Picture Pages: UMass Counter Reconsideration

Submitted by Brian on September 23rd, 2010 at 5:09 PM

Normally in these posts I've noticed something or understood something or am trying to explain something. No so much on this one. I grabbed this because it's a play where it seems like four different things could go right or wrong to turn it into a better or worse play than it ended up being, which is a five-yard run on first and ten.

It's the beginning of UMass's first drive of the second half; Michigan has just put up three touchdowns in four minutes of game time to surge into an 11-point lead. UMass starts off in an Ace formation with twin WRs to one side of the formation and twin TEs to the other. Banks and Kovacs are to the top of the screen, RVB and Gordon to the bottom:


At the snap the tailback starts running to the left side of the line. Martin gets under the center and starts pushing him back. Michigan linebackers start stepping to the playside, and Kovacs starts burrowing into the line:


A moment later the action has continued but the tailback has started coming back to the right. The move left was a feint; this is a counter. It's pulled the Michigan linebackers to the right:


A couple things on the above frame: it certainly looks like Martin is in a position to tackle in the A gap if the play ends up there, and he has gotten into a position where he is useful. But: I +1ed him on this play I shouldn't have since the defense is trying to seal him to one side and has. I don't think this is a negative play since he hasn't gotten blown off the line or anything, but it's not a win for the D.

The confusing thing about the linebacker play to me is if Martin is going to go to one side when the play starts it would make sense for the MLB to immediately go the other. This is "making the nose tackle right" if the nose can cut off that gap, which it certainly looks like Martin has.

HOWEVA, in the UFR comments, Steve Sharik made a point I hadn't ever thought of: when you pull linemen you are putting more blockers in a gap than there are defenders if everyone just takes a gap. So it makes some sense that Michigan LBs were in a read-and-react mode against UConn, which was pulling linemen all over the place. ND also makes heavy use of pulls, and frankly I'd be surprised if Michigan bothered to change their gameplan for UMass. The Minutemen did their share of pulling, anway. So it's more complicated than that. If Ezeh hammered it up behind Martin on this play and the opponents were pulling around into Martin's gap they would find a lot of space.

By the next frame the tailback has taken the handoff and the defense realizes it's a counter. Banks and Kovacs are engaged in a shoving match with the OL on the right side of the line that is going nowhere, which is usually a +0.5 in my book. It won't be here, as we'll see.

Mouton and Ezeh are free, though Ezeh is about to get a guy peeling off Martin:


Mouton sitting in that gap dissuades the RB from trying to hit it up; a step later he's still moving outside as the center attempts to get out on Ezeh:


Ezeh gets his face across the blocker and Martin is fighting through his guy; no place to go (except maybe cut behind Martin for a big gainer, but RVB seems like he'll shut that down):


A moment later this is obvious. Mouton is nearing the second TE, giving M three defenders on three blockers and Floyd ready to handle a bounce:


Here is some confusion. The RB fakes outside…


…which causes Mouton to hop outside the TE and surprises Floyd; Kovacs has finally yielded to the physics of his leetle body, giving the tailback a crease:


Kovacs and Floyd close the crease down, but it's six yards:


Video; watch how the tailback's little juke outside gives him the crease:

I misidentified this play as an inside zone, which it kind of is but that does not take into account the counter action. These are the things I think about it after some consideration:

