Writers of Pro Football Prospectus 2008

Most Recent FO Features

KerriganRya11.jpg

» Week 2 DVOA Ratings

Stomping the Jags leaves Washington No. 2 behind only Denver. But what can we really learn from one big win early in the season, before we are applying opponent adjustments?

30 Jul 2013

2012 Pressure Plays, Offense

by Danny Tuccitto

We're back for the second installment in our two-part series on pass pressure in 2012. Last week, we presented stats for defenses; today we'll do the same for quarterbacks. All of these stats appear in Football Outsiders Almanac 2013, which is still on sale if August is when you re-engage with the sport of football (PDF here, paperback here).

We only have pressure stats split out in this particular way for the last three seasons, because of changes in how we've done our game charting. Still, there are certain things we know for a fact about quarterbacks and pressure. First, every last one of them is worse when they encounter it. In 2012, Philip Rivers ranked last with a DVOA under pressure that was over 200 percentage points worse than his rating without pressure. Second, only the absolute worst of the worst quarterbacks can't muster a positive DVOA even on plays with no pressure. That group last season consisted of Ryan Lindley, John Skelton, and Brady Quinn. Finally, no matter what defenses seem to do, the following quartet finishes among the least-pressured quarterbacks when healthy: Peyton Manning, Tom Brady, Drew Brees, and Eli Manning. (Only the first two actually appear immune to pressure.)

Like last week, today's table is sorted from highest to lowest in "percent pressure" and the "Difference" column is ranked from smallest to largest in terms of pressure drop. Unlike last week, today's table contains quarterbacks, not defenses -- more specifically, quarterbacks with at least 100 pass plays in 2012.

