2013 Offensive Line Rankings

| January 13, 2014

OL-Rankings-2013-finalAwards week is still ongoing and so what better time than to announce our 2013 Offensive Line Rankings. That’s right those big guys up front who don’t post positive stats with their reputation often a product of where they play or where they were drafted are now getting the attention they deserve.

For some that’s good and for that is most definitely bad. One thing not in question though is that each player was judged by the same standards as his peers from our tireless tape watchers to give the most comprehensive list out there.

Numbers in bracket indicate last year’s rank while “PB” equals their pass blocking rank, “RB” their run and screen blocking rank & “PEN” their procedural or disciplinary penalty rank.

Updated: 20/03/2014

A mea culpa here. Some erroneous adding up of grades left the run blocking rankings out of whack and obviously had an impact on the total grades. This has now been corrected and double checked.

32. Arizona Cardinals (32nd)
PB: 32nd, RB: 27th, PEN: 22nd

Stud: Our 18th-ranked center on the year Lyle Sendlein (+0.5). The former undrafted free agent is the definition of a solid player, rarely wowing you in either a good or bad way. The team would have hoped Jonathan Cooper would be this guy, but that will have to wait for a year.

Dud: There was some serious competition here but Bradley Sowell (-38.4) takes the cake. In 12 games he earned 11 negative grades with 10 of them -2.2 or worse. The NFC West did feast on him.

Analysis: It’s rare to see a team finish last two years on the trot, but then it’s nothing Cardinals fans wouldn’t have seen coming. It wasn’t helped by the aforementioned loss of Cooper, while incredibly the team’s long overdue divorce from Levi Brown actually left them in a worse position at left tackle. It’s hard to imagine them being worse next year.

31. Jacksonville Jaguars (29th)

PB: 23rd, RB: 32nd, PEN: 29th

Stud: Hmm. We’re not allowed to pass so Uche Nwaneri (-0.3) gets that nod here. Probably best we move along from this quickly.

Dud: He’s not getting any better is he? Will Rackley (-32.7) is a disaster waiting to happen and in a division that features J.J. Watt that’s just asking for trouble.

Analysis: They traded away their star left tackle, watched as their rookie first round pick went on Injured Reserve and at the season’s end said goodbye to their long-term center. Performance was perhaps secondary to the coaches getting a chance to evaluate which of the 11 lineman they used they wanted to keep around. Unfortunately, the talent level just isn’t there right now.

30. Atlanta Falcons (15th)
PB: 30th, RB: 26th, PEN: 18th

Stud: I hesitate in calling Justin Blalock (+6.8) a stud because he’s symptomatic of some of the wrong moves this franchise has made in years gone by. Namely, treating a good player like a great one when it comes to contract negotiations. Still you can get by with five Blalock-level talents on your line. You can’t cope when he’s the best of the bunch.

Dud: The team didn’t envisage Lamarr Holmes (-32.3) starting on the left side and before Mike Johnson’s injury probably didn’t see him starting at all. As it was, he had a baptism of fire giving up the second highest number of quarterback disruptions of any tackle.

Analysis: They overpaid Sam Baker who promptly got injured, cut the reliable Tyson Clabo and had to deal with Todd McClure retiring. It was likely to be a rough transition but nobody saw this coming. The team had to alter their passing attack (screens and quick get outs for Matt Ryan) because whoever they fielded at tackle couldn’t defend the outside rush, and they allowed far too much penetration up the middle. Gave up more pressure than any other line. Without serious personnel questions answered it’s hard to see how this isn’t a disaster next year.

29. Oakland Raiders (24th)
PB: 21st, RB: 31st, PEN: 32nd

Stud: With Jared Veldheer missing most of the year it was left to Stefen Wisniewski (+10.4) to lead this line. He was easily the most consistent player on the line and looks set to lock down the center spot for years to come.

Dud: Starting Lucas Nix (-44.3) was a disaster. He was nowhere near ready for what was put on his plate and but for a midseason benching would have set records for low grades that you’d think would be hard to surpass in the years to come.

