Writers of Pro Football Prospectus 2008

Most Recent FO Features


» A History of Undrafted NFL Quarterbacks

Given the historical success of undrafted quarterbacks in the NFL, Tony Romo might as well be a national treasure. We look at the impact of developmental leagues on undrafted quarterbacks, and just how many players have tried to break through in a recent season.

06 Dec 2013

Under Pressure: Sacks by Confusion

by J.J. Cooper

Peyton Manning isn’t mobile. He doesn’t have Dan Marino’s blink-and-you-miss-it release. But if you’re an offensive lineman, Manning is your best friend.

Manning simply doesn’t give up cheap sacks. In part because of his lack of mobility, Manning’s internal clock is extremely highly tuned.

With left tackle Ryan Clady missing much of the season, the Broncos offensive line has struggled to protect the NFL’s oldest starting quarterback. But even with that Manning has only been sacked 15 times, and of those sacks 14 have come because a Broncos blocker was physically beaten.

This year to get a better idea of how sacks are occurring for different teams, each sack is logged as either a blown block (a blocker is simply physically beaten by a defender), a result of confusion (a problem with determining who should block a rusher), or a coverage sack (where the quarterback held the ball for too long until solid pass protection broke down). For the table below, QB fault sacks where a quarterback simply dropped the ball or fell down untouched were lumped into the coverage sacks.

Admittedly it can be a subjective distinction at times, but most sacks are rather straightforward.

In looking at how the sacks broke down by team, it’s stunning how the Broncos sacks were almost always a result of a physical breakdown. The only other sack was one where Manning dropped the ball untouched. Denver is one of two teams who have not allowed a sack because of a confused blocking assignment this year. (San Diego is the other.) The Broncos are also one of only two teams to allow only one coverage sack. (New Orleans is the other.)

On the other end of the spectrum, a full 50 percent of the Raiders 38 sacks are coverage sacks. The Ravens, Chiefs, and Patriots are three potential playoff teams with some significant problems with their assignments: all three have allowed 10 or more sacks where they struggled to figure out how to pick up every pass rusher.

Offense Blown
Pct Confusion Pct Coverage Pct Total
ARI 25 71.4% 7 20.0% 3 8.6% 35
ATL 22 73.3% 3 10.0% 5 16.7% 30
BAL 20 51.3% 10 25.6% 9 23.1% 39
BUF 14 43.8% 5 15.6% 13 40.6% 32
CAR 13 43.3% 7 23.3% 10 33.3% 30
CHI 8 38.1% 9 42.9% 4 19.0% 21
CIN 12 46.2% 8 30.8% 6 23.1% 26
CLE 20 47.6% 10 23.8% 12 28.6% 42
DAL 16 57.1% 7 25.0% 5 17.9% 28
DEN 14 93.3% 0 0.0% 1 6.7% 15
DET 8 53.3% 3 20.0% 4 26.7% 15
GB 20 64.5% 6 19.4% 5 16.1% 31
HOU 16 51.6% 10 32.3% 5 16.1% 31
IND 22 75.9% 5 17.2% 2 6.9% 29
JAC 20 51.3% 12 30.8% 7 17.9% 39
KC 14 43.8% 10 31.3% 8 25.0% 32
Offense Blown
Pct Confusion Pct Coverage Pct Total
MIA 35 77.8% 6 13.3% 4 8.9% 45
MIN 18 52.9% 7 20.6% 9 26.5% 34
NE 18 56.3% 10 31.3% 4 12.5% 32
NO 20 83.3% 3 12.5% 1 4.2% 24
NYG 22 71.0% 6 19.4% 3 9.7% 31
NYJ 19 45.2% 8 19.0% 15 35.7% 42
OAK 14 36.8% 5 13.2% 19 50.0% 38
PHI 13 37.1% 8 22.9% 14 40.0% 35
PIT 20 54.1% 4 10.8% 13 35.1% 37
SD 18 81.8% 0 0.0% 4 18.2% 22
SEA 15 50.0% 6 20.0% 9 30.0% 30
SF 11 35.5% 8 25.8% 12 38.7% 31
STL 18 62.1% 6 20.7% 5 17.2% 29
TB 14 42.4% 9 27.3% 10 30.3% 33
TEN 17 58.6% 5 17.2% 7 24.1% 29
WAS 13 39.4% 9 27.3% 11 33.3% 33
NFL Total 549 54.2% 212 21.4% 239 24.4% 1000


