like that. And it was good and all, but I just didn’t think he was maximizing my skill set.”

 

This went on long enough that there was a breaking point. That breaking point came to a head when the Dolphins shipped Fitzpatrick off to Pittsburgh for a 2020 first-round pick just two years after Miami spent a first-round pick to get him. For the Dolphins, not bad for trying to return something past the 30-day return policy. For the Steelers, that price tag was absolutely worth it.

 

Before Fitzpatrick arrived in Pittsburgh, the Steelers allowed 320 passing yards per game in their first two games of the season. With Fitzpatrick in their lineup, they allowed just 176.7 passing yards per game the rest of the season. For further evidence, pre-Fitzpatrick, the Steelers’ defense gave up 30.5 points per game, 125 rushing yards per game, and had a point differential of minus-32. With Fitzpatrick, the defense allowed 17.3 points per game, 107.4 rushing yards per game, and had a point differential of plus-18.

 

The Steelers also led the league with 38 takeaways last season. Fitzpatrick was a heavy contributor to that with five interceptions, one forced fumble, and two fumble recoveries. A big reason for that—in contrast to his lack of production in Miami—is because the Steelers saw him as a different player than the one the Dolphins tried to deploy.

 

“When I came to Pittsburgh, they immediately just plugged me into the free safety spot and said, ‘this is going to maximize your skill set by playing you back here’”, Fitzpatrick said. “‘You’re an instinctive guy, you have range, you’re athletic and you go and get the ball, you like being around the ball. And we think that free safety is going to maximize that skill set.’”

 

 

Instincts over the middle are a highly coveted trait. It’s why you see single-high type safeties get prioritized in the draft. If you have range and recognition, the good defensive coordinators will find a way to build their back end around the ground you can cover and the impact you can have.

 

The play above is one of many examples of what Fitzpatrick can do when you allow him to focus on the quarterback while in space on the back end. Fitzpatrick understands how to cut routes and bait quarterbacks. That stuff tends to happen less when you don’t put him in the position to be able to do that.

 

 

Over the course of his first seven games with the Steelers, Fitzpatrick notched seven takeaways in the form of interceptions and fumble recoveries. He should have had another one in the play above. Eventually, teams got wise and just stopped throwing his way, as Fitzpatrick was only targeted seven times in the last eight games, per Mark Madden.

 

 

Turnovers are not the only way to judge how successful Fitzpatrick is when allowed to roam as a free safety. When you let him freelance and read the offense, you get results like the one above, where he flies all the way across the field to break a pass up. This play happened because Fitzpatrick was put in the position to see the field as a whole.

 

Fitzpatrick yielded just a 46.3 passer rating over the course of his 14 games in Pittsburgh. For context on how good that is, Stephon Gilmore, who won Defensive Player of the Year in 2019, allowed a passer rating of 44.1.

 

 

When you allow Fitzpatrick to sit back and scan the entire field, you allow him to truly elevate the rest of the defense.

 

Take the play above as an example. In that play, the Steelers had the Ravens offense in a dangerous spot up against their own end zone. The Ravens, who are already a strong running team, were likely just going to hand the ball off. In order to counter their heavier trench on that play, the Steelers had to occupy extra defenders in the box. Anytime you do that, you likely have to make sacrifices in coverage. But when you have a player like Fitzpatrick on the back end, you have that kind of freedom.

 

As seen above, it made all the difference.

 

Fitzpatrick leading the NFL in takeaways in 2019 had plenty of luck. A handful of his turnovers were him being in the right spot at the right time and the ball bouncing his way. Even though it is luck, you still have to be around the ball and in position to capitalize when luck finds favor with you—luck can’t do both on its own. 

 

Fitzpatrick constantly puts himself in positions on the field where his presence can have the most impact. In fact, him just being on the field is a big part of what made Pittsburgh’s defense one of the best in 2019.

