press coverage

NOTE: I am looking to purchase a pair of tickets to Northwestern. If you've got a couple extras email me to discharge built-up beveled guilt.

Power vs zone read. A couple weeks ago I wondered if running a bunch of power had opened up the zone read again or if it was just an effect of playing Bob Diaco and Ron English. Frequent correspondent Tyler Sellhorn provides some insight:

WLBs are the bugaboo defender for the power play (double team frontside = WLB difficult to block/unblocked).  They are coached to hit the window created by the inside OL stepping to the double.  Playside combos of inside zone are difficult to distinguish from straight doubles. 

The best defense vs. ZR is to exchange gaps between the DE and WLB (you already know this).  Therefore, these two plays in concert screw with the WLB assignment-wise from a gameplan standpoint.  Gap-exchange weakside means that the free defender versus power is no longer paying any attention to the RB running said power.  Leaving the DE to defend the ZR by his lonesome, though, against DR...hell to pay.

Hope that enlightens.

God Bless,
Tyler Sellhorn

Since then we've seen San Diego State defend the zone read (and nothing else) ably and Minnesota defend nothing (and nothing else). A test of this theory will come against Northwestern, which may have given up 38 to Illinois but held the Illini rushing game to just 82 yards. Sacks factor in but even without those Scheelhaase and company managed just 3.1 YPC.

They also gave up 400 yards passing, so don't get too frightened.

Stretches versus outside zone. I've been using the two terms interchangeably, which Tyler suggests is mistaking rectangles for squares:

…the zone stretch, the various sweeps (including QB sweeps), pin/pull, and when the G tries to "log" the end/OLB on Down G, the Dash (frontside zone read) all try to accomplish the same thing: circle the defense and (usually) carry the ball between the numbers and the sideline.  

What I am getting at is that you have made the statement that there have been zero stretches and it feels like you are implying that M is not trying to get outside when you make that statement.  There are lots of ways to get the same thing as "stretch" conceptually, and Borges is trying to fit the concept into what he already has experience calling and know what to call when.  For example, QB sweep was the first call against WMU. 

So yeah, you keep harping on "zero stretches" when there have been plenty of attempts to get the ball outside, but M is using different blocking schemes to do the same thing.  You just need to be clearer about what you are trying to say in regards to this: we should be running outside more or we should be using stretch to run outside.  That is the distinction I am encouraging you to make.

Tyler Sellhorn

Right, then: I'd like to see more outside zone blocking from Michigan because they're pretty good at it and don't seem particularly good at getting outside with pin and pull stuff or toss sweeps.

Advanced not looking at the ball. Chris Brown of Smart Football had a couple of things to add in re: Michigan's NOBODY CARES WHEN WR LOOKS FOR BALL coverage technique:

Saw your picture pages on Michigan DBs playing the fade and having success playing the man versus the ball. Thought you might find this of interest from Saban.

Basically if you are even with the WR, you play the ball. If the receiver looks over his inside shoulder you look back that way; if the WR turns his outside shoulder back you turn into the WR (toward the sideline) to play the back shoulder fade.

But if you're out of phase with the guy, ie trailing him, you don't turn back to find the ball because you never will and they'll catch it; you play the man and his hands and eyes. (I get the impression that this wasn't the case last year.)

From the photos I saw on your site the Michigan DBs are doing a good job playing the man, but that's because they aren't "in-phase" with the WRs. If the throw was better they'd probably be completing the fades. But you're closer to this stuff than I am; mostly wanted to pass along the Saban points.

So Michigan's trail technique seems born of necessity. Since they don't have Charles Woodson or Leon Hall back there the best they can do is go for the PBU. We've seen Blake Countess look for the ball because he's in better position a few times.

If Countess proves to be the real deal and Michigan can get a second corner at that level we may see more DBs look back for the ball. As it is the current technique is at least an excellent stopgap.  

A little outdated. This came in before the Minnesota game:

Brian,

Do you think Denard would be as effective a runner from the RB position as he is from the QB position? My gut says he would not be but can't explain why. I bring this up given his continued poor passing performance with some people clamoring for him to change positions.