Analysis: Given the cap situation and injuries problems that befell this line, 27th might be considered a minor victory. They brought in players like Tony Pashos and Matt McCants and got surprisingly decent play out of both of them as rookie Menelik Watson found himself limited to 177 snaps that were not hugely encouraging. If they can re-sign Veldheer and get him healthy then there are pieces here to make a slow rise up these rankings.

28. New York Giants (11th)
PB: 31st, RB: 20th, PEN: 4th

Stud: In good news for the G-Men, rookie Justin Pugh (+7.1) got better and better as the season went on. Derided by many as something of a reach and with questions as to whether he can hold up at tackle, he was the biggest success story in a bad year for the Giants.

Dud: He’s been kicked around every position but center in his time as a Giant, but David Diehl (-26.5) really didn’t take to life at right guard. He earned this grade despite missing five games and at this point just can’t handle better athletes lining up opposite him.

Analysis: Injuries didn’t help, with the loss of David Bass seeing Kevin Boothe move to center where he would struggle. That created all sorts of shuffling and was further compounded by Chris Snee having his season ended after 188 snaps. Throw in Will Beatty responding to getting paid with a huge drop off from his 2012 season and you have the perfect storm. A once proud unit is now a major question mark.

27. Seattle Seahawks (20th)
PB: 25th, RB: 23rd, PEN: 30th

Stud: With injuries depleting the ranks, it was left to Michael Bowie (+7.1) to lead the team with their highest grade. He may eventually end up at guard (as he was for their recent playoff victory over the Saints) with his run blocking particularly impressive.

Dud: The team has to hope they never, ever have to start Paul McQuistan (-24.8) at left tackle again. It went very badly and he wasn’t much better at guard.

Analysis: An interesting year. Losing Russell Okung hurt but when they did get him on the field his play was a level or three below it’s usual high standard. At center Max Unger had a down year as a variety of combinations on either side of him failed. Essentially, they did enough at times for Marshawn Lynch to make yardage, but this had the feel of an experimental group with the coaches trying to luck into the right combination.

26. New York Jets (3rd)
PB: 15th, RB: 30th, PEN: 25th

OL-rank-fergusonStud: For a line with as much talent as they have, it’s really surprising that D’Brickashaw Ferguson (+0.1) would walk away with the highest grade. A bad year by his standards where he gave up eight sacks and was poor with his run blocking.

Dud: Whoever they put at left guard. Vlad Ducasse (-9.5) had a breakout game against the Patriots and then proceeded to stink the joint up, forcing rookie Brian Winters (-28.5) into action. It did not go well.

Analysis: The awe factor of watching Nick Mangold has gone. He finished the year strong but his streaky early season play was anything but what we’ve come to expect from a usual contender for first team All-Pro center duty. He epitomized why this line took such a nosedive. Losing Damien Woody was the first step in the team’s gradual decline and it seems saying goodbye to Brandon Moore may have really accelerated that process.

25. Indianapolis Colts (31st)

PB: 28th, RB: 22nd, PEN: 13th

Stud: They spent big on Gosder Cherilus (+12.2) and while the grade is good perhaps they’d want a little more from him? Particularly in the run game where the team struggled all too frequently.

Dud: Will Mike McGlynn (-23.4) be starting next year? After reviewing the 2012 and 2013 tape it’s hard to see how the team allows that to happen without his game reaching a whole new level that we haven’t seen.

Analysis: It didn’t go well in losing Donald Thomas early on, leaving them starting a rookie guard. But for all the investment here more is needed to help Andrew Luck and whoever is running the ball. The tackles are good enough (though it wouldn’t hurt for Anthony Castonzo to give up a little less pressure), but the interior need fixing.

Turn the page for the next eight on the list…

 

Pages: 1 2 3 4

  • bz2

    Interesting to see Philadelphia ending only at 9, after some of the praise we heard from PFF for them, plus having Kelce, Mathis and no-one in the red. I guess the praise they got was all predicated on the run. Also, in the Philly paragraph, “you’re” should be “your”.

  • Lord Mad

    Patriots run blocking at 21st? Right….no way is the pass blocking better than the run blocking.

    • Bellini

      Could the team at PFF please sort out their comments system please? It must be so frustrating to have your posts cut midway through. Rather than the coherent rationale he was about to give for his view, poor old Lord Mad is now left looking like a petulant child.