The Falcons have not had many problems with assignments in pass protection during the 2013 season. It’s one of the few things that have gone well for Atlanta. But that all seemed to fall apart against the Bills.

Buffalo rushed only two men on a third-and-2, but still managed to sack Falcons quarterback Matt Ryan as one of the two rushers, Manny Lawson, came unblocked off the edge for a 1.4-second sack. In five seasons of logging sacks, there has been only one other two-man rush that led to a sack, and that involved Joe Flacco rolling out and running out of bounds when he found no one to throw to. The Bills managed to get a free rusher while bailing out to put nine men into coverage.
This one provided the third fastest sack of the 2013 season.

You may ask how a two-man rush can get a rusher free? Well, when you’re playing the Falcons this year, anything is possible. This sack was just part of what was a disastrous day for the Falcons’ pass protection. The Bills had six sacks. Jerry Hughes picked up a pair of sacks, once with a nice stunt to the inside and another when he simply drove left tackle Lamar Holmes into Matt Ryan.

But the biggest problems were a lack of blitz recognition and cohesiveness from the Falcons offensive line. In addition to Lawson’s 1.4-second sack, defensive end Corbin Bryant was allowed to come in unblocked on a 1.8-second sack. Once again, the Falcons just couldn’t figure out who to block. At least this time it was a six-man rush, but when guard Justin Blalock blocked down to pick up a man on his inside shoulder and left tackle Holmes blocked out to pick up a man on his outside shoulder, it did all but send Bryant an Evite to sack the quarterback.


Bucs quarterback Mike Glennon and 49ers quarterback Colin Kaepernick shared the honors this week as each had a 6.7-second sack. Glennon’s came when he tried to backpedal away from pressure and ended up dropping the ball without being touched. Kaepernick’s long sack came when he tried to buy time with a scramble to his left, but he was eventually run down for a short sack by yardage (one yard lost) but a long one on the clock.

Posted by: J.J. Cooper on 06 Dec 2013

38 comments, Last at 27 Jan 2016, 4:59am by agariogames1


by nat :: Fri, 12/06/2013 - 12:57pm

Why ever do you do this table as a percentage of sacks, instead of a percentage of dropbacks? You give a ton of false impressions this way.

For example, Denver's OL (with Manning's help, naturally) is actually quite good at avoiding sacks due to blown blocks. They rank 5th in that category. Of course, they're 1st in avoiding confusion sacks (they don't blow assignments and the protection calls are good), and 2nd in avoiding coverage sacks (Manning doesn't hold the ball too long, and people get open).

by Perfundle :: Fri, 12/06/2013 - 1:22pm

Because this table is only looking at what percentage of sacks were the fault of the QB, not at how well each teams's OL pass blocks. The pass protection column on the OL page is plenty sufficient in showing the strength of Denver's pass protection.

by nat :: Fri, 12/06/2013 - 1:47pm

But why do the table that way? It gives false impressions and leads directly to the statement "It’s stunning how the Broncos sacks were almost always a result of a physical breakdown" when the reality is that the Broncos are very good at avoiding physical breakdowns in blocking.

It's not a false statement, per se. J.J's allowed to be stunned by whatever he wants. But if the point was to highlight how much better they were at avoiding confusion, this is a very backhanded, even insulting way of saying it.