Qualifying QB Plays Pct
Pressure
Rk with Pass Pressure without Pass Pressure Difference
Yds DVOA Rk Yds DVOA Rk Yds DVOA Rk
M.Vick 419 30.8% 1 2.5 -107.9% 20 7.7 59.7% 8 -5.3 -167.6% 34
K.Kolb 219 27.9% 2 1.5 -130.7% 32 6.6 27.0% 25 -5.1 -157.7% 33
R.Wilson 488 27.5% 3 3.7 -48.7% 2 8.2 71.4% 2 -4.5 -120.1% 12
A.Luck 710 26.3% 4 3.8 -51.9% 4 7.4 28.3% 23 -3.6 -80.2% 1
C.Ponder 563 25.0% 5 2.1 -85.5% 16 6.8 35.6% 19 -4.7 -121.1% 13
C.Kaepernick 260 24.6% 6 3.4 -66.1% 9 8.8 71.4% 3 -5.4 -137.5% 19
C.Newton 561 24.2% 7 3.9 -60.7% 7 8.1 46.2% 14 -4.1 -106.9% 7
R.Griffin 473 23.7% 8 3.7 -68.8% 11 8.5 66.7% 5 -4.8 -135.5% 18
B.Quinn 233 23.6% 9 0.1 -144.4% 34 6.3 -6.7% 38 -6.2 -137.8% 20
N.Foles 294 23.5% 10 2.5 -115.2% 28 6.6 26.4% 27 -4.2 -141.6% 24
J.Freeman 606 22.9% 11 3.3 -57.5% 5 7.8 29.2% 22 -4.5 -86.7% 3
B.Roethlisberger 489 22.7% 12 3.6 -39.4% 1 7.4 42.3% 16 -3.8 -81.7% 2
S.Bradford 601 22.0% 13 0.7 -110.1% 23 7.6 43.2% 15 -6.9 -153.3% 30
Qualifying QB Plays Pct
Pressure
Rk with Pass Pressure without Pass Pressure Difference
Yds DVOA Rk Yds DVOA Rk Yds DVOA Rk
P.Rivers 592 21.1% 14 0.3 -160.1% 37 7.2 51.1% 9 -6.9 -211.2% 39
J.Flacco 578 20.9% 15 2.1 -111.7% 24 7.6 42.1% 17 -5.5 -153.8% 31
C.Henne 350 20.9% 16 0.8 -134.4% 33 7.2 20.0% 30 -6.4 -154.4% 32
J.Locker 373 20.6% 17 2.8 -97.6% 18 7.0 13.5% 32 -4.2 -111.0% 8
T.Romo 698 20.2% 18 4.1 -79.4% 13 7.6 49.8% 10 -3.6 -129.1% 15
J.Skelton 218 20.2% 19 0.1 -150.5% 35 6.0 -0.7% 37 -5.9 -149.7% 28
A.Rodgers 646 20.1% 20 2.8 -50.4% 3 7.9 64.8% 6 -5.1 -115.2% 10
R.Tannehill 541 20.0% 21 2.1 -118.2% 29 7.3 26.4% 28 -5.3 -144.5% 26
J.Cutler 507 19.5% 22 2.7 -91.5% 17 7.1 26.1% 29 -4.4 -117.6% 11
C.Palmer 603 19.4% 23 2.8 -69.8% 12 7.6 27.5% 24 -4.8 -97.3% 4
R.Fitzpatrick 569 19.3% 24 1.4 -113.1% 26 7.3 26.7% 26 -5.8 -139.7% 21
M.Schaub 580 19.3% 25 3.0 -67.5% 10 7.7 34.4% 20 -4.7 -101.9% 6
M.Sanchez 505 18.6% 26 -0.2 -173.0% 39 6.9 9.9% 35 -7.1 -182.9% 38
Qualifying QB Plays Pct
Pressure
Rk with Pass Pressure without Pass Pressure Difference
Yds DVOA Rk Yds DVOA Rk Yds DVOA Rk
M.Cassel 310 18.4% 27 3.2 -113.3% 27 6.8 1.4% 36 -3.6 -114.7% 9
B.Weeden 560 18.0% 28 1.5 -108.9% 21 6.9 16.3% 31 -5.3 -125.2% 14
A.Dalton 595 17.5% 29 1.9 -112.8% 25 7.0 30.4% 21 -5.1 -143.2% 25
E.Manning 576 17.0% 30 2.5 -98.6% 19 8.0 47.7% 13 -5.4 -146.3% 27
B.Gabbert 311 16.7% 31 -0.3 -130.0% 31 6.1 10.9% 34 -6.4 -140.9% 22
M.Ryan 669 16.4% 32 2.0 -83.1% 15 8.0 47.8% 12 -6.1 -130.9% 17
A.Smith 259 15.8% 33 -0.6 -109.6% 22 8.1 62.1% 7 -8.7 -171.7% 35
D.Brees 700 15.6% 34 1.0 -123.3% 30 8.2 48.9% 11 -7.3 -172.2% 36
T.Brady 679 15.5% 35 1.3 -64.4% 8 8.2 76.7% 1 -6.9 -141.1% 23
M.Stafford 778 15.3% 36 2.5 -60.2% 6 7.1 38.9% 18 -4.6 -99.1% 5
R.Lindley 186 14.5% 37 -1.4 -158.7% 36 4.6 -27.9% 39 -6.0 -130.8% 16
M.Hasselbeck 243 14.4% 38 0.1 -164.9% 38 6.5 10.9% 33 -6.3 -175.9% 37
P.Manning 618 13.3% 39 2.8 -81.7% 14 8.4 69.4% 4 -5.6 -151.1% 29
NFL 408.9 20.2% -- 2.2 -90.9% -- 7.5 38.2% -- -5.2 -129.0% --

Michael Vick has been through a lot since 2003: an NFC championship game appearance, a failed personal venture, the NFL's only 1,000-yard rushing season by a quarterback, a failed business venture, 548 days in prison, and redemption in a usually unforgiving sports town. In our eyes, though, he's still the same guy who taught us 10 years ago that "pass protection is more dependent on the quarterback than most people realize." So as FO begins celebrating its tin anniversary, we appreciate his gift of a trip down memory lane: Fun With Sacks, Part II, written back in Football Outsiders' debut season of 2003 by Michael David Smith. Thank you, Ron Mexico.

Game Rewind: Relive every NFL moment…subscribe to Game Rewind.

Although, mobile quarterbacks like Vick -- and to a lesser extent Ben Roethlisberger, Jay Cutler, and Josh Freeman -- have consistently ranked among the most pressured quarterbacks, that pattern distinction hasn't always been as stark as it was last season. Of course, that's because 2012 saw a massive influx of first-year starters who can use their legs to make time for their arms. We're talking, of course, about Bill Barnwell's "Gang of Four."