 

https://thedraftnetwork.com/articles/minkah-fitzpatrick-pittsburgh-steelers

Edited July 27, 2020 by HTTRDynasty

[HTTRDynasty]

And here's another article on Minkah's impact:https://theathletic.com/1370602/2019/11/14/teds-film-room-minkah-fitzpatricks-impact-on-the-steelers-defense/

 

Ted’s Film Room: Minkah Fitzpatrick’s impact on the Steelers’ defense

By Ted Nguyen  Nov 14, 2019 13 

 

One of the reasons 2018 first-round pick Minkah Fitzpatrick wanted out of Miami was because he wanted to focus on one position — he wanted to be a free safety. After their first two games of the 2019 season, the Pittsburgh Steelers were giving up a ridiculous 8.85 yards per pass attempt — they were desperate for a playmaker in the secondary. So on September 16, the 0-2 Steelers traded away a 2020 first-round draft pick and swapped a couple of lower-round picks with the Dolphins for Fitzpatrick. Many critics raised their eyebrows at the move — Ben Roethlisberger was out for the year and the season looked lost. What if that pick they just traded was a top-10 selection?

 

Seven weeks later, ahead of their Thursday night game against the Browns, the Steelers’ defense is one of the best in the league (third in defensive DVOA) and 5-4  Pittsburgh is in the thick of the playoff hunt. If the season ended today, the Steelers would be a wild card team. In seven games with the Steelers, Fitzpatrick has 8 passes defensed, 5 interceptions, 1 forced fumble and 2 defensive touchdowns.

 

I asked Steelers film analyst Alex Kozora of Steelers Depot for his thoughts on Fitzpatrick’s impact on their defense and here’s what he had to say:

 

“Fitzpatrick has brought a level of playmaking in the secondary the Steelers literally haven’t seen since Troy Polamalu. The numbers don’t lie. Fitzpatrick is the first Steeler to have at least four interceptions in a season since Polamalu in 2010, a nine-year span that broke the longest drought in NFL history. He’s the first Steelers safety with two defensive TDs in a season since Carnell Lake in 1986. The Steelers’ biggest issue last year was a complete lack of takeaways.  No team dropped more interceptions than they did and they finished the year with 15 turnovers. This year, they’re at 26, the most through nine games since 1987, and he’s a big reason behind that. The defense is the strength, which is good timing given how much the offense has struggled.”

 

...

 

“It is football justice as far as we are concerned,” said Steelers head coach Mike Tomlin after the Steelers’ 17-12 Week 10 win over the Rams. “It is not anything mystical. (Fitzpatrick) doesn’t have a rabbit’s foot in his pocket. He is preparing and taking advantage of opportunities. It is reasonable to expect it to continue, provided we are in an environment where those plays happen. You are up on people, you get them one-dimensional. You get more opportunities for those type splash plays. It is not occurring in a vacuum for him. It is part of how games are unfolding as well.”

 

...

 

Fitzpatrick bet on himself. He thought that he could make the biggest impact as a free safety and demanded a trade and so far, he’s been spectacular in that role. The Steelers also took a gamble by trading a first-round pick for Fitzpatrick when they were 0-2 and so far, they’ve come away looking like big winners in the trade because they finally have an instinctual, play-making safety that their defense has sorely needed since Polamalu retired.

 

Edited July 27, 2020 by HTTRDynasty

[Always A Commander Never A Captain]

40 minutes ago, HTTRDynasty said:

 

Are you saying PFF grades vs. TD/INT ratio isn't a big difference?  That's what I was comparing it to.  The other stats are still inferior, but are much better than traditional stats like TD/INT ratio.

The sample size is only 11 for that stat, and the difference in correlation is negligible.  Saying they're lesser/worse/inferior as a statement is weird.  Saying TD/Int or some basic counting stat is worse, sure.  But the others?

 

40 minutes ago, HTTRDynasty said:

Actually, PFF is much more transparent than QBR.  It's definitely not infallible, but they do a good job of explaining their process for each position.  Where exactly are you seeing a lack of transparency?  Maybe you're looking in the wrong places.

Serious question...do you work for PFF?  It sure seems like I'm talking to PFF sales here.