The overall impression this gives is that J.J. Cooper believes that the table shows the Broncos OL is bad at the physical aspect of blocking. If so, he has confused himself. The data shows nothing of the sort.

That's the problem with doing the table this way.

by nottom :: Fri, 12/06/2013 - 2:18pm

Obviously people are free to interpret things as they wish, but I don't get that impression from the table at all. I see that the only time Denver gives up sacks is when they just get physically beat, not because they screwed up their blocks or because Manning held the ball too long and they don't get beat physically very often either.

by JoeyHarringtonsPiano :: Fri, 12/06/2013 - 2:56pm

Yea, my intepretation of the table is that, Manning and his Oline have their shit together when it comes to protection assignments, and the few times they allow sacks, it's because they physically got beat, and not due to mental errors/missed assignments, etc. I'm sure the majority of people who visit this site are aware the Broncos have spent the entire season at or near the top of the Adjusted Sack Rate rankings.

by tuluse :: Fri, 12/06/2013 - 4:59pm

Without putting the number of sacks in the context of the number of passing plays it makes it very easy to draw that conclusion from statistical noise instead of good data.

by JoeyHarringtonsPiano :: Fri, 12/06/2013 - 5:16pm

I guess it's a matter of knowing your audience. I'm laboring under the assumption that most people who read this article has scanned the O-line tables in the stats sections. If this article showed up on ESPN.com, then you're right that the readers would probably draw conclusions from data with no context.

by tuluse :: Fri, 12/06/2013 - 5:27pm

Do you know how many passing attempts each team has? I certainly don't. While I can check ASR to correlate the numbers presented here, that's a lot of extra work on my part to do for 32 teams.

by theslothook :: Fri, 12/06/2013 - 5:46pm

Have to be honest, the heading of this article really just suggested it was pointing out which quarterbacks take which kinds of sacks. I didn't really read it as an indictment of o line quality. I'm pretty sure even the Fo staff knows they do NOT have the stats to measure o line quality with regards to pass protection.

by tuluse :: Fri, 12/06/2013 - 6:21pm

This article creates answers that don't exist by presenting information with poor or non-existent context. It makes you think there is a large difference between 1 and 9 when it could easily be statistical noise.

by JoeyHarringtonsPiano :: Fri, 12/06/2013 - 5:48pm

I thought you were referring specifically to the Broncos, not all 32 teams. And even then, the extremes at either end are more interesting (to me at least, I'm not presuming to speak for you) in context with this table.

by Chrisgmiddle (not verified) :: Sat, 12/07/2013 - 1:00am

Actually what the table doesn't show is how long the QB held onto the ball. Sure it shows coverage sacks, but it does nothing to indicate how much pressure a QB would have been under if he wasn't lightning fast at getting the ball out of his hands. And nobody is faster snap to pass than Manning. His offensive line rarely has to hold their blocks for much more than a couple seconds.

by Neffarias_Bredd :: Fri, 12/06/2013 - 1:24pm

If you read the bit before the table he makes it pretty clear. This article is focusing more on QBs and why they take different sacks not on O-lines and how often they're sacked. The % number gives you a clear picture of where each sack comes from.

by tuluse :: Fri, 12/06/2013 - 2:56pm

Nat is right. It makes the Bronco's line look bad at something when they're actually good at all 3 things, just less good at one aspect.

Edit: imagine a line that gave up 2 sacks from being physically beat, 1 from confusion, and 1 from coverage. 50% of their sacks would be from being physically beat, but they only gave up 4 sacks in total. No one would be "at fault" for anything. It would be the most dominant offensive line of all time.

by LionInAZ :: Fri, 12/06/2013 - 7:14pm

I don't agree with this. The last column, which shows the total number of sacks, shows that the Broncos don't give up very many sacks at all. No analytically competent person could conclude from this table that the Broncos OL stinks.

