Monday notes — about PFF, Shawn Oakman, Eric Striker & more

Pro Football Focus received a lot of attention this summer when broadcaster Cris Collinsworth bought a stake in it. In many ways it further legitimised the product. We’re going to be hearing even more about PFF’s grades as a consequence. Bob Condotta at the Seattle Times dedicated a lengthy blog post to their breakdown of the defeat in San Diego.

I hope the extra attention and respect doesn’t stop people challenging some of their analysis. PFF’s system isn’t always water-tight, as the following tweet suggests:

Wilson’s two-touchdown, highly productive display against San Diego graded almost identically to Colin Kaepernick’s four-turnover meltdown against Chicago.

Here’s the issue as far as I can see it — PFF’s grades rely on one man’s opinion on a players responsibility during a given play and his ability to execute. As far as I’m aware they don’t use all-22 tape, it’s based on the broadcast output. They’re also judging based on their diagnosis of the scheme and play call. They’re grading a guard, for example, and saying he didn’t do his job because of X, Y or Z.

Vikings Head Coach Mike Zimmer was criticised by some for the following remarks…

I look at the grades and I can’t tell you what a 0.7 is or anything like that, but I know that the people that are grading our games and our defenses and our offenses, they don’t know if the tackle gets beat inside, if we weren’t sliding out to the nickel or who our guys are supposed to cover. I guarantee they don’t know who is in our blitz package and what they are supposed to do. I would just ask everybody to take that with a grain of salt, including our fans.

… and yet instead of appearing dated or out of touch, Zimmer simply points at the elephant in the room. An analyst sat at home watching the game on his TV or computer is grading every player based on what he interprets his role to be. Sometimes they’ll be right, sometimes they’ll be wrong. To take those grades on face value is to take a leap of faith in the individuals ability to break down every play call and scheme.

I’m not doubting their ability to get it right most of the time. However, it would only take one misjudgement on one scheme to potentially eschew a grade dramatically. If a player isn’t doing what they think he should be repeatedly, he’ll get marked down — possibly unfairly if he’s doing the exact job he’s been set by the coaching staff.

It’s also my understanding they don’t take into account the opponent. So a player competing against J.J. Watt is graded in exactly the same way as a player facing Benson Mayowa. Neither does it take into account supporting cast. J.R. Sweezy was marked down during a stretch where he was the only healthy starter remaining on the offensive line. He was criticised, to some extent, because he played on a hopeless line featuring Paul McQuistan and rookie Michael Bowie at tackle against the likes of Watt, Robert Quinn and Calais Campbell.

In the case of Kaepernick, whoever graded the Bears game clearly decided the quarterback wasn’t responsible for the four turnovers. They decided that his display was on a par with one of the top QB performances over the weekend.

It just doesn’t seem right.

I first became sceptical of the PFF system when I noticed they’d graded Earl Thomas as one of the worst starting safety’s in the league during the 2012 season. Perhaps he wasn’t at his Defensive MVP best like we saw in 2013, but one of the worst?

There is some science involved in the way they grade. It’d be wrong to suggest this is all based on a take. Yet there is a lot of one man’s opinion involved. The Collinsworth-inspired attention has legitimised PFF in a way it never previously experienced. I’m not sure that’s a good thing, because it seems like people are accepting their analysis as gospel. It isn’t. Use it, quote it, debate it. But take it for what it is.

For what it’s worth I’m not one of those anti-analytics types. Football Outsiders is a fantastic website and a weekly must-visit.

Shawn Oakman tape vs Buffalo

He stands out on a handful of plays here — especially the sack. He moves so well despite a reported 6-9, 280lbs frame. For that reason I’m almost certain he’ll be a high pick. There just aren’t many human beings with this level of size and freedom of movement. If Ziggy Ansah can go in the top-five based purely on potential, Oakman could easily be a very high pick next year.

