Writers of Pro Football Prospectus 2008

11 Jul 2009

On Rex Ryan And "Blitzing"

Pro Football Weekly has an interview with Bart Scott (hat tip: PFT) that discusses Rex Ryan's love of the blitz.

PFW: Ryan’s defense has the reputation of being very blitz-happy. Is that accurate?

Scott: A lot of people think that we blitz a lot, but when you break it down and see how many people are actually rushing, it’s usually not more than four. So if only four people are rushing, how is it a blitz? It’s all about switching up your rushes and not having a blueprint. If you’re constantly changing things up, how are teams going to get a read on you and expect you to be in certain places at certain times? You can’t.

Is Scott correct? Well, this is the exact sort of question the Game Charting Project can answer.

The answer is, somehow, both yes and no. Here's the Ravens' blitz frequencies in 2008:

Rushers Frequency Rank
3 11.7% 5
4 52.3% 29
5 27.6% 5
6 8.4% 15
7+ 2.2% 8

Scott is technically right -- the Ravens did blitz four or fewer 64 percent of the time. However, as you can see from their league-wide ranks, they blitzed five or more far more than the average team.

They did not, however, lead the league in blitzes of five players or more.

Team 3-4 Rushers 5+ Rushers % of Plays
w/5+ Rushers
DAL 297 268 47.4%
CHI 398 247 38.3%
WAS 328 201 38.0%
PHI 353 216 38.0%
ARI 339 205 37.7%
BAL 352 207 37.0%

So then, it all depends on your baseline. If your baseline is in a vacuum, Scott is right. If your baseline is the rest of the league, Scott is wrong.

Of course, if you want to get this sort of information for every team, you can head to our Online Store and purchase the PDF of Football Outsiders Almanac 2009.

Posted by: Bill Barnwell on 11 Jul 2009

19 comments, Last at 13 Jul 2009, 7:06pm by Mr Shush

Comments

1
by Dice :: Sat, 07/11/2009 - 2:54pm

Kinda the usual suspects on the league-wide list, but interesting to see the Bears. Trouble in Tampa-2?

3
by shake n bake :: Sat, 07/11/2009 - 4:11pm

The Bears have always had a more aggresive, blitzing, style of T-2, but yeah I'd say that much blitzing showed trouble. Specifically on the front 4. A Tampa-2 that can't consistently get pressure with it's front 4 is a bad Tampa-2 (Just look at how Indy's #1 scoring D did in the 2007 playoffs with Freeney on IR, Mathis playing on two bad knees and Raheem Brock playing DE with a rib injury while still bulked up to be a starting DT.

Billy f'ing Voleck tore up the pass D for the winning score, because a T-2 with no pressure is a terrible defense.

10
by drobviousso :: Sat, 07/11/2009 - 11:00pm

This is actually specifically covered in FOA in the Bear's chapter.

2
by Telamon :: Sat, 07/11/2009 - 3:38pm

Kinda unclear what the second graph means.

4
by phillyangst :: Sat, 07/11/2009 - 4:12pm

Yeah, what the heck is the 2nd set of numbers telling us?

5
by Bill Barnwell :: Sat, 07/11/2009 - 4:35pm

Sorry about that, column headers added to the second table.

Not to sound like TOO much of a shill (only a proper amount of shilling is acceptable), but I do cover the Bears' newfound interest in blitzing extensively in the team chapter.

6
by Marko :: Sat, 07/11/2009 - 4:58pm

I don't think it's quite accurate to say the Bears had a "newfound interest in blitzing." I would say that the Bears had a continued interest in getting pressure on the QB. Since the front four failed to generate any consistent pressure (due in large part to Tommie Harris' injuries and backup DE Mark Anderson's continued disappearing act since his stellar 12-sack rookie year in 2006), the Bears resorted to blitzing heavily to try to generate that pressure. However, their blitz schemes were very ineffective, which resulted in their DBs getting burned repeatedly. That's why the Bears' overall defense (and specifically their pass defense) had such a bad year.

Hopefully the addition of Rod Marinelli as DL coach will help rejuvenate the pass rush from the front four and allow them to blitz less this year.

7
by njjetfan12 :: Sat, 07/11/2009 - 5:01pm

As a Jets fan, I can say that any amount of extra pass rushers/ actual pass rush not caused by Kris Jenkins collapsing the pocket will be a welcome sight

I'm a little confused though, because the first table says that the Ravens rushed 5 or more people 38.2 % of the time, when you calculate the totals, but the 2nd table says 37%.

Also, by these statistics does a "blitz" only occur on a pass play, because on run plays everyone eventually gravitates to the ball? And, on a well-executed play fake where certain defenders end up close to or at the line of scrimmage, are they counted as blitzers?

8
by Sergio :: Sat, 07/11/2009 - 9:31pm

Blitzes may only occur on passing plays in these stays because the game charting project does not tally number of rushers nor blockers in running plays.