by Chrisgmiddle (not verified) :: Sat, 12/07/2013 - 1:03am

Manning's offensive lines are always proficient at preventing sacks. What's the common denominator here? Manning is. He gets his team into the right plays and gets the ball out of his hands so fast that the offensive line almost might as well not even be there a lot of the time.

by Bobman :: Sun, 12/08/2013 - 1:31am

and often enough over the past 7 seasons or so, they haven't been.

by Kal :: Fri, 12/06/2013 - 3:06pm

Just curious on your opinion - do you think that confusion sacks are a result of poor line play or poor line assignment? Usually line assignment goes to the QB as well (or at least he helps a lot) which would indicate even further how good Manning has been this year.

by nat :: Fri, 12/06/2013 - 3:21pm

I'm just saying that the table draws your eyes to the wrong things and invites you to draw the bad conclusions. It takes extra work on our part to see the right things and draw the correct conclusions.

With a table of sack rates (rather than relative sack rates among the three types of sacks), J.J. would have been "stunned" by the right thing: how low the Broncos "confusion sack" and "coverage sack" rates are. Instead, he was "stunned" by how high the "blown block" rate was -- which was exactly the wrong thing to latch on to.

by Kal :: Fri, 12/06/2013 - 3:27pm

If you say so; he mentioned how good it was in the article, and the Broncos are the best in the league in both coverage and confusion sacks. Both stand out to be pretty heavily. It's also kind of surprising how bad the Denver line is; it's comparable to most of the bottom third of the league.

by Perfundle :: Fri, 12/06/2013 - 6:30pm

"J.J. would have been "stunned" by the right thing: how low the Broncos "confusion sack" and "coverage sack" rates are."

But that table won't be as useful in telling you how much of the very low number of sacks were on Manning, would it? An extremely high blown block rate is the same thing as an extremely low coverage sack rate, and what's so bad about that?

by LionInAZ :: Fri, 12/06/2013 - 7:17pm

If you're not willing to do the extra work to extract meaning from a table, what are you doing on an analytics site? Do you need to have everything pre-digested for you?

by nat :: Fri, 12/06/2013 - 9:23pm

Oh for crying out loud.

It's bad stats. You don't compare "blown block" sacks to "coverage" sacks to conclude that a line blows a lot of blocks. You compare to how other lines blow blocks. And you don't say it's stunning how many they blow. That's just misleading.

I get that he means that the ratio is stunning. But the presentation makes it look like the problem is blown blocks. It's not. The Denver line is one of the five best at avoiding blown block sacks.

Obviously I did the work. Is FO supposed to be doing bad tables so you and I get some homework? Really?

by The Ancient Mariner (not verified) :: Sat, 12/07/2013 - 12:54pm

Correct. You don't. Which, since this is *explicitly not* what this article is doing, is utterly irrelevant.

I'm amazed how many people are criticizing the author here for not writing the article they think he should have written.

by tuluse :: Sat, 12/07/2013 - 1:49pm

The Broncos have 483 pass attempts. Can you say with certainty that the difference between something happening 1 in ~500 events vs 9 times is statistically meaningful?

by nat :: Sat, 12/07/2013 - 3:20pm

Taken on its own terms, the article tries to use the blown block sacks as a percentage of all sacks to argue or at least illustrate how Manning is the "best friend" of an offensive line that has "struggled". It does this by being "stunned" by the high percentage.

It turns out that the high percentage represents the fifth best blown block sack rate is the league. Combined with the "confusion" sacks, the Broncos are near or at the top in avoiding line-caused sacks. They are not struggling. They are doing great.

This leaves the implicit idea that a low "coverage" sack rate might somehow indicate that the QB's internal clock is saving the line's bacon in other ways. That's a stretch, since it could indicate receivers getting open, a low number of opportunities, etc. Or maybe Manning's good at throwing the ball away to avoid a long sack.