Like Ansah, however, there’s a lot to work on. He has a tendency to get blocked out of plays too easily given his size. He can play with a greater intensity and his technique overall needs some refinement. Yet if he was a really polished, dominating defensive end he’d be a lock to go first overall. Nobody should be surprised he needs a bit of fine tuning.

A coach or GM somewhere will fall in love with the idea of developing this guy. The top-10 doesn’t seem unlikely.

Weekend scouting notes:

I watched the Tennessee-Oklahoma game and focused mainly on nose tackle Jordan Phillips. He’s a talented player who moves well for his size. He had a sack in the first half — stunting around and bursting to the QB. He could develop into a first or second round pick — but he wasn’t the most impressive player on the field in this game.

Eric Striker (6-0, 221lbs) is used mainly as a pass rusher but he’s really an athletic, roaming linebacker who just makes plays. He’s a junior and had three sacks in the Sugar Bowl last season versus Alabama. In this game he was constantly involved and stood out a mile.

With smaller, athletic linebackers getting more attention (see: Ryan Shazier) — Striker could be a coveted prospect. His ability to rush from the outside, cover and fly around will be attractive to many teams. If you get a chance to watch the Sooners this year take a look at #19.

It was also another impressive day for senior Chuka Ndulue. He looks bigger than 6-3 and 289lbs and while he might not end up being a high pick — he’ll add quality depth to a defensive line rotation. He’s a NFL player playing college football right now and it shows — he dominated the true freshmen on Tennessee’s O-line.

If we’re debating athletic linebackers destined to be high draft picks, Washington’s Shaq Thompson deserves a mention. He had two defensive touchdowns against Illinois and will almost certainly be a first round pick. The Huskies are getting a lot of love right now, but I’m not convinced Danny Shelton or Marcus Peters will go anywhere near round one. Thompson is a different case altogether.

Shelton is a big, productive tackle but is he enough of an athlete to warrant a high grade? Is he not the prototypical mid-round DT? As for Peters — his recent team suspension is a concern and while he showed what he’s capable of against Stanford last year, his performances are quite inconsistent. With good coaching he could be special, but it’s probably a risk too high for the first frame.

I mentioned it on Saturday but Markus Golden continues to dominate for Missouri. He and team mate Shane Ray have already notched nine sacks between them. They could both go in the first two rounds next year.

Kevin White’s fast start continues. The West Virginia receiver is a fantastic prospect — he had a big game against Alabama in the season opener and he’s already up to 460 yards (2nd in the nation overall) and two touchdowns. He’s 6-3 and 210lbs, sudden in his movement and he high points the football. He has suffered with a lack of confidence in the past but he’s showing no signs of that as a senior. He’s another player to keep an eye on this season.

Congrats to Austin Davis

Three years ago I had a chance to interview Davis — a highly motivated and talented individual at the time playing quarterback for Southern Miss. He went undrafted in 2012 and landed with the St. Louis Rams — and this week had his first NFL start against Tampa Bay. He completed 22/29 for 235 yards in a 19-17 upset victory on the road.

If the Rams want someone to control their offense and make the most of a difficult situation, I’ve no doubt Davis is up to the task. And while he’s now playing for a NFC West rival — I wanted to take the time to congratulate him on his first victory as a pro.

21 Comments

  1. Dan

    Shawn Oakman is just a tremendous athlete for his build. He also doesn’t seem to struggle with leverage as much as the similarly sized Morgan Hunt did. Do you think he could play inside or as a 3-4 DE, or is he strictly a 4-3 DE? Also you play him as a LEO in our 4-3 under or rotate him around like Bennett?

    Ray Drew of Georgia is a player that caught my eye against South Carolina and then I watched the videos of him available on draftbreakdown. He’s quick off the snap, powerful and gets good push on offensive lineman, something Bill Walsh said he valued in defensive tackles and maybe something the Seahawks are lacking now.

    • Rob Staton

      I’ll have a look at Drew now.

      On Oakman — for me he could be the type of player who kicks inside on passing downs and he can play the five-tech in a 3-4. For Seattle I think you move him all across the line.

