On Pattern Matching: Saban's Cover 3-Man Hybrid Comment Count

Seth

Next time you see this you'll know what's going on

In previous layman's discussions on how fancy newfangled anti-spread defenses function I've talked about how Quarters works, and how MSU used aggressive alignments with it to dominate the run game at the cost of greater risk of getting beat over the top. Each time I alluded to the fact that Saban's defense is similar in concept except where Quarters is a Cover 2/Cover 4 hybrid, Saban's is a Cover 3/Cover 1 hybrid.

We will see it this year. Every defense uses some Cover 3 and Cover 1 as a changeup, but Saban's base system, now all over the SEC, has spread into various Michigan opponents. Penn State kept it around while transitioning to Bob Shoop's version of Quarters. Maryland had it last year; not sure if their 4-3 transition includes a coverage shift. I think BYU (which is going back to 3-3-5 with Bronco Mendenhall overseeing it personally) is expected to as well. Michigan State has played with it, since it's similar to what they do normally. Anyway I thought it'd be fun to get into it now, so we'll have it to reference later.

Resources/hat tips:

  1. Rufio of Cleveland Browns SBNation blog Dawgs by Nature.
  2. Matthew Brophy's incomparable series on Alabama's D: part i, part ii, part iii, and his "Rip/Liz" video.
  3. Eleven Warriors' Kyle Jones's film study
  4. Ricky Muncie of Crimson Tide SBNation blog Roll Bama Roll
  5. Chris Brown, of course. Of course.
  6. Pre-emptive thanks to actual football coaches who post in the comments and point out where I got something wrong or over-simplified.

I'm Not a Coach Disclaimer

I'm not a football coach. I'm a guy on the internet who read a lot about football.

Basics of One-High Defenses

Cover 3 is probably the most basic defense in existence. It is the defense you learn on Day 1 as a high school freshman, if not before. At that level it is a "go to this spot and then find work" scheme, past that there are techniques coaches teach to cover the gaps. Here are the two basic versions that Saban uses against standard 2x2 formations:

alabamacoverages

If you picked up on the fact that "Liz" and "Rip" begin with the same letters as "left" and "right" (or you know your port and starboard colors) you have my permission to eat a cookie.

Joe Paterno used variations of this (Rip is very close to his base defense*) since the Chatelperronian, and like Neanderthal toolkits it only looks crude until you see it in the hands of a master.**

Some things to know that we'll use later:

  • The receiver numbering system is the same as in Quarters: start from the sideline and work your way in until you're at the center. It's where they are at the snap, not before, in case motion messed with that.
  • The path you take to your zone matters a great deal. Note how guys running toward their zones are actually going through weak points in the coverage. This is for "routing" purposes: if you're there a receiver can't be.

The latter is true for all zone defenses, but it's a stress point for Cover 3 because the holes in the zone are places the offense can attack either quickly (7-9 yards downfield in the seam) or easily (deep downfield once the free safety has committed). Cover 3 coaches teach defenders to be in the way so receivers have to re-route to covered places.

The tradeoff is natural coverage strength to the middle of the field, to the detriment of the flats—if you've ever watched an NFL defense that seems to constantly be tackling fullbacks squirting out of the backfield, that's why.

The problem with Cover 3 is the same problem with Cover 2: those frikkity vertical routes:

cover3v4verts

cover3vpoles

The problem remains with pretty much any set of routes that stem from a vertical release.

The old-fashioned answer to this is play more man defense, and certainly Cover 1 (example diagram) is a complementary coverage to any Cover 3 team. In Cov1, aka "Man Free" defense, corners stay on the receivers, the erstwhile "curl/flat" guys stay on the #2's, and the middle linebacker over the RB takes the RB.

But if you're playing man-to-man defense, you'd better have men who can win their battles 97%+ of the time against theirs. If you need to activate that free safety to double up a dangerman, now you're giving up "front"—how many defenders are participating in your run fits, and once it's not an 8-man front anymore you're weak against the run. Offenses will also use rub routes, or exploit matchups, e.g. have a quick slot receiver sprint across the formation until he loses the linebacker trying to keep up.

These were problems for Saban to a much greater degree when he was dealing with the kind of talent the Cleveland Browns drafted during his DC days. By the time he got to MSU he already had his Rip and Liz and his Cov1 amalgamated into a hybrid scheme he called "pattern matching."