 

Can you link me to the grading process they would use for an outside corner.  Pre-snap alignment, pre-snap leverage, self-assignment, assignment of help defenders in area, then responsibilities post-snap, coverage technique after the WR release.  How they lay the fault on coaching or the player.  Some of this stuff PFF can not know, but they pretend to.

 

40 minutes ago, HTTRDynasty said:

 

 

People love to use HHCD as some "ha, gotcha" example against PFF, but he ranked 19th last year, 17th the year before that and 37th in both of the two years before that.  It's not like they've had him ranked as some super stud.  And note: those rankings don't even include all the players who ranked higher than him that played less than 300 snaps that season.  Is he the 17th best safety in the league?  Probably not.  Has he been somewhere between 17th and 37th best throughout his career?  Yes, I would say so.

 

I'm not talking about full season, I'm talking about his grades earlier that season with the Packers.  The grade below was just after we traded for him.  After the trade, someone went back to look at his Packers film, compared it to his Redskins film, and determined there must have been an issue with the PFF grades, as the film showed the same errors/issues with the Packers that he had with us.  Unfortunately I can't find that article right now.

 

 

 

40 minutes ago, HTTRDynasty said:

PFF has discussed their view on Slay in several articles.  They still view him as a top-tier CB who had a down year.  He had similar assignments in all the years PFF had him rated highly - so I doubt it's some massive error on their part and more likely players/bloggers displaying an anchoring bias based on how they've viewed Slay throughout his career, rather than just focusing on 2019, when getting angry at Slay's 2019 grade.

 

It's not talking about his career as a whole, or what he'll do in 2020...it's just talking about what happened in 2019.  The point is PFF viewed him as having a below average year, yet others have responded by saying that can't be correct given his assignments, his technique, the difficulty of catches he forced WR's to make, etc.

 

He's had the same assignments in all the years PFF had him rated highly (I think I've read that was 5 years?), it's certainly possible, but the Lions have had 2 different coaching staffs over that time frame, and I have a hard time believing his assignments were that consistent play to play across multiple years.

 

But given that the Lions played the most man coverage in the NFL in 2019...I think I want to call shenanigans.  I have doubts they played that much man coverage throughout Slay's 5 or 6 years being good.

 

40 minutes ago, HTTRDynasty said:

 

I mean, they had Quinton Dunbar as their #2 ranked CB last year, so I'm inclined to think the fault lied with Norman rather than the coaches - especially considering how often he was benched.

 

I mean, Norman was flat out bad last year, he's lost a step and can't recover like he used to.  By the end of the season he had rookie UDFA's getting separation on him.  The point was not who was at fault, the point was that it's impossible to know for sure who was at fault...and then attach a grade to that.  It makes a stat misleading.

 

40 minutes ago, HTTRDynasty said:

 PFF can't know everything.  But I think it's foolish to completely dismiss an analysis just because they don't get something right 100% of the time.  

 

I'm not saying to completely dismiss, I'm saying it's an extremely subjective stat, that should not be viewed by itself as a sign of quality of play.  But rather, at the very least, be combined with other stats, preferably objective.

 

If a subjective stat says someone is good or bad, and an objective stat says otherwise...that should make you pause.  But if both a subjective and an objective stat say the same thing, you should feel more confident.  Darius Slay's 2019 season is an example of that.

 

40 minutes ago, HTTRDynasty said:

PFF was actually the first to recognize Josh Norman for his elite play well before he had made a name for himself among players/coaches/fans.  They go through an extensive grading process that has multiple redundancies, including former coaches who review their grades.  

You have gotta be on the PFF sales team.

 

 

[HTTRDynasty]

14 minutes ago, Alcoholic Zebra said:

The sample size is only 11 for that stat, and the difference in correlation is negligible.  Saying they're lesser/worse/inferior as a statement is weird.  Saying TD/Int or some basic counting stat is worse, sure.  But the others?

 

You're saying the difference between 0.54 and 0.26 is negligible?  Sorry, but I can't take that very seriously.  The others are not traditional stats and my focus was on how much better the grades were at telling the story than traditional stats.