  • I should not have given Martin a plus nor Ezeh a minus. Both plays are fine. Ezeh did react in time to get across the center and Martin cut off his gap, then fought back through his blocker in time to help close down the play's intended hole. But Martin did not force a cutback—that was the play design—and didn't help on the tackle.
  • On a later edition of this same play Banks should have been minused for flowing down the line too hard and opening up space for the tailback, but I still think Ezeh is slow to read and react, thus allowing him to be "blocked" by a center who's falling to the ground because of Martin's violent burst into the guard; I'd rather run the D like Martin is going to able to slant into the gap he wants against most teams and watch the cutbacks. Kovacs's ability to pursue hard when he has a gap to one side of Mouton to fill, then redirect and make a tackle when the RB cuts inside of Mouton is impressively aware.
  • This was the story of Banks's day: I'm not doing much but I'm not going backwards either.
  • Kovacs is small and this hurt him here as he tried to stand up to blockers, but really if it takes this long for help to arrive it's not his fault.
  • Mouton and Floyd are confused when it comes to edge play; here it seems like Mouton should make sure he bounces the RB to Floyd and instead he hops outside, creating a gap that the RB can use. If he bounces it to Floyd he should be able to tackle; if Floyd expects that Mouton will funnel it inside he should be able to tackle. Neither happens. More of this deficiency can be seen in the earlier Mouton picture pages and the easy touchdown UMass scored when Floyd let the RB outside.
  • I think I would prefer a chancier scheme that said "aww, the hell with it" and blasted linebackers into open gaps once they read run. If Michigan's going to get ground like they did attempting to play read and react—a lot of should-be-zero-yard runs like this one turned into four or six—they're going to give up a lot of drives like we saw Saturday. Getting those zero- and negative-yard plays on early downs seems more likely to get the defense off the field. This will put more pressure on the safeties when this doesn't work out, but they seem like good tacklers and guys who take good angles.



September 23rd, 2010 at 5:26 PM ^

Look at how far we have come:

they seem like good tacklers and guys who take good angles.

Who could have said that for our safeties past two years. Just for this, I going to give Gerg some credit - and if they just keep doing this with some level of competence, I will take my chances against any team as long as our offense remains Denardastic!


September 23rd, 2010 at 6:16 PM ^

but at times, especially on the play outlined above, it appears that these guys just don't trust each other.  If fundamentally Mouton knows that he is supposed to keep inside contain and force the RB into Floyd, is it that he just loses his mind and doesn't remember that Floyd will be there or is it that he doesn't trust what's behind him.  It's so strange.


September 23rd, 2010 at 6:24 PM ^

Isn't the definition of contain to be on the outside of the field as to make the runner cut back toward more players?

To answer your question, though, I think the sooner you contain a player, the better because that means there are more players to tackle him. If he goes inside to fill the gap, that leaves Floyd 1-on-1 with the RB, which is a match up I wouldn't want to see happen.

ND Sux

September 24th, 2010 at 7:44 AM ^

"Mouton should make sure he bounces the RB to Floyd and instead he hops outside, creating a gap that the RB can use"

Most of the week we've been talking about losing outside contain, now the suggestion is force the RB outside and into a 1-on-1 with the DB?  I like our chances better with Mouton keeping outside contain, forcing RB back inside where there's a logjam of defenders.  Mouton is +1 on this play, IMO.  I realize the guy got six yards, but only b/c he earned them. 

100% agree that we should attack the gaps a bit more aggressively, or at least MIX IT UP some to keep the offense guessing.  No gain or neg gets the defense off the field more often. 


September 23rd, 2010 at 6:19 PM ^

I think Ezeh should get a -1. At pic 4 its clear the play is a counter but by pic 8 he is STILL 3yds from the LOS and getting engaged by the OL. I think by pic 8 he should have been at the LOS trying to make a play. If he where there that would have made flloyd and moutons decision that much easier.


I also think Martin hasn't done anything here to warrant a +1. looks like a stalemate for the NT to me.


September 23rd, 2010 at 6:30 PM ^

Should he be 0 yards from the LOS and try to run through the linemen? If he reads run right and reacts and gets beat on a RPS call, he gets a -1, but if he reads run right, but hesitates to make sure its not a counter, he gets a -1? The play was designed to get him out of position and get the guard on him. What should he have done to not earn a -1?


September 23rd, 2010 at 11:32 PM ^

I didn't say ezeh should blow this play up for a TFL. Just that from pics 4-8 he can't get closer to 3yds from the LOS. In that same time Mouton covers a lot more real estate. In fact he overcommits twice in the same play...lots of movement.