[After the jump]

-----------------------------

* The Paterno-era "Hero", and "Sam" in the linked diagram were early examples of hybrid space players, and the zone-blitzing 8-man front it spawned was the basis of Rocky Long's 3-3-5 defense.

** …who discovered children were being sexually abused in his locker room and didn't tell the police because football reasons.

-----------------------------

Pattern Matching

Shown here is the left side of the "Liz" variant:

Pattern matching

------------------------

So all of this is kind of like Quarters: you're numbering receivers, and watching if they go vertical. But you're reading the receiver in front of you (in Quarters they both read the #2) and matching your coverage to what he does (hence the name). Simplified: if your guy goes vertical or angled in, you play Cover 1, if he goes hard inside or outside you play Cover 3, passing off and catching whoever enters your zone.

The thing it's most like is basketball zones, where you stay with your man until he crosses another defender, and you pass him off to that guy.

Benefits for a Modern (Spread) World

In an age when at least half of anybody's schedule is bound to be spread-to-run teams, and/or high-tempo, you need a good-for-all-seasons defensive front that doesn't take long to call and won't get torn to bits if you get stuck in that personnel.

To get the playcall in, the guys need only know if it's "Liz" or "Rip" to play relatively sound, and any time you get beyond that can be put to calling an atypical formation or a more interesting blitz. It's also a pretty balanced defense; the differences between the WLB and SAM have diminished as the SAM ends up over a slot as often as a TE.

The pattern matching looks and acts a lot like a Cover 1, except if a quarterback reads man and throws a drag, there's a guy who just got an "UNDER!" call suddenly coming down to intercept. Or if he reads Cover 3 and is attacking it with verts, there's a flat defender peeling off of that to run up the seam. And remember they can still throw a Cover 2 into the mix. A responsible quarterback could easily get frustrated trying to find spots to throw it, the more because the stuff he's used to isn't there. Quarterbacks without confidence are death to passing offense.

Special Personnel Requirements

Alabama's defense isn't much different from Penn State's back in the day, or Dick LeBeau's Pittsburgh defenses that informed both—lots of zone blitzes from shifting fronts. But more zone blitzes mean more generalized talent, since odd guys are taking on blockers or dropping into coverage more often.

To combat the spread Saban now uses hybrid space players as OLBs, and recruits safeties who can play in the box as well as range a deep zone. To give that Liz/Rip flexibility both safeties have to have Hero characteristics, and both OLBs need to be able to cover in space as well as blitz. The three down linemen have to be good at disrupting and even two-gapping to make sure the offense can't punish these lighter edge defenders—Saban tends to blow through the lives and livelihoods of a lot of DL to find a select few who can create the matchups he wants inside.

In Practice

Let's grab that first play from the 2012 Disaster in Dallas I put at the top of the article:

Michigan's throwing a deep slant off play-action, a good idea against this defense for reasons that become kind of apparent when you think about what everyone's assignments are. Alabama has a "Liz" call on:

pattern matching IV

And here's what they're thinking (click for bigger so you can actually read it):

pattern matching IV drawn

Note that this was a "Liz" call from a cover 2 look. The SAM ran in and the safety ran down right before the snap to reveal the coverage.

Despite the right read and a good throw this play was broken up by Dee Milliner, the right side cornerback, in part because Gardner's route was too skinny (he needed to cut in harder to get separation), mostly because Dee Milliner was worth way more than whatever Alabama was paying him.

I like the play design because it attacks exactly what Alabama did, or would have if Gardner's route had given Milliner an "Under" read (and Dee didn't see his curl/flat defender was tied up with the FB). The play-action off something Michigan plausibly ran (remember this was the first offensive play of 2012) forced a run reaction and made space against an excellent defense for any route in the area the WLB was supposed to cover. The bubble action to the strongside also got a guy open in the flat, which is pretty common against this defense (the SS and CB would be responsible for closing it down after a minimal gain).