      Sort it out!

      • Mr Garrison

        @Bellini Gotta love those internet intellectuals

      • Lord Mad

        Had to make room for your pretentious self absorbed ego.

  • Luke

    How can the Cardinals rank 32nd in the rankings? I feel this is more a predication based on last year’s line. To rank the Cardinals the same as last year is deeply misinformed. I would advise people to look at http://www.footballoutsiders.com/stats/ol.

    • pbskids4000

      The fact that FO think the Colts have the 15th best run blocking line makes it lose so much credibility. I’d rather stick with the website that shows performance of players. You can not give all credit to a offensive line if the RB gets a 5 yard gain or if a QB/RB got a first down on 3rd and 2. Stuid way to analyze.

      • Abdallah Awwad

        lol and given the fact that atleast 3 teams have given up 40 plus sacks, one of which gave up 50+ are in the top 15 also helps prove pff is a joke. the colts can’t block at all.

  • MT

    A Redskin fan for 21 years and this is one of the worst offensive line I’ve ever seen. I’m dead serious. Nobody can play QB behind that line. OMG. It’s beyond belief to rank them at 4th, even if they were ranked at 20, it would still be generous.

    • yankydave

      Exactly. Watch when at least four of the guys-not-named-Williams are released and don’t start anywhere else next year. That will prove how ridiculous these rankings are.

  • Bill Marcellino

    Football Outsiders has NE ranked 1st in offensive line RB; PFF has them 21st in RB. This is because of a difference in the unit of analysis, and the analytical assumptions behind that unit of analysis. FO is looking at the line as a whole as the unit of analysis, and PFF the individual player. You end up with two very different methods based on emergent vs. reductive assumptions: a) line play is an emergent quality b) line play is reducible to individual performance. If you think line play is an emergent quality, you can’t simply follow PFF’s reductive approach of grading players and then simply aggregating the grades.

    This is a good example of how PFF’s method is likely more appropriate for more individual performances (e.g. defensive backs), but fails when applied to more complex emergences (e.g. line play).

    • Lord Mad

      Awesome breakdown. Thank you!

    • Mike Renner

      There are so many variables that aren’t accounted for when crunching running numbers that it is impossible to equate their arbitrary definitions of running success with offensive line play. Things like running back talent, quality of other blockers(full backs/tight ends/receivers), and play calling are all huge factors to a running plays success. That analysis doesn’t really consider those elements though and puts the result of a running play solely on the offensive line.

      While you may feel we rank New England too low, that analysis also has two of the top three rushing offenses, Philadelphia and San Francisco, ranked 26th and 29th respectively. It wouldn’t take watching that many games of tape to realize that both those units can run block better than the Falcons and Cardinals.

      • Bill Marcellino

        “Arbitrary definitions of running success?” That makes no sense–for example, there’s nothing arbitrary about assigning “tackles at or behind the line of scrimmage” as a line failure.

        • Mike Renner

          So on a toss play when a safety beats a WR and tackles the RB for a loss…that’s the lines fault? Or if a draw is called against an overload blitz, 5 linemen are supposed to pick up 6 blitzers? Yes, usually runs for a loss come from line failures, but attributing blame without watching the play is guesswork. From the article:

          Losses: 120% value
          0-4 Yards: 100% value
          5-10 Yards: 50% value
          11+ Yards: 0% value

          Saying the offensive line has nothing to do with a gain over 10 yards, but takes all the blame for losses seems arbitrary to me.

          • Bill Marcellino

            Mike, I think you’re right that this doesn’t account for WR blocking. But I would point out that it is a proxy, not a measure–while you are right to point out that there are cases where stuffs aren’t attributable to the line, I do think MOST of the time, stuffs are the result of bad line play. Just like MOST of the credit for very long runs likely goes to the runner. Maybe this could be refined, but clearly it is not arbitrary.

          • Mike Renner

            I agree that line play isn’t just a sum of individual talents and understand why they chose those figures.What I’m saying is that their weighting system is arbitrary. 120%, 100%, 50% and 0% are arbitrary designations based on what they believe is the best way to eliminate running back skill. I just think you’ll never be able to control external variables well enough by simply using yards as a basis.