In addition to finishing among the top eight in pressure rate, Russell Wilson, Andrew Luck, and Robert Griffin also finished among the top five in rushing DYAR, while Colin Kaepernick set the record for most rushing yards by a quarterback in a single postseason, helping Green Bay's Erik Walden earn the X of Great Shame all along the way. But although the Gang of Four seems poised to carry on Vick's "poster boy of pressure" legacy in the coming decade, Luck stands out as the only one who finished 2012 below the league average in DVOA on plays without pressure.

Regardless, these quarterbacks were so good on plays with pressure that it makes some other quarterbacks in the table look especially bad. For instance, thanks to "Fun With Sacks, Part II," we know it's no surprise that Kaepernick was pressured far more often than Alex Smith. And thanks to everything we've learned about the tactical acumen of San Francisco's offensive coaching staff over the past couple of years, it's no surprise that both starters posted top-eight DVOAs when allowed to execute pass plays in a near-vacuum. Look at Smith's DVOA with pressure, however, and you get a glimpse into one potential reason why the 49ers made a permanent change. Perhaps most damning is that an inexperienced Kaepernick was also better under pressure in 2012 than Smith was during his breakout season of 2011 (-76.0% DVOA, ranked 14th).

Meanwhile, Wilson, Luck, and Griffin were busy relegating the rookie seasons of three other starters to the dust bin of history. Ryan Tannehill, Brandon Weeden, and Nick Foles each finished in the bottom half of the rankings for DVOA with pressure; each also finished that low without pressure. In a normal year, that's probably about what we would expect from rookies; 2012 was not a normal year. As a third rounder who only started six games, Foles probably gets the benefit of the doubt. (Also because comparing his pressure rate to Vick's allows us to mention "Fun With Sacks, Part II" again.) Tannehill's struggles -- fairly or unfairly -- are magnified given his draft position (eighth overall), while Weeden seems like a one man performance of One Flew Over the Cuckoo's Nest: Colonel Matterson's age, Cheswick's panic under pressure, Ellsworth's happy feet, and as likely to get "coached up" as Martini playing blackjack.

In addition to the six rookies (and Kaepernick), there were nine other qualifying quarterbacks last season with four or fewer years of starting experience, eight of whom qualified in 2011 as well. Both Christian Ponder and Sam Bradford saw considerable improvement in their DVOA without pressure, but not much with pressure. Blaine Gabbert and John Skelton were terrible across the board for a second-straight year (but both Jacksonville and Arizona fit the phenomenon described in "Fun With Sacks, Part II"). Meanwhile, Sanchez went from terrible to a $2 meal deal.

Freeman, Cam Newton, and Matthew Stafford, on the other hand, have been incredibly consistent according to these stats. (To compare other quarterbacks, you can click here for last year's piece.) Stafford's 2011 rankings (out of 34 qualifiers) were 26th, sixth, 21st, and third as you move across the table. Newton's were 20th, 10th, 15th, and 10th. Freeman's were sixth, seventh, 29th, and first. Except for Newton facing more pressure in 2012, both years paint similar pictures of each quarterback's "game." Stafford is rarely under duress despite dropping back as much as Miley Cyrus. All three quarterbacks have average DVOA or slightly worse on plays without pressure, but are much better than average on plays with pressure.

For Newton and Stafford that description makes sense, but if you've read any of Barnwell's recent Grantland pieces, seeing the words "much better than average on plays with pressure" associated with Freeman's name probably caused a spit-take. Much of his (and Ron Jaworski's) opinions on the matter came from film review, so it's entirely possible that this massive divergence is just another example of stats and scouting having complimentary roles.

However, what I can't figure out is how their pressure stats are completely opposite from ours.* After all, our determination of which plays counted as pass pressure this year came directly from ESPN Stats and Information. So how can Freeman finish second-to-last in Total QBR under pressure when we rank him fifth? This doesn't happen with other players; for example, both systems have Mark Sanchez bringing up the rear. Even if we ignore sacks, scrambles, and intentional grounding penalties, focusing instead on Freeman's actual throws under pressure, he still ranked 10th with a -9.8% DVOA. And remember what we mentioned earlier: Freeman (and Newton and Stafford) seem to have a distinct pressure M.O. Want more evidence for that? How about the following? Even in 2010, Freeman led the league in DVOA on plays with pressure (-3.9% DVOA), and finished fourth even if we only considered actual passes (39.8% DVOA).