Peter F

Denard wouldn't be as effective a runner because he excels in the space allowed by a spread formation. In a pro-style offense he would probably be too slight to be a tailback, at least full-time. He'd end up in the slot.

The main tactical innovation allowed by having your QB as a runner is it allows you to spread the field horizontally by adding more WRs without giving up the extra blocker. With the defense locked in on those slots—something the threat of the bubble screen enforces—a guy like Denard can pick and choose from big gaps that open up because the defense is stretched.

Handing it to a tailback without using the QB as a threat invites an unblocked guy through since there are fewer blockers in the area. Think of this like a power play: a 4-on-3 power play is more dangerous than a 5-on-4 because it's easier to find the open guy and there's more space. The shotgun provides the extra man by using the QB as a runner. That extra space means Denard can make yards by accelerating past tackles instead of breaking them.

Denard's still pretty good when things get tight, but the pounding would be worse if that was all he was doing.

Play action problems.

Brian, would like your view/analysis of Denard's play action fakes and the importance of these in the offense. It does not appear to me that Denard really sells the hand off as much as other QB's. I'll spare the comparison to Peyton Manning. A good play fake can open up zones in the secondary and give Denard more time to make his reads as the defense should be crashing on the running back. Or, is this less of an issue in a zone read offense since there is basically a play fake on the majority of plays.

It appears to me Borges likes to throw off play action and if the QB is not selling it, that might account for some of the pressured throws we have seen from Denard so far. (disclaimer about adjustment to learning a new offense a given)

Thanks,
Trueblueintexas

There are two entirely different playfakes Denard is executing. There's one from under center and one from the shotgun. It is possible that Denard's fakes from under center are not convincing, but I think the bigger problem is that the run game is not threatening. When you're averaging three yards a carry, safeties don't have to worry about your run game because it's not getting to them. I'll keep an eye out if we get more play action from the I-form later in the year. It's possible he's a problem there since he hasn't really practiced that skill.

The shotgun is a different matter. When Michigan goes play action from the shot gun it's either Denard stepping to the line or a zone read fake. Both are inherently convincing. In the first Denard is moving towards the LOS as the offense run blocks. In the second they are executing the mesh point exactly as they would on a running play. Unless the line is doing things that tip off the opponent there's no difference. The sheer number of hand-wavingly wide open dudes on shotgun PA should be sufficient evidence that Denard's doing just fine with his fakes there.

Linebacker blaming.

Brian,

I'm reading the SDSU preview and you say that Demens and Hawthorne have to get better at diagnosing plays quickly.  This appears to be a consistent theme with M linebackers over the last few years.  I would assume that this "skill" is probably the easiest to evaluate when recruiting high school players as HS offenses are pretty run heavy.  Did our coaches completely drop the ball in recruiting these guys or did they believe diagnosing plays is something that can be taught and, thus, focused more on the recruit's physical traits/potential? 

Thanks!

I'm not sure that skill is easy to evaluate because a lot of high school kids don't get much coaching and what they get is of debatable value. You might be able to detect a kid who just Gets It, but plenty of college-level athletes who look clueless early develop into excellent players with college coaching. Prescott Burgess and Shawn Crable are two examples in recent Michigan history.

In the case of Michigan's current starters, the Great Rodriguez Defensive Coaching Malpractice is probably more at fault than recruiting. The current LB crew has been coached by Jay Hopson, Greg Robinson, and Adam Braithwaite. Braithwaite has the best resume of all of those guys by virtue of not having one. They've also swung from one system to another and, in the case of Herron, Hawthorne, and Cam Gordon, from one position to another. If these guys weren't having trouble diagnosing plays that would warrant creating a golden idol resembling Mark Smith.

As it is I think they're doing as well as can be expected. Hopefully we'll see the improvement we never got under the GRDCM as the season progresses.

A few rows in front of me at the Western game was one of those guys who exasperatedly yells out a piece of football wisdom he's picked up over the years whenever he is affronted by its lack. His wisdom was "turn around for the ball," which he yelled at Herron a couple times and the cornerbacks a couple times.