Also, we chart a rusher as someone who rushes the QB (redundant and subjective, I guess). Players drawn by PA should not count as rushers...

-- Go Phins!

11
by Kibbles :: Sun, 07/12/2009 - 2:17am

I can't speak specifically to these numbers, but there is a such thing as a "run blitz" (or a blitz designed specifically to shut down a run). While it's true that everyone gravitates towards the ball *EVENTUALLY*, at the snap the defense has to make an educated guess about where the ball is going and how best to get a defender there, which means some people will rush with the intention of penetrating the backfield and tackling whoever has the ball (be it QB or RB), while others will drop back to protect against the possibility of the pass. Thus, even on called runs, you could in theory count the number of rushers simply by looking at how many defenders are moving forward immediately after the snap (before they know whether it's a pass or a run).

16
by Matt W (not verified) :: Sun, 07/12/2009 - 12:37pm

Re: I'm a little confused though, because the first table says that the Ravens rushed 5 or more people 38.2 % of the time, when you calculate the totals, but the 2nd table says 37%.

The numbers in the first table also add up to more than 100%. My guess would be that the "6" row should really be 6+, since rows 3 through 6 add up to 100%. But then Baltimore would be rushing more than four 36% of the time, not 37%. So who knows.

[Possible suggestion for future web designers: It's be nice to be able to see the post while writing a comment. I realize that this could be a pain to implement.]

9
by Theo :: Sat, 07/11/2009 - 9:58pm

Is there a correlation between blitzing and good defense?
I know that in youth football and lower league football all out blitzes are best way to win a game. (6 defenders on the 6 eligible receivers and the rest rushes into the gap)
I'd like to know if that's true for NFL teams.

12
by Kibbles :: Sun, 07/12/2009 - 2:55am

I think there's probably a very strong correlation between creating pressure and winning football games. On defenses where blitzes result in pressure, I'd expect a positive correlation to wins. On defenses that blitz and still don't get pressure, I'd expect a negative correlation to wins. I don't know which effect would wind up being the stronger, so I don't know what the league-wide trend would be.

I remember in 2005 the Denver Broncos went 13-3 and had the #9 defensive DVOA despite absolute dreck at the D-Line (that was the height of the "Browncos" years, right after Trevor Pryce walked) by becoming the most blitz-happy team I have ever seen. It was almost surreal some of the stuff they did. Against the Eagles that year, for instance, they opened the game by rushing 6+ defenders on four of the first five snaps, including a 7-man blitz and an insane "kitchen sink" blitz where they rushed NINE. That was absurd- they basically had two defenders playing cover-2 in case the Eagles managed to get a pass off, and everyone else just took off after McNabb the second the ball was snapped. I have never, ever seen another 9-man blitz before or since in the NFL. For all of his faults, I will give Larry Coyer this- he had cajones. I sometimes wondered if he specifically unveiled that beauty against the Eagles to try to show up Jim Johnson.

15
by Bowl Game Anomaly :: Sun, 07/12/2009 - 12:01pm

FO has studied this. I think they found that rushing 5 is almost identical to rushing 4, so they only looked at rushing 6 or more as blitzing. From there, they found that 6-blitzing was very effective in certain situations and less so in others. (I don't remember which situations were optimal for blitzing or not blitzing.) The article is in one of the PFPs.

18
by Mr Shush :: Mon, 07/13/2009 - 7:03pm

If I remember correctly, the finding was that rushing 5 from a 4-3 base was essentially ineffective - it didn't (on average) lead to more pressure, and it did lead to more completions and yardage compared to a 4 man rush. Blitzing more than 6 was also generally not a good idea (though the samples there are very small). 6 blitzes were, as you rightly say, quite effective in some situations and less so in others, as were 5 man blitzes from a 3-4.

13
by Alex51 :: Sun, 07/12/2009 - 4:23am

Not sure, but looking at that chart, it seems the most blitz-happy teams last year were pretty good. Four of the six teams in the chart had top-10 defenses. I also find it interesting that three of them are from the NFC East.

14
by Dice :: Sun, 07/12/2009 - 11:49am

Well, the skins can't get pressure with their front four, the Eagles have Jim Johnson, so of course they'll blitz, and I guess the Cowboys need to blitz to get some pressure not named DeMarcus Ware. The Jints didn't need to blitz as much with their very good D-line.

I wonder if I ought to buy a copy and send it to Greg Blache.

17
by tuluse :: Mon, 07/13/2009 - 12:57pm

The Giants do blitz a lot though, or they did with Spags.

19
by Mr Shush :: Mon, 07/13/2009 - 7:06pm

Wade Phillips has always been a blitzomaniac. His version of the 3-4 is a very blitz-heavy one, and so far as I know has never worked terribly well with anyone else calling the plays. I don't think much of Phillips as a head coach, but he's not far behind the likes of Jim Johnson or Dick LeBeau when it comes to defensive playcalling.