 

17 minutes ago, Alcoholic Zebra said:

Serious question...do you work for PFF?  It sure seems like I'm talking to PFF sales here.

 

No.  I just put more faith in an organization who watches every play from every game throughout the season than I do random fans who watch maybe 15% of plays league-wide.  As I've said over and over, I do not agree with all of their conclusions, but I put more faith in their process and opinions than I do yours, for example.  No offense.

 

27 minutes ago, Alcoholic Zebra said:

I'm not talking about full season, I'm talking about his grades earlier that season with the Packers.  The grade below was just after we traded for him.  After the trade, someone went back to look at his Packers film, compared it to his Redskins film, and determined there must have been an issue with the PFF grades, as the film showed the same errors/issues with the Packers that he had with us.  Unfortunately I can't find that article right now.

 

 

 

Okay... so I have three choices.  

1. I can trust that PFF provided a decent approximation of HHCD's play during that time.

2. I can trust the random person who "went back to look at his Packers film, and determined there must have been an issue with the PFF grades"

3. I can go back and watch his games myself.

 

I don't have the time or the will to do #3, and I will choose #1 over #2 the other 99% of the time.  The 1% is reserved for the minimal amount of film analysts I have a considerable amount of respect for.

 

PFF has said countless times that grades for coverage players are extremely volatile from year-to-year and even within seasons, so it doesn't surprise me that players could have periods of highly graded play, only to come back down to earth.  Regardless, it's not as if an 83.5 grade is even considered an "elite" grade from them.

 

35 minutes ago, Alcoholic Zebra said:

It's not talking about his career as a whole, or what he'll do in 2020...it's just talking about what happened in 2019.  The point is PFF viewed him as having a below average year, yet others have responded by saying that can't be correct given his assignments, his technique, the difficulty of catches he forced WR's to make, etc.

 

48 minutes ago, Alcoholic Zebra said:

I'm not saying to completely dismiss, I'm saying it's an extremely subjective stat, that should not be viewed by itself as a sign of quality of play.  But rather, at the very least, be combined with other stats, preferably objective.

 

And I've never claimed PFF was infallible.  I've gotten into debates with PFF analysts on Twitter about not adjusting their grades for level of competition.  Obviously you should use more than just PFF grade to evaluate a player.  I have never claimed that you shouldn't.  But I do largely find myself agreeing with their rankings moreso than I do the majority of other sources out there.

 

54 minutes ago, Alcoholic Zebra said:

If a subjective stat says someone is good or bad, and an objective stat says otherwise...that should make you pause.  But if both a subjective and an objective stat say the same thing, you should feel more confident. 

 

Did you know the sky is blue?

 

52 minutes ago, Alcoholic Zebra said:

You have gotta be on the PFF sales team.

 

And you must create straw men as a hobby.

[Always A Commander Never A Captain]

45 minutes ago, HTTRDynasty said:

 

You're saying the difference between 0.54 and 0.26 is negligible?  Sorry, but I can't take that very seriously.  The others are not traditional stats and my focus was on how much better the grades were at telling the story than traditional stats.

Oh come on now, you know that isn't the stat we've been referencing.  It's the right column.  I'm quoting two points where I referenced what could have only been the right column. 

 

But the difference is 0.01, which is nothing. 

...

Right now all I'd say is that none of these stats are that good at predicting future success.

 

A difference of .2 or .3 from TD/Int to the others is sizeable.  You'd be able to see that from looking at the raw data.  But a difference of 0.01 or 0.03?  Nope.

 

45 minutes ago, HTTRDynasty said:

No.  I just put more faith in an organization who watches every play from every game throughout the season than I do random fans who watch maybe 15% of plays league-wide.  As I've said over and over, I do not agree with all of their conclusions, but I put more faith in their process and opinions than I do yours, for example.  No offense.

 

Sure, that's fair.  And the reasonable thing to do.  Everyone here might just post 1 gif, some numbers probably from PFR, and then their gut take on things.  From a credibility standpoint PFF has everyone here beat.