  2. JeffC

    Are you sure that Peter Prisco doesn’t work at PFF?

  3. Mark

    Analyitics for football in general can’t be very accurate. It’s not like baseball where you have 162 games and all plays are essentially 1 on 1. I like FO as they concentrate on team statistics and will open admit that sample size is too small for meaningful statistics.

  4. Dumbquestions

    I saw those PFF ratings for Wilson and Kaepernick. Clearly they made no sense.

    I’ve been casting around for an analytic that measures RW’s value *as he’s used,* as opposed to how people want to grade him. So far, not much luck. His signature is efficiency, which I’ve been thinking of (crudely) as scoring possessions minus mistakes.

    I found one tiny little set of stats on PFF that appeared to examine those stats – but they’re far from perfect. One measure listed Seattle as second only to Denver in that context, which makes perfect sense.

    RW does not make mistakes. Seems to me any stat that attempts to evaluate him should include that factor.

    • AlaskaHawk

      I’m a big fan of the Seahawks but I have to disagree with the statement that RW does not make mistakes. He is good for at least one bone headed mistake per game. Usually it is either holding on to the ball too loosely and getting stripped in the backfield, or scrambling backwards until he is sacked 15 yards behind the line of scrimmage. I haven’t seen any improvement in that area this year.

      • dean

        Honestly every QB makes a mistake or two every game…you can’t expect Wilson to be perfect. I think he is in the top 3 right now in regards to that though. Can’t do much better than that.

      • Jon

        are you referring to this past Sunday when the entire D-line was in the backfield within 1 second of the snap and he attempted an escape? I don’t think that is a mistake, I think it is how RW end up making many of his amazing plays after being let down by the O-line. I am not saying he does not make mistakes but this year he has taken 3 sacks, 1 of those was for 0 yds. Thrown O INT and has given O fumbles to the other team. Kaepernick turned the ball over 4 times against the Bears and scored just about the same as RW. Look at the context when someone makes a comment like RW doesn’t make mistakes.

      • Dumbquestions

        I’ll accept the idea that RW makes some mistakes, i.e., trying too hard to extend plays at times and getting sacked – but some mistakes matter more than others. He doesn’t give the ball away, and that’s the important thing. If RW throws 0 interceptions and Kaepernick throws 4, that’s a substantial difference. PFF’s metrics don’t appear to reflect that disparity, and without some explanation of the rationale, it’s pretty hard to give those metrics a lot of credence.

        • Chris

          RW has had a couple picks dropped.

      • AlaskaHawk

        How about RW hasn’t had any big turnovers this year. That is a fair statement.

  5. andy

    Regarding to PFF, personally, I think they are pretty good and consistent on their “grading”. We, as fans, tend to “cheer” when it favors your players then “boo” when it does not. The fact of the matter is RW, according to their grading, is having one of the highest scores, in the first two years for ALL qbs (I believe he has like a 20+ cumulative score over the two years). I don’t hear any complaints regarding to that score.

    • Rob Staton

      OK then, let’s take the names ‘Wilson’ and ‘Kaepernick’ out of the debate and call them ‘Bert’ and ‘Ernie’ instead.

      Bert doesn’t turn the ball over, has two touchdowns on the road and is pretty much the only reason his team has a chance to win on the day.

      Ernie has four turnovers in a home loss.

      PFF graded both players almost identically.

      This isn’t a Seahawks fan complaining about a snotty review. It’s a critique on a system of grading that in many ways just doesn’t make any sense. It’s just a coincidence that on this occasion Wilson and Kaepernick’s two performances are the example used to illustrate the point.

  6. dave crockett

    The problem I have with PFF is that without knowing the methodology the grade is pretty meaningless. PFF would actually be more useful if they published their raters’ notes rather than the numerical grade. In other words, it would almost certainly be more insightful to understand the raters’ thinking about the QBs in the SF/CHI & SEA/SD games than the final number.