How to Beat It

A few years later Urban Meyer would attack it virtually the same way: manball play-action with a spread to draw down the linebackers and commit the curl/flat defender. What made Ohio State's attack so devastating was the threat of the running game kept stressing the linebackers to react to it. This in turn stressed the cornerbacks—not Dee Milliner anymore. Ohio State did what they did to everyone who tried to play the run aggressively: test that outside man coverage with their dangerous downfield threats. They were doing so much that the defense began to expect it, and were unprepared for a wrinkle as simple as the outside receiver coming down for a crack block.

Then came the play Ohio State fans put on a t-shirt. It was remarkably similar to the run Michigan showed in play-action in 2012 (also the one play that worked for Michigan in the Spring Game), and they got the same defensive playcall: "Liz."

OSU fans labeled this "Big Ten Speed" but what beat it was old-fashioned manball (run from a spread look) overwhelming those hybrid defenders when they pinched down. One difference was the X receiver cracked down on the "backer", who was attacking the running back coming to his side and, because OSU had been sending that guy deep all day, was not expecting to be sideswiped. The cornerback did his job by attacking the edge and forcing the cutback.

But here's the coup de grace: notice the pulling guard got to his hole, and the Mike linebacker get erased by the X receiver's block too. That means the puller can move on to the last guy: the free safety coming up to fill from way back. Elliott would run away from defenders after, but what sprung him was a muscular attack on lighter hybrids.

Ultimately, like every defense, pattern matching is only as good as your players.

Comments

Space Coyote

April 29th, 2015 at 10:26 AM ^

I don't really have much to add, just wanted to say this covered things pretty dang well. Like you said, you have to have the players for the scheme to work. This scheme is sound, it's also complex (which can cause people to think more than react), and when the offenses starts dictating, just like any other defense, you can start finding gaps.

Good work. This was a good read. I really like that your Rip/Liz diagram shows the path to the zones, those are small details that often go unnoticed.

Space Coyote

April 29th, 2015 at 11:21 AM ^

So he deserves the praise and I like the work he does. Trust me, if there was something that stuck out to me in this post as needing corrected, I would have done it as I've done with a couple of Seth's previous posts. I've never had any issue pointing out things I disagree with Seth about, and Seth's rarely had issue with it for what it is (additional insight, difference in opinion, different perspective), which I appreciate and respect.

The only issue I have with some of what Seth says and some of the other main posters say are the pot shots at Hoke and Co that I find unnecessary (not in this post, which is nice); but frankly, it's not worth going down that path again, particularly because he's not coaching at Michigan anymore.

Space Coyote

April 29th, 2015 at 11:25 AM ^

The communication. I talked the past couple years about the issue with young safeties, and it boils down to all the communication responsibilities. Pre-snap and post-snap. These guys are constantly communicating, and diagnosing the plays and doing your assignment and communicating others their assignment is a lot of work and something that often goes overlooked because a guy is an athlete but is making mistakes back there. It's hard, it's a lot of responsibility for young guys. That's something that I don't think gets enough focus when people talk about safeties having problems.

Seth

April 29th, 2015 at 11:42 AM ^

When I got to talk to Marcus Ray for a bit at one of Marlin's events last year I asked him about young safeties and communicating and he was adamant that it's not about youth: football at his level forces kids to grow up early, and there are guys who are confident and communicative and guys who are hesitant and will always be that way, and sometimes there are guys who are way too aggressive but that's rare. He was way more down on Jarrod Wilson than anyone on MGoBlog. But he also brought up Jamar Adams as an example of the right kind of guy and I'm like "they're exactly the same player, down to the uniform number!"

Space Coyote

April 29th, 2015 at 11:44 AM ^

That's why a guy like Peppers can step in there and take over the job, and partially why I think WIlson was in the doghouse with Hoke's staff at times. But there is also a confidence with knowing you're right, and that comes with experience in my mind. So I disagree that youth doesn't play a factor in it. I think it's more a combination of things.

Magnus

April 29th, 2015 at 12:01 PM ^

I think communication is very key up the middle of your offense and defense, so QB, C, ILB, S, etc. are important communicators. They communicate to both sides of the field, and they see things from both sides of the field. I don't think it's a coincidence that it seems to take time to find a good center, a good middle linebacker, a good quarterback, etc. Most of those guys aren't instantaneously ready for those jobs, and you tend to see more experienced guys rise to the top.

That's why you can play Mason Cole at LT but not C, why you have Brandent Englemon winning the starting safety job, etc. There are exceptions (young quarterbacks, young safeties, etc.) but generally, the older guys are more confident and more mature.