          • Bill Marcellino

            Mike, these aren’t arbitrary–they’re approximations based on judgement. If I tell you, “Top Marine officers usually have 1 or 2 bad fitreps in their record book,” that numerical range isn’t arbitrary. It’s an approximation, and it represents judgements. It could be a pretty crummy judgement (maybe it’s based on unrepresentative experience), or it could be a good judgement (based on a retroactive study of officer performance and retention).

            AYL is a proxy developed from observing football in the real world, and an understanding of how running plays actually develop. It’s approximate of course–maybe losses should be weighted more or less, or maybe distances could be shortened/lengthened. So there is a range of acceptable numbers, but you couldn’t plug just any number in. I couldn’t put 0-90 yard runs are fully weighted, but 91-99 yards are slightly weighted in favor of running back performance, at least I couldn’t if I wanted to match actual observation of football.

            I have no problem with you critiquing this as a proxy, or the construct of the proxy–I think it’s reasonable, but I don’t know that it is is perfect. But saying “it’s arbitrary” indicates an analytical error.

          • Mike Renner

            I guess I keep using the word ‘arbitrary’ because any system that you could create based on yardage is meaningless without actually watching the play. Is there anywhere on FO that explains how they came up with these values? I’d be curious to see if their reasoning arose from watching and charting a bunch of plays and seeing for what gains line play has the biggest impact.

          • Ajit

            Coming from someone who likes both products, its worth stating that pff’s analysis is focused on the microlevel details of football, while FO is about the macro. Fo’s basis is all about trying to sort through relevant variables for optimal predictions – an excercise entirely different from what Pff is trying to do. Both are useful.

            I think this deviated too far from Bill Marcellino’s original point, which was, trying to explain the gap between what pff’s findings and FOs. I wouldn’t say Pff failed anymore than I’d say FO succeeded. They may just be telling us two different sides of the elephant.

            As an aside – I think it would be fun to try and see what kind of bridge can be built to connect Fo’s aggregate statistics with Pff’s micro level statistics. Ultimately, we’re dealing with event analysis and they should be related(assuming we’re all watching the same games).

          • Bill Marcellino

            Ajit, I think you’re quite right that the source of the difference is a micro vs macro approach to data. It would seem in cases like this, where there is a really huge divergence, that one or the other approach is missing something.

          • Jef Rogers

            Guys, just want to let both of you know that I appreciate the interesting dialog presented in a constructive way. This stuff usually devolves into a catfight in two posts and I learned something from reading this entire thread. Namely, that the B’s I earned in statistics were likely lucky grades.

          • Bill Marcellino

            I’m surprised constructive dialogue hasn’t torn a fabric in the internet…

          • JJ

            Which brings us back to the days before PFF and fancy oline metrics, when X oline was good and Y oline sucks.

          • Ben August

            Hmm, this is all very interesting. I’d suggest evaluating it through Data Envelopment Analysis (DEA), using an Output Perspective (can choose multiple input and output vectors). I think DEA’s ability to define a best-practice frontier, combined with ad ho/quasifactor (post-analysis externality analysis), would be very useful (ie, quasi factors could at least take in to account, quality of competition). Something to look in to, if you enjoy stats and sports- highly applicable.

          • Ajit

            In statistical parlance, the problem here is with endogeneity, or more generally, separating cause and effect. There are all kinds of linkage issues in football, ie- how much is play calling /receivers/ qb affecting how good the o line is playing.

            I think PFF still does a great job, but for instance, can we really know for sure how good den’s o line would fair with tebow at qb? In theory, pff’s grades are attempting to separate that, but even they would admit that such extremes would definitely affect their ratings. Again, it’s still very informative and they should be applauded, but trying to tease out those factors statistically on a microlevel is extraordinarily difficult.

          • Ben August

            I would agree that it is difficult- and in fact, may be utterly impossible.

            As far as addressing externalities, I typically suggest running the primary analysis, and then applying ad hoc analysis later, in order to attempt to explain or highlight how different externalities may impact whatever is being observed.

            For instance, though we don’t know how different QB’s can impact a line, we can try…IE, the biggest quantitative factor would probably be “time to decide” and/or “time to release,” regarding how fast a QB typically gets “rid” of the ball.