 

45 minutes ago, HTTRDynasty said:

Okay... so I have three choices.  

1. I can trust that PFF provided a decent approximation of HHCD's play during that time.

2. I can trust the random person who "went back to look at his Packers film, and determined there must have been an issue with the PFF grades"

3. I can go back and watch his games myself.

 

I don't have the time or the will to do #3, and I will choose #1 over #2 the other 99% of the time.  The 1% is reserved for the minimal amount of film analysts I have a considerable amount of respect for.

I can't even find the article anymore, so that's unfortunately meh without a source.  The basic gist was highlighting the poor angles and hesitation when reacting we saw of him with the Redskins, with gifs of him doing that earlier that season.  It's possible the writer cherry picked the only times he did that with the Packers that year, and every other pursuit of the ballcarrier was done well.  Yet given that we had just seen HaHa Clinton-Dix do poor plays against the ballcarrier immediately upon being traded, it didn't seem like a sudden and inexplicable regression in play quality was likely.

 

His QB Rating allowed when in coverage was similar enough with the Packers (86.3) before the trade and with the Redskins (88.2) after the trade, that it's probable his coverage grades were not worlds better with the Packers.  I also think I recall something (maybe from Mark Bullock?) showing that how HaHa handled his responsibilities in identifying and rotating to Deep Digs/Crossers and the like were the same for both teams.  

 

I'm also definitely not going to go out myself or expect others here to buy NFL League Pass (or whatever it's called) for $100 to backup claims we make.  That would be unreasonable for us amateurs/fans/whatever here on ES to expect.

 

So the question here is, do you think HaHa had errors in grading while with the Packers, or do you think HaHa suddenly was unable to take proper angles to the ballcarrier with us?

 

PFF is good with their stats.  But their grades, it's a point to make in a conversation.  Their grades should not be relied upon by itself.

 

 

 

 

[Skinsinparadise]

Ledyard who I think does good work on draft issues, ranks who he thinks are the best in the NFL per position

 

 

[HTTRDynasty]

29 minutes ago, Alcoholic Zebra said:

Their grades should not be relied upon by itself.

 

And I've never thought, or said, that they should be.

 

30 minutes ago, Alcoholic Zebra said:

Sure, that's fair.  And the reasonable thing to do.  Everyone here might just post 1 gif, some numbers probably from PFR, and then their gut take on things.  From a credibility standpoint PFF has everyone here beat.

 

This is basically what I've been saying.  Maybe I haven't expressed it as well as I should have.

 

But I guess I'm taking it a step further in saying I think PFF rankings are far more credible than this annual player's Top 100 list.  And I even found myself agreeing more with their rankings than I did ESPN's top 10 rankings from coaches and GM's that they recently published.  But that doesn't mean I think PFF grades are the end-all, be-all.  I use what I watch myself, what I hear/read from film analysts, what I hear/read from NFL coaches/FO, etc. in addition to PFF grades, to form my final opinion on players/teams.  But I do happen to think PFF is the best mainstream ranking system out right now, in the form of their grades.

 

Yes, we can nitpick and show examples where their grades are questionable, but by and large I've found their grading for Washington players to be pretty consistent with what I've seen from watching the games, so I have no real reason to doubt their grading league-wide isn't just as good.  In fact, I find myself agreeing with them more the more time I spend re-watching old games.  For example, I was not very high on SDH from watching the games live.  But after going back and re-watching every game from last season and focusing on SDH when he was playing, I now agree with the great coverage grade PFF gave him.  On the other hand, I also agree with the below average pass rush grade they gave Jonathan Allen last season (and there are objective metrics that back that up as well).

[Est.1974]

Minkah would have been great here. Shame the price was a first rounder. We get to see what the player he replaced is like instead......

[DWinzit]

8 hours ago, Skinsinparadise said:

Ledyard who I think does good work on draft issues, ranks who he thinks are the best in the NFL per position

 

 

Things that jumped out to me per this list:

Julio Jones #1 WR. For some reason he still seems underrated even though he is proven. His name never seems to be in the top WR discussions With the other top guys on the list.