*We checked in with our friends at ESPN Stats and Information, and they tell us Freeman ranked 10th in Total QBR when pressured over the course of the season. Re-reading the transcript of what Jaws said, he may have been only referring to Freeman's atrocious six-game stretch at the end of 2012. Nevertheless, I think our general point holds: Both Barnwell and Jaworski intimate that Freeman is not good under pressure, but his pressure stats suggest the exact opposite. Wouldn't be the first time there's a conflict between stats and scouting, but this particular conflict is worth pointing out because Freeman's stats in particular have been so consistently above average for three seasons in a row.

UPDATE: In Comment 15 below, valued FO reader Sander linked to this piece on Bucs Nation wherein the venerable Greg Cosell seems to have provided another clue to the mystery. He said Freeman was second-to-last in Total QBR against "the blitz."

This has become such a stand-alone issue, that I'll have a Freeman-centric, stats-filled XP on it shortly.

Posted by: Danny Tuccitto on 30 Jul 2013

40 comments, Last at 06 Jan 2014, 12:41am by air jordan pas cher

Comments

1
by MilkmanDanimal :: Tue, 07/30/2013 - 3:26pm

Well, this completely destroys my take on Josh Freeman's performance the last few years. Great with time, utterly terrible with pressure in his face. I do think there's been a pretty marked difference in how Freeman handles pressure from the outside (pretty well) vs. up the middle (utterly awful), but the idea that he's actually done quite well under pressure shocks the heck out of me. Yes, stats vs. scouting and all, but I've been expecting a good year from Freeman largely because of the return of Nicks and Joseph and there being a presumably solid block there in the middle of the offensive line.

2
by JoeyHarringtonsPiano :: Tue, 07/30/2013 - 3:48pm

I was confused, as well, after I read Barnwell and Tanier's piece on Josh Freeman, and then looked at this table. Could it be that the game charters are marking a high amount of pass plays as being "with pressure" compared to what others are seeing with film study?

5
by theslothook :: Tue, 07/30/2013 - 5:58pm

As a charter, I can tell you how pressure is done. ESPN has a set of charters who mark a number of things, including plays where there is going to be pressure. FO provides those log books to us charters who then are asked to describe some background info behind those plays already marked as pressure. This includes, how the pressure was generated(ie, blown block, unblocked, ran into, coverage, etc etc) as well as the name of the player(s) involved. We don't designate which plays have pressure and which do not, though we can mark things we BELIEVE should be marked pressure in the comments section.

So no, charters aren't given liberty to mark pressures, only espn charters are. Total qbr uses those same log books so that's why it was so strange.

7
by JoeyHarringtonsPiano :: Tue, 07/30/2013 - 6:23pm

Thanks for the info. I didn't know the exact mechanics of game charting. So there goes that theory. My other theory is that Freeman has particularly terrible/catastrophic outlier plays that stuck in Barnwell and Tanier's head (Barnwell's article has links to videos of Freeman's particularly bad INT's), and they're underselling how he does overall with pass pressure.

Also, since Freeman's overall performance took a nosedive in weeks 12-17, it would be interesting to see his numbers above broken down before and after week 12.

3
by Aaron Brooks Go... :: Tue, 07/30/2013 - 3:57pm

I'm not going to complete bag Vick for a high pressure rate when 60% of his O-line was dead, or worse, and even his replacement linemen were dropping like flies.

Foles finished in the last 10, too.

12
by shah8 (not verified) :: Wed, 07/31/2013 - 2:17am

I find the difference in ranking between unpressured and pressured to be really quite interesting. I definitely have thought that Vick played better when he had protection, but the difference is pretty stark for him and Philip Rivers. Meanwhile, Jay Cutler does not seem to generate *that* much pressure (and Jay Cutler is a mobile QB along the lines of Romo), while also not having a good ranking when non-pressured. Tell me if that's not a big, red, flag for something being wrong upstairs beyond gunslinging.