I was with him, but then a funny thing happened: no one could complete a fly route on these mediocre corners. Here's everything I've got marked fly/go/fade (which I am totally inconsistent about) from the first two weeks:

Your score excluding the miscommunication: two legit pass interference penalties, one horsecrap call, one 26-yard completion to Michael Floyd, five incompletions. What's more, in each case save one pressure-forced incompletion and the two legit PI calls the corners are 1) there and 2) making a play on the ball.

That's seven out of nine legitimately good plays from the DBs on accurate deep balls. On all but one—the legit Avery PI—the corners were on an island as Mattison sent at least five. No bracket here. The Avery PI was a zone, the rest of it was man coverage, much of it press.

Michigan's press-ish coverage success in fly routes in 2011 including a game against Michael Floyd: 88%. The exception was virtually unstoppable and still drew a plus from the ol' softie who does these things. That's miraculous in last year's context. Hell, it's miraculous in a lot of contexts. How has this happened?

Michigan Press Coverage As Explained By Underpants Gnomes
IN STEREO

STEP 1: Line up a yard off the LOS with inside leverage.

STEP 2: When receiver releases outside, turn hips and run with him real fast.

STEP 3: NOBODY CARES WHEN RECEIVER LOOKS FOR BALL

STEP 4: NOBODY CARES!

STEP 5: When receiver reaches up for ball, punch him in the face.

OPTIONAL: grab his jersey a bit and get away with it
OPTIONAL: scream SHORYUKEN.

STEP 6: Profit: arm-waving motions indicating that the pass was incomplete.

OPTIONAL: shake head to indicate "no."
OPTIONAL: pick up horsecrap pass interference call.

Videos

Floyd on Floyd action:

Avery on Jones action:

Interesting Items

Why it works. That whole find-the-ball thing is hard. Todd Howard was coached to do it but always did it late, whipping his head around just in time to see the ball zing by. When you do that you've given yourself an even tougher job than the WR, who's been tracking the thing since it left the QB's hand. Lots can go wrong there. He can slow up and you bowl him over. He can slow up on a deliberately underthrown ball. He can slow, then extend a la Manningham. Or you can just not find the ball quickly enough.

In contrast, the shoryuken technique seems pretty easy. Focus on the WR's chest. When his arms go up, get your arms/head/body in between those arms. Faceguard the guy for bonus points. Net result: incompletion or spectacular Prothro-style catch. Mostly the former.

It's hard to get lost because you're following the WR's chest everywhere, and the only bomb you can't defend is the one that's just past your outstretched arms. That's hard to throw and hard to catch.

Gibson –8. Two games in I am a believer in Tony Gibson Was The Worst. These are the same guys as last year making these plays. Notre Dame clearly identified these fades as a weakness to exploit, especially in press coverage, but got little out of them. If you discount the Avery PI, on the eight fade attempts against press coverage opponents got 41 yards, just over five yards per attempt. Even if you count the Avery PI that hops up to 6.9 YPA—still worse than the NCAA average of 7.2 YPA.

Compare that to last year, when even doing something right meant you did something wrong:

Small sample size disclaimers apply, but Tony Gibson? The worst.

Downsides and low upsides. So this style of coverage seems pretty effective, obviously. There are two major downsides to my eyes:

1. Low upside. Since you are never looking for the ball you are highly unlikely to intercept it.
2. A tendency to pick up PI calls. Refs give you more leeway when you are looking for the ball. Bumping a guy with your back to the ball is always going to be an issue, but you can get away with "look and lean," as Spielman calls it.

I'm a little concerned about our corners' speed when asked to run real fast. Against Western Floyd gave up a yard or two of separation to a MAC receiver on his successful fly defense; in the second clip above it kind of feels like on a longer route Jones will pull away from Avery. Those are hypotheticals, though, and whatever limitations of Floyd and Avery have do not currently include a tendency to get burned deep.

This allows cool stuff. Michigan can press with one high safety because of this, which opens up the blitz possibilities that produce big plays. While the coverage style precludes big plays from the cornerbacks it allows them from other parts of the defense, and those big plays are bigger. What would you rather have, an interception 30 yards downfield or the quarterback fumbling the ball?

Tony Gibson. The worst!