I just think that Ezeh should have tried to hit that gap created by Martins double team. He wouldn't have made the tackle due to the OL pealing off Martin but IMO, just the threat of him approaching the line would have cause the RB to bounce this play outside...and into Flloyd and Mouton for maybe 1yd.

Blue in Yarmouth

September 24th, 2010 at 7:54 AM ^

That is the silliest thing I have ever heard. By charging to the line of scrimmage on this play Ezeh would have simply run himself out of the play or got swallowed in the garbage at the line. He did the right thing by shuffling down the line to make a tackle.


September 23rd, 2010 at 6:21 PM ^

They ran this play all day. I wonder how many plays Martin got +1'd for and Ezeh got -1'd for in similar situations. It might explain why Martin was so astronomically high and Ezeh might be pretty close to even if there enough plays like this.

The scary thing is that this is the same play that EMU ran down our throat last year, the difference being that UMass was able to bounce it outside more, while EMU had to cut it up for a gain of 4-8 yards.


September 23rd, 2010 at 6:49 PM ^

I know this is pretty much a repeat of what Brian is saying, but if Martin or Banks can not beat a double team here, it seems by design that no one even has a chance to engage the runner until he has passed the LOS with momemtum.  I just don't see how this scheme can possibly not yield 3-5 yards minimum. 
  And just another thought, if Mouton does blow up into that gap and the RB bounces outside, is Floyd 1 on 1 against a RB in space a good option for Michigan? Maybe with UMASS, but if that's John Clay, yikes.

Space Coyote

September 23rd, 2010 at 9:45 PM ^

Then it's a pretty good option to have.  If he tries to cut it back in Michigan would have lots of help there.  I'd say it would probably be about a 2 yard gain at that point, so, while it's not the best option to have, it's not bad.

I also see what Mouton was trying to do here.  There's a good chance he had C-gap in a run that direction and didn't want to get caught inside.  I'm not sure if the corners were playing man or zone, but if it's zone then it's at least understandable why Mouton would jump outside.  I still think he should have just slammed the B-gap though, as it was a counter with no action in the direction of the ball.  This would end up forcing the RB to make a cut in the back field and running him into the outside contain.


September 23rd, 2010 at 6:55 PM ^

Given personnel, the idea that this defense is playing bend but don't break seems rational.  If so, you accept some long time-consuming drives, waiting for the offense to make a mistake (as we saw UConn and ND do).  UMass executed well, so they moved the ball.   I'm content to accept that outcome if it means avoiding the big play. FGs, not TDs! The faith in the safties taking good angles seems premature to me.

Amp up the aggressiveness when the risk is low (e.g. short field... or when your D is getting gashed anyway)  or when your opponent seems adept at executing (like UMass).  I don't have great faith in most teams in the Big10 doing that;  Bad OL or inexperienced QBs are abundant, so execution will be spotty. The defensive strategy will work well in that case. 

Against MSU and Wisconsin, who have good QBs and seems to execute pretty well, I'd be more inclined to shift the gameplan a bit towards aggression.  Against Purdue, Illinois, PSU and Iowa, I'd sit back and let them shoot themselves in the foot.

Zone Left

September 23rd, 2010 at 7:20 PM ^

I'm with you here. Most college teams are not capable of consistently executing quality offense. A team with severe personnel deficiencies and what appears to be a very dynamic offense would seem to be well served forcing college offenses to methodically move down the field, forcing one or two mistakes per game, and accepting some points in exchange for never being run off of the field a la the much more pressure heavy team from 2008. It's frustrating to watch, but it is almost certainly better than putting this secondary on an island very often.
<br>On another note, I'm interested to see how Big 10 teams play Martin. A solid center can probably use his slashing against him and create some significant holes in the middle.