13
by theslothook :: Wed, 07/31/2013 - 5:03am

In Denver - he actually had a very strong sack rate and really from what I saw, was pretty accurate and had good fundamentals. Now, he had ace offensive mind Mike Shannahan, but still, he looked great. Now hes been in chicago and hes been mediocre to say the least.

there are a few things at work here. I would say his receivers still by in large suck. Yes, Marshall was great, but he's been there for only one year. The rest of the receivers are really just 3rd and 4th receivers and so hes a bit hamstrung by that.

Another possible issue could be that he has become shell shocked. We thought this phenomenon might explain Rivers recent decline. Sam bradford might have it too. Years behind a poor o line gives you the yips and you start not trusting either your protection or your receivers. The culmination of which looks like gunslinging.

It could be one, or both, but I suspect his lack of receivers is the biggest issue now. After all, being there are plenty of effective gunslingers - Brees, Romo, Favre to name a few.

14
by theslothook :: Wed, 07/31/2013 - 5:03am

In Denver - he actually had a very strong sack rate and really from what I saw, was pretty accurate and had good fundamentals. Now, he had ace offensive mind Mike Shannahan, but still, he looked great. Now hes been in chicago and hes been mediocre to say the least.

there are a few things at work here. I would say his receivers still by in large suck. Yes, Marshall was great, but he's been there for only one year. The rest of the receivers are really just 3rd and 4th receivers and so hes a bit hamstrung by that.

Another possible issue could be that he has become shell shocked. We thought this phenomenon might explain Rivers recent decline. Sam bradford might have it too. Years behind a poor o line gives you the yips and you start not trusting either your protection or your receivers. The culmination of which makes him look bad.

It could be one, or both, but I suspect his lack of receivers is the biggest issue now. He may be a gunslinger by trait, but there are plenty of effective gunslingers - Brees, Romo, Favre to name a few.

18
by JoeyHarringtonsPiano :: Wed, 07/31/2013 - 12:03pm

I really do like Cutler as a QB, but I think what he is right now is what he'll always be. He has a top 5 WR1, and a top 5 RB, and no more 7 step drops with 5 man protection. Yes his line is not great, but very good/great quarterbacks have overcome worse offensive lines (do you remember how the 2010 Packers O-line played?). I think he's out of excuses at this point.

19
by Aaron Brooks Go... :: Wed, 07/31/2013 - 12:33pm

Brees is horrific if you can pressure him.

17
by Uncle Rico (not verified) :: Wed, 07/31/2013 - 9:42am

Been his MO since forever, healthy/competent OL or no. Take a look at Roethlisberger's line last year. Or pretty much any year.

As an aside heads-up, the league's game stats and info site (where all these snap counts are originally coming from) now include lineup details for each team, both sides of the ball. It's not broken out by game, rather (I'm guessing) updated weekly in the accumulative. But if someone were to collect/save that data each week, they could get that kind of detail mentioned in this string. Along with countless other data, like nickel/dime/base, specific and most effective personnel groupings, package comparisons, maybe a more refined value over replacement metric, approx value, when injuries specifically occurred, etc.

4
by jchavlik :: Tue, 07/30/2013 - 5:15pm

Interesting, for winning a super bowl, I'd have to say that Flacco certainly had an average/below average year by all accounts.

38
by RickD :: Thu, 08/01/2013 - 5:00pm

We need splits comparing his performance under Cam Cameron against the games under Jim Caldwell. My impression is that he was much better with Caldwell running the offense.

39
by theslothook :: Thu, 08/01/2013 - 5:17pm

The trouble is...he took over in week 15. IN week 15, he posted a -31 dyar. In week 16, he had a dyar 205. And then of course, they rested starters in week 17. Then of course, the playoffs he was magnificent. Not really enough evidence to gauge the Caldwell effect.

6
by theslothook :: Tue, 07/30/2013 - 6:01pm

What makes me sad in this is that pressures, at least in theory, should be able to help us distinguish which offensive lines are better than others. After all, pressure, on the face of it, appears a result of how well your offensive line can hold up against the guy in front. But then you breakdown the numbers and you see the same pattern appearing with sacks as you do with pressure - namely, its the qb thats the biggest story. Clearly, there's some true quality of offensive line play that exists independent from your qb, just as there is for your receivers, but we haven't found the right stat for it. Yet...