            Anyways, you run the initial analysis, and then re-order based upon a classification (measured time), then plot it across a graph, and look for trends.

            I’ve had what I believe to be great success using “ad hoc” analysis to account for externalities, and have applied it to factors such as coaching style, weather, elevation, play choice, etc…with regard to the output of various positions within football (NFL).

            But like you said- it is difficult, and time consuming, and ultimately it is hard to know the direction of the linkages (SPSS and PASW would probably help, but didn’t take that far).

          • Ajit

            I don’t know this for sure, but they did mention they used regression analysis to determine which values produced the most predictive statistic. It certainly seems reasonable I would think.

          • Bill Marcellino

            Mike,

            1) You’re quite right these aggregate statistics tell us nothing about the individual play. That’s the nature of this kind of proxy. It’s like BMI–I’m just under 30 BMI (obese), but that’s inaccurate–in this micro-level case my BMI reflects strength training. But at the macro-level of populations, BMI is a very reliable proxy.
            2) I think your question about the method is valid–I would like to hear how they got to those figures as well. They seem reasonable at face value to me, but I’m no scout. I would prefer an explanation that makes clear why weighting 0% for long runs is more accurate than weighting 25%.

          • Mike Renner

            It’s interesting that you bring up BMI because it suffers from the same input data problem that I think FO’s analysis has. Height/Weight may come close to representative on a Macro-level, but wouldn’t you say body fat % would be a more precise depiction on both micro and macro levels? BMI doesn’t regulate a huge variable in body composition, muscle vs fat.

            FO’s analysis has the same problem in that the input data doesn’t control for RB skill/running style. They are using yards, but wouldn’t yards prior to contact be more accurate again on both micro and macro?

            Some of FO’s stats are great in that they approach complex problems with complex equations that take into account loads of aggregate data. This formula, along with their sack rate formula, just seems too simplistic to tackle something as multi-faceted as line play. I think they are headed in the right direction with their analysis, but they simply need better input data.

          • Bill Marcellino

            Mike,

            1) Yes, body composition would be more accurate at the level of populations. But getting that data is very costly, so BMI is used as a proxy–it is cheap to get that data, and at the level of populations it is a very reliable proxy. Communities with high BMI are chock-filled with overfat people.

            2) I think adding in yards after contact, or using broken tackles, misses, etc as a way to weight aggregate yards would be more accurate. My question is about whether the increase in accuracy would justify the cost.

    • Budahmon

      The grading is also subjective. If the grader likes the player they will get a better grade….that is just human nature after all the graders are fans and have their own personal biases. Case in point lets look at the Chiefs (and I to have my own personal bias); PFF has liked the Chiefs’ Asamoah for several years. They have also slammed Fisher on a continuous basis all year. Yet, the Chiefs line play improved dramatically once Asamoah was injured. Specifically, Fisher’s run blocking and pass protection greatly improved with Schwartz to his left. Yet, PFF grades out Asamoah highly and takes the time to slam Fisher again in this write up. Here I will quote from the article….(WRT Schwartz) “his play ensured that the team couldn’t switch back to Jon Asamoah (who wasn’t playing badly himself).” Asamoah was the worst OL the Chiefs had this year. His play affected the entire line. When he was injured and Schwartz replaced him the entire line play took off. Charles had more productive runs, getting into the second level of defenders on a more consistent basis and the OL pass protection improved. Asamoah has always been inconsistent from game to game, but this past year he was consistently bad…yet PFF ranks him highly. They had the same feelings about Lilja last year…..and where is Lilja now? Out of football since he was cut by the Broncos…