Richie Incognito right after Zach Martin. This guy is a freak in multiple ways

Richard Sherman top CB, tell me how old and slow he is again. He would not be nearl this high on my list but he surely had a good year on a great team.

Love me some Minkah

3 hours ago, UK SKINS FAN 74 said:

Minkah would have been great here. Shame the price was a first rounder. We get to see what the player he replaced is like instead......

I crushed on Minkah and Roquan Smith in the 2018 draft! The fact that he was available last year, damnnnn

[Going Commando]

21 hours ago, HTTRDynasty said:

 

Perhaps.  It may be useless, or it may be ahead of its time.  I know PFF grades as a whole were clowned on for years before it became so widely accepted that every NFL team now has a subscription and players and teams use it in contract negotiations.  It may be the same with PFF WAR, which has already been studied and refined by those that coach, play and draft in the NFL.

 

Like I said though, I personally don't agree with all of their conclusions, but I do agree a great safety has more of an impact than a great 1-tech in today's league.

 

It's worse than useless, it leads to bad actionable analysis.  It's not like their grading system, which is a laborious and somewhat subjective method of stat counting.  PFF WAR is a derived stat.  Like I said before, the grading system is really only useful for making apples to apples comparisons of performance by guys playing the same position, or playing the same roles.  That way the systemic flaws in the way they assign grades apply the same.  It is not at all useful for comparing the value of performances of guys playing different positions/roles.  It has weighting errors in the way it counts stats that underrate the impact and value of line and run game play and overrate skill position play, as well as create highly volatile counting metrics for skill position play that hurt their accuracy in judging performance/quality.

 

PFF WAR doesn't account for these systemic weighting flaws, they're baked into the metric, and it leads to really awful analysis about the value of players in general.  Stuff that should be really obvious red flags for anyone coming to the conclusions their stats suggest, like putting Chase Young fourth or fifth in their mock drafts.  So PFF acknowledges that their DB, WR, and QB grades are highly volatile, but they don't seem to get that this indicates there is a pretty significant problem in the way they are grading these positions, particularly in comparison to the other ones.  Instead they assume the variability must be an accurate reflection of performance at those positions, and move on to derive conclusions that they are the most valuable and least replaceable positions in the game.  But what seems more likely to you?

 

1 - that QBs, WRs, and DBs can essentially forget their talent or make enormous, epiphany-like breakthroughs season to season, often in non-linear progressions, or that they suffer systemically greater rates of debilitating injuries

 

Or

 

2 - that PFFs grading system is flawed and really noisy for those positions in particular

 

I think you're also suggesting that PFF's work is more broadly accepted and valued by industry insiders than it probably is.  I have no doubt that agents use their work to try and get more money for their clients in contract negotiations when it suits that agenda.  And PFF cited discussion of their work on SNF as a basis of their credibility in that paper you linked (not mentioning that I'm pretty sure Chris Collinsworth works for them and was hired to promote their work).  But that doesn't mean that coaches and front office employees put a lot of stock in PFF's work when they're doing evaluations.

 

But let's get back to the original debate about the value of coverage vs line play and break down the value of a great DB vs a great DL, and in this case specifically, S vs 1 tech.  I agree with the premise that coverage is extremely important for defensive success.  But individual contribution in coverage is a cog in the machine type of contribution.  You can play perfect coverage and make virtually no impact on the outcome of the play because your pass defense still got shredded due to another guy's failure.  That is much less the case with line play.  If you win your match up as a DL, then you're likely going to impact the outcome of the play, either by wrecking that part of the pocket and altering the throw, or closing a run crease.  You can be much more of a routine gamewrecker as a DL in this way.  Thus I think it's much better for a team to have five or six good but not great DBs as opposed to one or two great ones and then three or four weak ones.  But I don't think that's the case with linemen.  Two great ones can wreck a game on their own even if the other two or three guys are nobodies.  I think you saw this reality play out with our secondary last season.  Dunnie was like the second best graded CB for PFF last season and he made virtually no impact on the success of our pass defense or wins.  But the best linemen much more routinely correlate to elite defensive performance. 