8
by Danny Tuccitto :: Tue, 07/30/2013 - 6:40pm

It's an excellent point you make, but I'm a little more optimistic about it. There are a few examples of where changes in lineman seem to have made a big difference. For instance, the massive pressure rate difference between Kolb and Lindley can be at least partially explained by Massie's vast improvement after ARI's bye (i.e., when all of Lindley's starts happened). In other words, as absolutely atrocious as Lindley played last year, there's little or no reason to think his "true" ability to avoid the pass rush is that much better than Kolb's.

Don't have the time to go through all the various permutations, but I seem to remember TEN shuffling their line a bit with Locker back in the second half of the season.

Something else that makes me optimistic is that, with game-by-game snap counts now available, a variable like "changes along the OL" is something we can potentially control for going forward.

9
by Karl Cuba :: Tue, 07/30/2013 - 6:54pm

How do the numbers look if you include sacks in the under pressure category?

10
by Danny Tuccitto :: Tue, 07/30/2013 - 7:36pm

Sacks are included in the under pressure category in the above table. It's essentially (sacks+hits+hurries)/pass plays. Where sacks aren't included is with respect to the rate of pressure on throws only (i.e., [hits+hurries]/passes). That particular table is only in the book, but it really doesn't change much in the context of the "who's to blame?" conversation, though. For instance, Vick's still No. 1 in pressure rate on throws only, and the correlation between the two pressure rates is .913. Of course, that's pretty intuitive when you think about the fact that there are about three times as many hits and hurries as there are sacks.

11
by theslothook :: Tue, 07/30/2013 - 10:56pm

This is sort of an aside and maybe it was mentioned somewhere else, but since the snap count data is now available, will FO be using it to provide a richer AGL metric?

And also, this info may give us a much better idea of what sort of injuries cause what sort of impacts, though that I think will require at least a few more years of data gathering.

16
by Dan :: Wed, 07/31/2013 - 7:48am

There are sources of stopwatch data on QBs, which helps untangle QB & OL further.

PFF clocked Lindley as the quickest trigger in the NFL (min. 100 dropbacks), with an average of 2.38 seconds before the pass/sack/scramble (Brady & P. Manning were 2nd and 3rd, at 2.47 and 2.50). Kolb was in the slowest third of QBs, at 2.84 seconds, and Skelton was quicker than average at 2.60. That matches their Pct Pressure rankings here.

PFF stopwatch numbers for other teammates:

Jacksonville: Henne 2.53, Gabbert 2.56
Kansas City: Cassel 2.83, Quinn 2.78
Philadelphia: Vick 3.07, Foles 2.83
San Francisco: Kaepernick 3.18, Smith 2.89
Tennessee: Locker 2.84, Hasselbeck 2.50

The first two teams (JAC & KC) are essentially ties in timing but have a fairly big gap in Pct Pressure. The last 3 (PHI, SF, TEN) have a significant gap in timing and (like ARI) have a corresponding gap in Pct Pressure (although it's surprising how little pressure Alex Smith faced, given that he was one of the slower QBs in the league).

20
by Aaron Brooks Go... :: Wed, 07/31/2013 - 12:37pm

SF runs quite a bit of play-action, right?

22
by CBPodge :: Wed, 07/31/2013 - 2:32pm

I'm going to go out on a limb and say that this is the only list that you'll ever see that has a ranking of:

1. Ryan Lindley
2. Tom Brady
3. Peyton Manning

24
by Bobman :: Wed, 07/31/2013 - 2:35pm

Well, here, AND Matt Millen's fantasy draft board (after 37 WRs he had rated ahead of them).

Is it past the statute of limitations for bashing him as GM?

31
by Jerry :: Wed, 07/31/2013 - 6:50pm

Not even the City of Detroit's bankruptcy absolves Millen's debt to Lions fans.

32
by JoeyHarringtonsPiano :: Wed, 07/31/2013 - 8:02pm

Seconded. Making fun of Millen will have more mileage than making fun of Brett Favre's retirements/unretirements and proto-Carlos Dangering.

33
by theslothook :: Wed, 07/31/2013 - 9:10pm

Does it help any that Matt Millen WAS the gm that drafted Calvin? In some sense, he did leave detroit with a nice parting gift for the future.

34
by Jerry :: Wed, 07/31/2013 - 11:09pm

With all those high draft packs, and all the receivers Millen took, it's disappointing that only one worked out.