    • Daniel

      Football Outsiders just broke down Cam Newton last week and said he was a emerging talent and a good passer. They raved about how he is reading the field. This is a guy who is completing a Rick Mirer like 70+% to the left side of the field and 55% to the right side of the field. His TD/INT ratio is 8-1 throwing left and 13-10 throwing right. He is rarely even attempting passes over the middle of the field. He averaged 3 pass attempts a game attempted over the middle and had a 3-2 TD/INT ratio there. You know who can’t throw over the middle of the field? Poor passers! His completion % drops to 42% when he throws 11 yards or more. Cam’s yardage dropped to 3,379 this season. Also if you break it down Cam has ALWAYS put his stats up vs. weak defenses and struggled vs. good defenses. Check the game logs over the course of his career in you do not believe me. If you watch the Panthers, like in the SF game last week, when they give Cam the keys to the offense they struggle to move the ball and he throws picks. When they ask him to do less and run the ball more they play better. None of this is a breakdown of a top level QB or an emerging talent who is a good passer. The fact is the Panthers have an elite defense that carried them to their awesome record and a playoff birth. Cam is hands down the best scrambling QB in the league, but he is suspect as a passer when asked to do so. Yet, Football Outsiders somehow finds him to be a good passer in the pocket and emerging talent.

      • Nik

        Cam’s passing yardage going down is due to the Panthers switching to Mike Shula’s conservative run first ball control offense. With Chudzinski, they had an aggressive downfield passing attack. Shula’s conservative playcalling has restricted their offense.

        And have you seen Cam’s weapons. His #1 WR fell off a cliff. He’s 35 years old and lost his speed, leaping ability, and had uncharacteristic drops. His other options aren’t reliable either. Brandon Lafail and Ted freakin’ Ginn.

        Gimme a freakin’ break. The best QB in the league last year Peyton Manning underperformed severely against the Seahawks defense. So, you’re saying QBs do worse against good defenses? No way!

    • sam

      True, but Football Outsiders uses formulas involving just yards, and a good running back will create more yardage for his team. That’s why the chiefs are so high with Jamaal Charles. It’s definitely more important to look at the line as a whole, but PFF, despite the subjectivity of each rating of each play, doesn’t reward a lineman for a great move made by the running back. It’s best to look at both rankings and take them with a grain of salt.

  • phillydude123

    Wait, what? How is Miami #19 when they lead the league in sacks allowed… by a long shot? I love PFF, but you guys need to do some work on analyzing units as units, not individual sums.

    • pbskids4000

      WTF?! If you have ever gone to this site before you would realize how little sacks mean in the grand scheme of things.

      • Trey Warner

        Dude if you actually watch one snap of a Dolphins game you will see how horrendous their O-line was

    • Jonas Salk

      Sacks are not necessarily the offensive lines fault. There can be a bunch of things that can occur that lead to a sack and none of them are any indication of how the offensive line played on that particular play. For example, a RB, FB or TE can miss a block or a blitzer. You also have QB’s who simply have a slower release or just take longer to read a defense and get rid of the ball.

  • yankydave

    Redskins at three completely destroys whatever formula you use. What a joke.

  • Geo McDowell

    “Arizona Cardinals (32nd, again)”
    This is what happens when you draft a Notre Dame WR when you have a HOFer at WR and desperately need help on the OLine. Would the 49Whiners even be in the playoffs if the Cards had drafted one of the many quality tackles still on the board in 2012?

  • Daniel

    I can’t agree with the Eagles line not being at least top 5. Their worst guys graded out positive. I didn’t see anyone else with a grade like that. Not to mention, as you stated, they had 3 all pro candidates and in my opinion the best lineman in the NFL (Mathis). However, I liked the article good read, and thanks for giving us some awesome analysis all season!

  • Stating The Obvious.

    “you can get by with five Blalock-level talents on your line. You can’t cope when he’s the best of the bunch.” So there’s five Blalock level talents on the line and you can’t cope when he’s the best but you can get by with 5 of him and therefore he’d be the best offensive lineman by proxy.

  • Stating The Obvious.

    Love all the stats I get from this site though. You guys are a huge research when I need to research certain players.

  • Abdallah Awwad

    see this is the problem with pro football focus. you guys count a qb hurry as 3/4ths of a sack. that is not accurate at all. the sacks are the worst thing allowed, followed by the qb hit. big ben was sacked 41 times. the green bay qbs were sacked 52 times. yet the bears recieve lower grades then them. i think it should be hurry lvl 1 =1, hurry lvl2 =2, qb hit =3, and a sack =4. it should be like era in baseball, the lower the better.

  • Don417

    Brandon Albert a star? On what planet?