 

Another example relevant to the discussion of Derwin James's value--PFF has recently claimed that the Chargers have the best secondary in the NFL, presumably in terms of talent.  Their raw team pass defense stats were good last season too.  But they were 21st in pass defense and total defensive DVOA last season.  The individual talent they've stockpiled didn't translate to defensive success or wins last year.  Judging by their low rankings in a suite of team rush defense stats as well as team rushing offense stats, they had significant problems in their defensive front and even bigger problems with their run game.  Those failures superseded the impact of that secondary, and they lost a bunch of games.

 

There are just too many examples of winning and championship teams that were carried by their line play for me to agree that S > 1 tech.  Fletcher Cox and a dominant DL and OL group carrying a ho hum secondary and back up QB to a championship.  The Giants winning their championships wholly on the strength of line play.  Denver's great secondary play a function of a dominating front.  Carolina winning 17 games on the strength of their line and front play despite having anonymous receivers and having only one good DB and then Cam Newton falling off the face of the planet when the line deteriorates and Josh Norman looking mediocre playing behind a weak front.  The Rams winning 15 games based on their line and run game play despite sketchy secondary and WR/QB play.  The 49ers winning 15 based on their line play and their run game.  The dynamic plays out every level of the sport too, as Alabama has won 37 championships using the run game plus line play formula.  Even when you look at Troy Polamalu as the gold standard of elite all around playmaking at the safety position, it was James Harrison and Lamar Woodley and Casey Hampton and frickin Joey Porter being game wreckers, including getting the best of an all time great Seattle OL that got those Superbowl wins, not Troy Polamalu.  Hell Darnell Dockett almost carried that Arizona defense on his own in that game.  You can play dominant defense, get off the field, control the ball, and win games with a strong complimentary run game and a great DT as your best player and not a whole lot of quality lined up behind him.  I definitely don't think that's the case with a great safety as your best player and not much quality lined up in front of him.  They don't have the same level of impact.

 

Also, replaceability is a major reason why DLs are more valuable than DBs.  PFF claims that WRs and DBs are more valuable and much less replaceable than the players who play closer to the ball based on the delta of their performances vs the performances of the crap DBs and WRs.  But then they claim that the entire general value of the IOL and IDL position groups according to their metrics is propped up by the extreme outlier performances of the best players at those positions.  That doesn't really compute, as it suggests that the best IDLs and IOLs are extremely valuable and irreplaceable.  In the real world, I think it is obviously true that linemen are less replaceable than offensive skill players and DBs.  I think teams demonstrate a very clear pattern of making roster decisions based upon knowledge of this truth, and the reasons why it's true are straightforward and fundamental: size plus speed is much more rare than just speed, and OL and DL assignments are harder than secondary and WR assignments.  Their reads are harder and less forgiving of mistakes and it's much harder to block someone or beat blocks than it is to run free on patterns, fits, coverage assignments, etc.  And blocking/block beating around the ball is orders of magnitude more intense than blocking away from the ball.  Not to mention the contact they experience on a snap by snap basis is more brutal and being able to play through all that **** is harder.  Safeties have the easiest assignments of anyone on the D on pass plays.  Nine times out of 10 they're either single or help covering the slowest pattern on the offense or sitting in a deep zone with the most amount of time to read out of anyone on the D.  Even when they blitz they're generally coming free.  Their assignments are harder than corners' in the run game, but it's still mostly just running alleys and playing force against bad blockers.  It takes a player with superhuman instincts like Jamal Adams to make a big individual impact from that position.  And even when he's the best guy at the position in the league and is on a HoF track early in his career, he still got traded.  Could you see the Rams trading Aaron Donald after he made his first All Pro team?  The Eagles trading Fletcher Cox after he made his?  These guys are pretty much the least replaceable and most valuable non-quarterbacks in the NFL.