35
by Vincent Verhei :: Thu, 08/01/2013 - 1:09am

As long as Matt Millen is being paid money to go on TV and give his opinions on football, it's perfectly fine to make fun of him.

26
by RoninX (not verified) :: Wed, 07/31/2013 - 2:57pm

Regarding the discussion of the quality of offensive lines: I wonder how comparing "QB pressured %" with "percent of plays that opponents brought pressure" would look. Clearly scrambling (are we calling them "mobile" these days?) QBs would have a different profile from pocket passers but I wonder if this would line up with scouting impressions of offensive lines moreso than focusing on the QB.

15
by Sander :: Wed, 07/31/2013 - 7:42am

Ron Jaworski wasn't referring to plays with pass pressure, but plays where the opponent sent pressure (ie. blitzed). That isn't clear from Jaworski's original comments, but Greg Cosell later clarified that.

29
by Danny Tuccitto :: Wed, 07/31/2013 - 4:00pm

Thanks for that! Will add to the editor's note at the bottom of the piece.

21
by CBPodge :: Wed, 07/31/2013 - 2:25pm

I suspect the thing with Freeman might be that when he makes mistakes, he makes noticeable ones. You don't really notice a guy completing ten checkdowns under pressure, you notice a guy throwing a bad interception under pressure. From a numbers standpoint, the positives from the ten checkdowns might cancel out or even outweigh the negatives from the interception. But from a scouting standpoint, you remember the interception and forget the checkdowns.

I wonder if you could see the outcome of the plays under pressure that Freeman makes, in terms of categories (i.e. whether he passes short, mid or long, completions, yardage etc) and see how they stack up against some other QBs?

23
by Bobman :: Wed, 07/31/2013 - 2:33pm

This is a new and wonderful concept for me, but someone please help me understand: Wilson, for example, was 2nd ranked with pressure and 2nd ranked w/o pressure (nice job, rook) which results in an overall (or difference) ranking of 12. But Roethlisberger was #1 with and #12 without, meaning pressure affects his production less, and he's #2 ranked by difference? Is that how it works?

Luck (Good lord, look at that boy's workload--nearly twice league average) was ranked #4 (nice) and #23 (nasty) which results in a top difference ranking since pressure affected him the least with respect to his non-pressured performance... right?

Cool stuff. Reminds me of the WKRP in Cincinnati episode about how drinking affects one's reactions. Johnny Fever was much better when blasted, but with no "pressure" was incompetent. Jeez am I old.

27
by RoninX (not verified) :: Wed, 07/31/2013 - 3:16pm

I think you are looking at it correctly (though Rothlisberger is #16 w/o pressure), but I'm not sure I agree with the conclusion you draw regarding Luck. Remember that several of the top ten "difference" performers got their slot by performing quite poorly w/o pressure (Palmer, Locker, and Luck were all more than 10% under the 38.2% DVOA league average%). Achieving a smaller drop by not having as far to fall isn't necessarily an indication of talent under pressure. The w/ and w/o rankings seem to be the better marks to compare across QBs rather than strictly applying the difference value. Luck stacks up quite well on that merit alone... at least in the "with pressure category"!

25
by RoninX (not verified) :: Wed, 07/31/2013 - 2:56pm

In what sense are Manning and Brady "immune" to pressure (statistically speaking)? Just that year-in, year-out they are among the least frequently pressured? Because their "difference" DVOA% when actually under pressure (-151.1 and -141.1 respectively) both exceed the NFL average substantially.

28
by Danny Tuccitto :: Wed, 07/31/2013 - 3:58pm

What I meant by that was that, as long as we've been tracking this, their DVOA w/ pressure is much better than average.

Something I was going to discuss in the article, but just felt would get too convoluted, is that the "Difference" column needs to be viewed in context. Since the difference depends on the 2 splits, a guy like Peyton ends up with a crappy difference when he's actually well above-average on both of the splits. Basically, the way I try to wrap my brain around this stuff is to group the quarterback rankings into a 3x3x3 matrix of "profiles," for lack of a better term (empty "profiles" omitted):