The problem is that, by using this bad table, J.J. can't argue what he tries to argue. In fact, this kind of stat is useless to write almost any imaginable article. To get anything sensible out of it, you have to build your own table with opportunities and rates. When you do, you find out J.J. had it exactly backwards. The line isn't struggling, and is probably helping their QB more than most lines do.

by ChicagoRaider :: Fri, 12/06/2013 - 3:39pm

When Terrelle Pryor's offensive line got physically beaten at one location and was confused in another, but he ran around behind the line of scrimmage for 4 seconds before getting tackled, how did you score the sack?

by Aaron Brooks Go... :: Fri, 12/06/2013 - 3:50pm

I don't know, but it probably drove Ben Muth to drink.

by theslothook :: Fri, 12/06/2013 - 4:55pm

I do also wonder how much its qb that sets the protection vs its the o lines job to read and correctly block things. I hate the statement that says its a combination of both because that tells us really nothing. Combination could mean 50 50 or 1 99 in either direction.

I suppose it varies across teams and schemes, but I still like to think they hire o line coaches for a reason and that getting protection assignments should come from training the o linemen. The qb should probably be a last second fail safe. After all, the qb already has a lot to worry about.

There is one instance I think the qb has a very direct impact on setting the protection, and that's calling out the edge blitzers or rushers coming from the slot. There, I would imagine its hard for the o line to get a complete view of the defenses formations and will likely require the qb to tell them how far to set. The rest of the other things, like blocking stunts or setting proper gaps - seems like an o line thing.

by PerlStalker :: Fri, 12/06/2013 - 6:00pm

In an interview a week or two back, Manning was quick to point out how good his center is a changing the protection while Manning is back there doing the chicken dance. That tells me that it's the line that's responsible for setting the protection though Manning works hard to ensure he calls a play that gets them out of bad situations.

From what I've read over the years, it's fairly similar for most NFL teams.

by Perfundle :: Fri, 12/06/2013 - 6:22pm

Well, it's not like Manning is going to take credit for himself there, even if he was responsible.

by Quark (not verified) :: Fri, 12/06/2013 - 5:03pm

Would love to see Eagles split into Foles vs Vick.

by steveNC (not verified) :: Fri, 12/06/2013 - 5:51pm

I like the last line, where we have exactly 1000 total sacks for some reason. We see that 549 is 54.2% of 1000, 212 is 21.4%, and 239 is 24.4%?

by LionInAZ :: Fri, 12/06/2013 - 7:36pm

And the total number of sacks is exactly correct by my count. My guess is that the percentages are a hangover from an earlier draft that wasn't corrected.

by Friv de friv (not verified) :: Fri, 12/06/2013 - 10:43pm

But the pressure is all that can bring more good things then.

by silarude (not verified) :: Tue, 12/10/2013 - 12:57pm

The one thing this table shows, and quite clearly, is how bad Miami's line sucks.

by linuxbasiccommand :: Mon, 01/18/2016 - 10:29am

Tänan Teid väga mõistva suhtumise eest ning jään ootama Teie seisukohta eespool nimetatud ühist huvi pakkuva teema suhtes.
http://www.agariospielen.de http://www.elsaspelletjes.nl http://www.giochiminions.it http://www.gryagario.pl http://www.jeuxdeagario.com

Kui nõukogu algatab soovituse esitamisega ülemäärase tasakaalunihke menetluse, peab selle menetluse lõpetama samalaadse õigusaktiga, st soovituse või otsusega, mis on kehtiv tervikuna/kehtiv selle adressaatidele.
http://www.jeuxdeelsa.com http://www.jeuxdehappywheels.com http://www.juegosdehappywheels.com http://www.juegosdemonkeygohappy.com http://www.juegosdewheely.com

by agariogames1 :: Wed, 01/27/2016 - 4:59am

Just wanted to say that I read your blog quite frequently and I’m always amazed at some of the stuff people post here. But keep up the good work, it’s always interesting. See ya,