True Blue In Ohio

September 23rd, 2010 at 7:01 PM ^

I could not agree more with the last bullet point (or arrow point to be exact).  We need to shoot a line backer from the interior into the gap when reading the run.  This can do two things:

1)  Martin gets double teamed all the time.  The O-Line will be thinking of trying to somehow to contain the linebacker shooting the gap.  A miracle might happen and the linebacker shoots up the right gap and makes a tackle for a two yard loss or the linebacker at least delays the execution of the play and help arrives minimizing the running back falling forward for one or two yards.  The latter most likely applies in our defense, but a different element is needed to our scheme.

2)  If your shooting guys from every gap at random, the offense will adjust and start looking for different zone blitzes.  However, it gets them thinking and reacting to what our defense is doing.  It seems our defense is always reacting to the offenses.  We just might be very inept on defense, I hope not.

steve sharik

September 23rd, 2010 at 8:00 PM ^

...once again shows why I prefer my ILBs to key guards first, then backs.  Our ILBs are keying backs, and they have different gaps depending on flow.  What it looks like is the following:

  • Ends have C gap.
  • The Quick and the Will ("The Quick and the Dead?") scrape to the edge on flow to, empty A gap (the one not occupied by The Hulk) on flow away. 
  • The Mike has B gap to flow.
  • Kovacs is playing D gap (b/w the TE and Wing).
  • The RB intial flow action puts the ILBs toward the incorrect gap, and the counter makes them late to the party.
  • The OL is blocking inside zone but with a wrinkle: the backside T, TE, Wing base block, turning this into a designed inside zone windback/iso play into the backside B gap, but with no lead blocker.  UMass has done an excellent job scouting our defense, noticing that when flow goes one way, we are short a gap on the backside, since only 1 stack backer is left to play both the backside A and B gap.  Yes, we have 8 defenders for 8 gaps, but when the DE is in the C gap and the playside stack backer has edge, they're both in the same gap! Yipes, we're one gap short.
  • If the ILBs had been keying guards, Mouton would've seen open window in that backside B gap, slow-shuffled with the initial RB flow, then when the RB countered to that backside, shot the B gap and nailed it for worst-case-scenario 1 yard gain.


September 23rd, 2010 at 8:46 PM ^

What's the downside of moving the DEs out of their 5-techniques and into a 4-technique, then having them play a 2-gap system where they attack either the B or C gap depending on the flow of the lineman?  Wouldn't that eliminate your problem of being outnumbered on the backside of the line? 

Is it a personnel issue (i.e. RVB, Banks aren't big/strong enough to play a 2-gap system)?  Does it open up other lanes in this stack formation?


September 24th, 2010 at 11:39 AM ^

Is this a case of our LBs trying to do too much and UMass taking advantage of that with the counter?  Since we're slanting our DLine almost every play, in order to have gap integrity when the OLine pulls don't our LBs have to read OLine so that we don't have two players in one gap and no one in the new gaps created from the pulling linemen?  I'm no coach but I don't think I've heard of a one gap defense where the LBs key on the RBs first, is there even an advantage to doing that?


September 23rd, 2010 at 8:28 PM ^

I am making the following analysis of this play based on the assumption Ezeh reads the guard in front of him (to the left of Martin) through to the action of the nearest back (ball carrier), Roh is reading the guard to Martin's right and through to the RB and Mouton is reading the same keys as Ezeh.

Ezeh: has Guard action down (doubling Martin). Because Martin has already slanted to his right a nice hole opens up in front of Ezeh. Two things- Ezeh's keys tell him "run" and his first step should be forward and right; however, since Martin is slanting right his first step should be forward into the hole created by Martin's stunt and the double team block action.

Roh: has Guard action at him (reach block), his first step should be right and forward attacking the guard's inside shoulder. RVB's appears to have outside contain to this side.

Mouton: has same read as Ezeh. Based on Kovacs' action it appears the defensive play call dictates Mouton has outside contain.

I think Mouton play is correct, although I would have liked to see him be more aggresive in taking on the TE.

