After two nonconference games against teams not expected to finish in the top 25 in run defense, Michigan's running game hasn't produced the way we expected. You'd figure with two of the best three running backs in the country, three starters back from the two-time Joe (Sherrone) Moore-winning offensive line supplemented by two of the position's top transfers, and the hottest quarterback in the country, the Michigan run game should be steamrolling opponents. They're not.

There are several reasons for this, including…

1. ECU committed their entire defense to stopping the run, and Michigan was content to only play-action them to death as much as they needed to build a 30-point lead through the air.
2. UNLV committed their linebackers to stopping the run, and Michigan was content to only use intermediate passing on early downs as much as they needed to build a 28-point lead.
3. OTs Myles Hinton and Karsen Barnhart aren't grading out well, though they remain ahead of projected starters Trente Jones and LaDarius Henderson.
4. The RBs, especially Edwards, haven't been as patient in letting things develop as they were last year.

5. The WR blocking this year has been a major step down from what they got with Ronnie Bell (I'm going to keep saying I told you so!)
6. Except in a few circumstances Michigan has been loathe to use JJ McCarthy's legs to even the numbers, and both opponents have been taking advantage of this by overplaying the edge on (fake) zone reads.

But the one I want to address today is that Michigan has been using Stretch (aka Outside Zone) as more than a changeup this year, and they're not yet all that good at it, nor likely to become masters.

WHAT IS OUTSIDE ZONE?

The run play where you see all the blockers sliding sideways while the running back picks a gap. My "What Is" series is aging out but here's my 9-year-old explanation:

For the OL it starts with the same rules as inside zone: if you're "covered" by a defensive lineman you block that guy, if you're not you release to block a linebacker, and if you're lined up playside of a DL you combo then release. Then you block depending on what those defenders are trying to do: if they shoot inside you seal, if they slant away you shove and make them over-slant—they'll really have no choice but to try to ride you and stay playside. Then the RB picks his hole based on whichever block went the best.

[After THE JUMP: Michigan isn't made this way.]

WHY IS MICHIGAN RUNNING STRETCH IF THEY'RE BUILT FOR POWER?

The smaller, programmic reason is the NFL these days expects linemen to be able to run just about everything. To make Michigan's linemen more attractive to pro teams, and therefore make Michigan more attractive to the kinds of athletes who translate into awesome linemen, it behooves the program to demonstrate they can dominate with a variety of concepts.

However the major reason I think they're adding more Stretch to the mix is it's strong against the best way to defend Michigan's favorite run play: Duo.

Duo is a great base play for Michigan. They have heavy OL who can consistently move DTs with double-teams. They have very good RBs who can consistently put an LB in the wrong gap and get to a different one before the LB can react. This play was a Bo staple, a Harbaugh staple at Stanford, and has come back into favor in the last eight or so years as offenses found it effective against teams that use smaller DTs.

If you don't have those planet-sized DTs, you can still defeat Duo with aggression. That means firing your front seven defenders upfield before the doubles have a chance to blow the DTs back or the RB has a chance to put the LB in the wrong gap.

Outside Zone then is a good way to punish defenders who are overplaying Duo. We got a good example of that versus UNLV:

The Rebels here gave Michigan a light 3-3-5 look with a plan to go all ECU at the first sign of run action. Everybody fly upfield at a gap. I want you to particularly notice that the DTs are lined up to the left (from the offense's perspective) of center Drake Nugent and right tackle Karsen Barnhart.

Michigan snaps it and both of these DTs are moving upfield, sure, but they're also now on the back side of the play.

Nugent and Barnhart use that leverage to help their respective guards also get around their DTs. Suddenly this gap-sound defense is getting overwhelmed. The linebackers are flowing to replace, but because the DTs got scooped while trying to get upfield they're getting outflanked and Corum has a blue lane off Barnhart's butt.

If you recall back to the Indiana game that was way too close in 2015, that happened for this same reason. Mo Hurst and Will Henry were young DTs firing upfield to upset interior runs, and Indiana was punishing that behavior by running stretch around them.

I bet you this is why Michigan is running a lot of Stretch this year: If they can get good at it, they can punish the way defenses want to play Duo, the play that works because Michigan's OL are big hosses who can blow DTs downfield with double-teams and Michigan's RBs are lightning quick dudes who can force the unblocked linebacker behind that double to pick one side of it before the RB hops to the other gap.

However, it's got costs.

THE TECHNIQUE IS EXPENSIVE

By expensive I mean that Stretch takes a lot of program resources to get good at it, most especially the practice time to rep it enough that your linemen can anticipate all the little subtle ways defenses are going to react to it. Compared to other base running concepts, Stretch is highly reactive to the defense, which means you have to have everyone super-well drilled at all of the defense's reactions and the correct counterreactions to be consistently successful.