1) Good difference, good w/ pressure, good w/o pressure: Wilson, Rodgers,
2) Average difference, good w/ pressure, good w/o pressure: Kaepernick, Griffin, Romo, Brady
3) Good difference, good w/ pressure, average w/o pressure: Luck, Newton, Freeman, Roethlisberger, Palmer, Schaub, Stafford
4) Average difference, average w/ pressure, good w/o pressure: Ryan
5) Good difference, average w/ pressure, average w/o pressure: Ponder

6) Bad difference, average w/ pressure, good w/o pressure: Vick, E.Manning, Smith, P.Manning
7) Bad difference, bad w/ pressure, good w/o pressure: Rivers, Brees
8) Average difference, average w/ pressure, average w/o pressure: Fitzpatrick, Dalton
9) Bad difference, average w/ pressure, average w/o pressure: Bradford, Flacco
10) Good difference, average w/ pressure, bad w/o pressure: Locker, Cutler

11) Average difference, average w/ pressure, bad w/o pressure: Weeden
12) Good difference, bad w/ pressure, bad w/o pressure: Cassel
13) Average difference, bad w/ pressure, bad w/o pressure: Quinn, Foles, Tannehill, Gabbert, Lindley
14) Bad difference, bad w/ pressure, average w/o pressure: Kolb
15) Bad difference, bad w/ pressure, bad w/o pressure: Henne, Skelton, Sanchez, Hasselbeck

At least for the sake of my own comprehension, I think that helps put the "difference" in better context.

30
by RoninX (not verified) :: Wed, 07/31/2013 - 5:18pm

Thank you. Great list, I'd like to think I was working my way around to viewing the data this way.

I actually was trying to make a similar point about "
difference" in an above post, so of course I should not have used the "difference" split to compare them in mine. Still, I think your matrix list displays (at least in Manning's case) what I was getting at: he is in the Bad Diff, Avg w/o, Good w/ category. To me that is not the profile of a pressure immune passer... however, your post already answered the obvious follow-up real question: historically Manning has been in the 2) row instead of the 6) row.

36
by peterplaysbass :: Thu, 08/01/2013 - 11:25am

I was thinking about slapping the table into Excel and formatting the two DVOAs and the difference to rank according to average - you've summarized nicely. A few observations:

1) interesting that Brees is bad with pressure. I'm guessing that he was better in years that aren't 2012.

2) Group 12 is interesting. I assumed the QB (Cassel) was bad w/o pressure but terrible w/ pressure (hence the "good" difference), but it turns out he's terrible w/o pressure and just bad w/ pressure. Unless the contextual difference between the 2012 Chiefs and the 2013 Vikings is closer to the difference between the 2012 Chiefs and the 2008 Patriots than I think, Ponder ought to keep his job barring injury. Although Cassel's 2010 performance in Kansas City is intriguing.

3) Hasselbeck is beyond done as a starter in this league

4) I don't understand some of the Locker/Cutler love I'm hearing

5) Flacco gives hope to good teams with meh QBs - St Louis could win it all sometime in the next few years

6) (my 6th point got a little Vikings heavy, so I'll post in a new comment)

37
by peterplaysbass :: Thu, 08/01/2013 - 11:35am

Matt Ryan and Christian Ponder are very similar with two main differences: Ryan saw opposing defenses send pressure a lot less often(16.4% vs 25.0% - that's 52% more pressure for Ponder), and his average yards per attempt on plays with no pressure was much higher (8.0 vs 6.8).

Ponder needs to increase his yards per attempt to be closer to Ryan's to experience similar success. Something tells me Roddy White (9.4 yards per target) and Julio Jones (9.3 y/t) help that difference. Gonzalez did well also (7.5 y/t vs. Rudolph's 5.3 y/t), but that isn't a knock on Rudolph. Hopefully Jennings and Wright/Patterson will close that gap better than Harvin (8.0), Jenkins (6.2) and Simpson (5.3) did last year.

I'm not sure how to get the Viking's opponents to send pressure less often - although better WRs may help. It makes me wonder if defenses are loading the box against Peterson and then sending extra rushers when the play changes to a pass play - sure defenses weren't gearing up to stop Turner last year.

I hope to see this metric again next year - if Steven Jackson changes the Falcons to be a little more like the Vikings and the new WRs change Minnesota to be a little more like Atlanta, it will be interesting to see how the numbers of Ryan and Ponder react.

40
by air jordan pas cher (not verified) :: Mon, 01/06/2014 - 12:41am

"La sous-Pr