If Ezeh steps into the hole the guard has no chance to block him. He either makes a TFL or the RB has to bounce outside right now to a waiting Mouton.

We are lucky the ball carrier's vision and feet aren't elite because Roh's missteps create a huge cut back lane.

OK, I said all that to say Ezeh could have played this better, but it was a good play by the offense.

EDIT: I type way to slow. Didn't see the SS's ILB read post.

Deep Under Cover

September 23rd, 2010 at 8:54 PM ^

One of the biggest problems we have is when RVB or Banks slant they slant so hard the Olineman only needs to use his momentum and push him in that direction creating a huge hole behind him.   Why our DTs slant that hard makes no sense.  The slant is to get them in the gap they are slanting to and restrict it.  But our guys allow their momentum to get them blocked.


September 23rd, 2010 at 8:55 PM ^

I've felt this way since the first analysis of the defense against UMass went up.  If the LBs are not going to consistently react well to these plays I would rather see them aggressively shooting into gaps created by the defensive lineman and make at least some zero-negative stops.  It would cut down on the ability of the RBs to watch the line and make decisions.  There have been way too many of these slow-developing runs go for 5-10 yards.  Make the offense commit earlier - they will break off some plays but it will be less consistent and we'll have a chance for some big plays from our defense.


September 23rd, 2010 at 8:59 PM ^

I'm OK with this however they played it. I think we'll see more run blitzing in the B10 season
<br>What I could not stomach was repeatedly getting destroyed on the edge in terms of contain
<br>Our B-B-Don't-B D can't do that


September 23rd, 2010 at 10:12 PM ^

Nice analysis Brian.

I kind of agree with the last point about sending LBs into the gaps when it is clearly a run, or at least sending one guy.  While I understand the logic that predicts college offenses will make enough mistakes to slow themselves down if you just play bend-but-don't-break, that -3/4 yard stop on first down also tends to screw with offenses, especially ones like UMass that execute a couple of plays very well.  We saw them give up on a couple of third-and-longs that seemed manageable, and the safeties and DBs should be able to cover any gaps that are created with the lost LB.  Plus, Ezeh and Mouton seem far more dangerous as straight-ahead tacklers than a guys who try to flow across a line and maintain containment, as they seem to be lost in the wash far too often. 


September 23rd, 2010 at 11:14 PM ^

For the life of me, I do NOT know why we have Kovac crashing inside the EOL, only to get pinned in by the end or slot.  Even if the entire rest of the line is slanting to the right (in this case), I would have Kovacs maintain outside leverage while the LB would have responsibility for the C gap.

Kovacs isn't big enough/Barwisized to fight off a seal block from a bigger tight end or a slot who has position/leverage on him.


September 24th, 2010 at 9:13 AM ^

Brian quoted a scout a few posts ago saying the problem he has with GERGs defense is too passive...most people (myself included at the time) probably took this as "does not blitz alot" or "3 man rushes argh" but the "i'd rather them shoot the gap then read and react" seems this fits into that comment's thesis. LBs and the box safety waiting for the play to come to them instead of shooting the gaps or forcing the play. I don't know if its a trust issue with the LBs/safety (be it their talent level to effectively shoot the gaps, their ability to read the run correctly and shoot gaps instead of just shooting gaps nonstop) or if its schematic. It sure looks intentional but it would be interesting to see what the scheme would be for the LBs if we had better LBs.... It seems Mouton and Ezeh have the talent and skill set to be effective at the aggressive force the play idea but don't have the brains to effectively read it making it too risky in the event that its a PA or slow developing pass which would == burn.


September 24th, 2010 at 11:45 AM ^

It seems rare that anyone but Martin, although Roh did well at ND, ever just beats their man on a run play.  If Banks, RVB, Sagesse, Roh, and the rest of the crew could make a couple more plays a game (as a group, not each)  instead of just occupying blockers it would make a huge difference.