Magnus

April 29th, 2015 at 12:01 PM ^

I think communication is very key up the middle of your offense and defense, so QB, C, ILB, S, etc. are important communicators. They communicate to both sides of the field, and they see things from both sides of the field. I don't think it's a coincidence that it seems to take time to find a good center, a good middle linebacker, a good quarterback, etc. Most of those guys aren't instantaneously ready for those jobs, and you tend to see more experienced guys rise to the top.

That's why you can play Mason Cole at LT but not C, why you have Brandent Englemon winning the starting safety job, etc. There are exceptions (young quarterbacks, young safeties, etc.) but generally, the older guys are more confident and more mature.

dragonchild

April 29th, 2015 at 10:57 AM ^

This is all pretty involved as an implemented scheme, but rep it and the reads/communication can be done very quickly with practice.  But HSPs don't grow on trees.  You can find guys who fit the mold, but these are inherently difficult tasks, football at a very high level.  I'll stay away from the DL talk as I've always been a space guy, but as discussed ad nauseum here, the HSPs need the speed to get to their spots but enough power to blitz and defend the run.  Teams know how to stop the spread, in theory.  The tough part is finding the players that can do it.  I don't think MSU's D was a gimmick or "too aggressive" or unsustainable so much as they lost the guys who could execuate the very demanding tasks, especially safety.  Well, maybe that's how it was unusustainable.

This is the value of Peppers as a FS, and why I've been screaming for Michigan to put him there since he arrived on campus.  As a corner he could be a lockdown guy who could take pressure off the safety, but the safety is still targeted by spread offenses.  Why do that when he could BE the safety and handle what spread offenses will throw at him?

The focus is on the offense, but next season's defense has the potential to be very, very good.  Not only are they experienced, the 2014 defense was unfortunately a mix of players that didn't fit that well.  They were good at some things (Countess at zone, JMFR at attacking the edge) and I'm big fans of 'em all, but cobbling together a cohesive defense with all these mismatched specialists was probably the defense's biggest challenge last season.  With Peppers and Morgan back, oh man. . . If we can just get some decent DE play, Michigan will be very tough to score on.

Space Coyote

April 29th, 2015 at 11:38 AM ^

First, you have to remember, when an offense dictates terms, it can torch any defense. There just isn't a lot you can do there scheme wise.

Second, C. Jones was 18/35 for 243 yards, 1 TD and 1 INT and also lost a fumble.

Third, Alabama started the game in a 2-high safety look to try to keep everything in front of them. The goal was to prevent OSU from getting big plays and force Cardale to actually string together a lot of short and intermediate plays. The issue was that he kept on doing it, whether picking up 3rd downs with his legs scrambling, or the fact that the run game made it difficult to keep two safeties deep. So then Bama switched back to the single-high look, and that's when OSU started getting big gains. But, in the 2nd half (after the switch), Bama also forced 5 punts on 7 drives, it's just on the two TD drives they gave up big plays (the other TD was an INT return).

bronxblue

April 29th, 2015 at 12:24 PM ^

Nice breakdown. You could tell this year Alabama had talent issues on defense because teams were able to take advantage of favorable matchups way more than in years past. There are always holes in defenses, but in years past you had athletes in the secondary who could clean them up.

LBSS

April 29th, 2015 at 12:48 PM ^

Thanks for the writeup Seth, pieces like this are a big part of why I love MGoBlog so much. Especially in the football-starved offseason.

That said, MUST DESTROY AUTOPLAY AD WITH NOISE. KILL IT WITH FIRE!

Reader71

April 29th, 2015 at 2:35 PM ^

Great work! Defensive football, outside of the very basics, is alien to me. I love, love, love that the blog can teach this old dog new tricks. This is a huge compliment: your work has given me more ways to enjoy the game I love. Thank you sincerely.

wuzura

April 30th, 2015 at 10:02 AM ^

 working at home with Google! It's by-far the best job I've had. Last Wednesday I got a brand new BMW since getting a check for $6474 this - 4 weeks past. I began this 8-months ago and immediately was bringing home at least $77 per hour. I work through this link, go to tech tab for work detail.

 
----------------------> http://www.netjob80.com