We saw the Wolverines bring this out last week against ECU for a decent gain, but only because Corum made a play on the safety. This play is a pretty good demonstration of what Michigan's stretch aptitudes. Right off the snap their center Nugent got one of those all-important reach blocks, getting himself playside of the nose tackle (#40) who lined up playside of him.

There's an awful lot going on here, but that's the point of Stretch. The linemen are letting the defenders move where they will and are supposed to just help them along the way while the RB looks for the best lane. The best way to get a good lane in stretch is to find a gap between a defender who is letting himself get pinned inside and one who is going to get too far outside. From the setup above it looks like things are developing nicely for an attack to either side of Keegan, since the NT got reached and sealed upfield by Nugent and DE #3 is engaged with the B gap very far away.

This lane falls apart however because Zinter couldn't seal off the backside linebacker and Nugent started to lose that NT.

So Corum hops out to the lane outside of Barnhart, who got around the DE lined up inside of him. The only problem out there is ECU's safety was firing down on run action. But Barnhart and Loveland opened up enough space for Corum to get a shot at a Corum'ing.

At least until the backside LB that Zinter whiffed on catches up.

So despite the defense delivering an unblocked safety to the gap at 1 yard, this play ended up working because Corum shook him for a little. It *SHOULD* have worked however because the DTs shot upfield and let themselves get reached by Nugent and Barnhart. The gap that should have created didn't come off because Zinter whiffed on a backside LB he had position on. It wasn't an athletic thing—Zinter just got too far downfield and the WLB had room to shoot behind his back and still be relevant.

Stretch also requires adjustments to things that defenses commonly practice to defeat stretch. We saw a good example the first time Michigan ran this against UNLV on Saturday. On this one UNLV lined up with two DTs over the same OL.

Why? Jon Duerr explained what UNLV did here to blow it up.

By doubling Keegan the defense was able to get past the blocking, make Barnhart useless, and cut off access to all of the other lanes. If you're wondering if Stretch has a response to this, well yes of course it does. As soon as he releases into air, Barnhart needs to check inside and Zinter needs to wall off to create a cutback lane behind the double so Nugent can ride the DT away from that lane. They also need to prevent the backside end from crashing down on this (zone reads and bootlegs punish that kind of behavior).

THE BODIES ARE EXPENSIVE

By this I mean you need the personnel who can execute the blocks that make Stretch effective. Size helps of course, but more than size you need linemen who are super agile as they try to out-dance defensive linemen trying to avoid getting outflanked. Stretch thrives on blockers getting across the defenders who are supposed to be attacking their opposites' playside gaps. Reach blocks (1-on-1) and scoops (when two OL combo each other around a defender) are what create most of the big gashing gaps in Stretch. The lesser way Stretch gets yardage however is when the OL manage to spread out the defenders so much that the RB has room to put a move on an unblocked defender in a wide gap, like Corum did to the safety above.

Execution has been an issue with Michigan's Stretch, and is likely to remain so because they have big hulking OL who aren't made for it. This one was blocked correctly except Myles Hinton, a verifiable hoss, couldn't get across the backside DE.

That's a tough ask for a guy his size. Teams that run a lot of Stretch usually recruit light-footed tackles who can make that block with more consistency.

The other thing you see about the above is Barnhart is dealing with a guy in the backfield who forced a cutback. Again, a tackle with super light feet might be able to pin or ride that guy inside, but Barnhart isn't really that guy. He gives up too much ground trying to get around, and that guy combines with the guy Hinton missed to tackle.

Here's another Stretch from this game that got defeated at the blocking. This time it was Hinton getting shed on his kickout block and Keegan not completing the scoop with Nugent on the NT.

We've seen Keegan make oodles of excellent blocks by identifying the right defender in Michigan's power schemes and by overwhelming DTs he's doubling. This block wasn't about identifying the right guy to block; it was about getting around the guy, which is just a matter of foot agility, which isn't Keegan's strong suit.

Michigan's linemen aren't complete stiffs however. Barnhart and Hinton both got effective blocks on their defenders to create a decent gain when UNLV presented them a wide split to attack, with Barnhart getting out on a hybrid linebacker and Hinton getting across a WLB.

I think Michigan hoped to burn UNLV a few times with this so they could go back to running Duo or forcing defenses to expend their safeties defending it. Stretch isn't very forgiving on teams that don't want to specialize in it however. How and whether it develops as a complement to their Duo game will have to unfold over the next few weeks. If they've already repped it enough to be running it enough to draw responses to it I doubt they're going to put it on the shelf.