  7. CestrianHawk

    Apologies if this doesn’t seem to be the place to ask. In the SD game last Sunday and concerning Harvin’s ‘out of bounds’ touchdown, is the line considered out of bounds? I ask because in soccer the line is ‘in bounds’. Harvin’s foot was clearly on the line but no part seemed outside the line.

    • Rob Staton

      On the line is out of bounds in the NFL. If any part of his foot is touching the line, the play is over.

  8. CestrianHawk

    Thanks Rob.

  9. Radman

    PFF’s methods are very similar to the grades in many advanced metrics used in baseball defensive metrics. One key aspect of those measures is to not dwell on the micro to take a macro approach. A player can have wide variance from game to game and season to season. But, in two or three season samplings, you can get a pretty good idea of what that player is.

    Fangraphs and other places often to fan ratings, and they very often end up being very close to what some of the UZR ratings are. This is a good test- the old “eye test”. So, while it’s true that a single game performance might not align with what one sees, at the end of the year, or a two season sampling…more often than not, the scores at PFF very closely mirror what most intelligent observers “See”. There are some exceptions, and they are usually supported by a pretty sensible narrative.

    So, my advice with PFF or any other metric or scouting, is to not take it as “gospel” and also not expect a single sample, measure, or observation to tell the whole story. But, over the course of time, the repeated, systematic observations are a good data point worthy of note.

    I don’t think anyone should take any scouting agency, blog, or write as the single source of truth. But, if you take in a few you can get a fair enough picture. PFF is unique in that, if you take a season long or two or three season sample, you’re getting several hundred systematic observations coded over time. No one else does that. Everything else is very impressionistic.

    Their effort to systematize should be commended. And utilized. But not relied upon solely, nor disregarded because a few games or even a single season of a single player doesn’t make sense to you.

    • Rob Staton

      Whether you get a good result over 2-3 seasons or not, the point is PFF still publish grades on a game-per-game basis. Their system is open to debate over the long haul anyway (see: Earl Thomas’ 2012 mark), but while ever they’re saying, “this player played a good/bad game and here’s the grade to prove it”, it still needs to be judged as one man’s opinion on a player’s performance, all while trying to project what his specific role is on an individual call or within a scheme.

      If we ignore the occasions where the grades seem strange (positive or negative) then we might as well see them as the gospel. Every grade influences the overall mark for the season. If we’re saying that he didn’t deserve a 0.8 this week but it’ll work out as an accurate grade down the road, we’re basically admitting he’ll be harshly judged in a future game to even it out.

      I’m not totally disregarding PFF in the same way nobody has claimed you need to listen to one scouting agency, blog or writer for the ultimate source. I wanted to write a piece to break down what PFF actually is in it’s rawest sense. One person watching a TV broadcast judging a player’s performance, without taking opponent or circumstance into account, while trying to estimate his role within a specific scheme or gameplan.

      • Arias

        Going from memory I do recall that ET3 did grade out really poorly in 2012 because he was a tackle missing machine. I seem to remember an article by them ripping him to pieces for it. But that was the only noted flaw too, so unless they penalize disproportionately against missed tackles it’s hard to understand how he graded out as one of the worst simply from his mountain of misses in ’12.

        • Radman

          I find the fixation on ET3 one year rating here hilarious. If you look at how they’ve rated other Seahawks players throughout their career, they’ve been well ahead of the curve of other scouts and media outlets in terms of rating their players.

          ET was not “One of the worst” in ’12. As has been posted before, there was a very tight cluster of safeties rated very closely together that season, and Earl was near the bottom. But the spread was very thin from about 8-20. Just a few plays.

          Earl corrected his self admitted tackling issues and the very next season he bounced right up to the top of the rankings. I think the Earl Thomas example actually is a credit to the PFF system, not a criticism.

          Like WAR in baseball stats, unless you look at the numbers in context of the others, and from one year to the next, the numbers are not that valuable. The story they help tell over time are more valuable.

© 2024 Seahawks Draft Blog

Theme by Anders NorenUp ↑