Hokepoints Empties the Bubble Drawer

Submitted by Seth on December 3rd, 2013 at 3:34 PM



The best reason I've been able to come up with for how this Michigan team could put up that kind of yardage against Ohio State is that Ohio State's defensive players are—man, how do I say this without being a total jackass homer rival?—more prone to mental errors than your average Big Ten starters.

I hereby dedicate this post In memory of the too-short MGoCareer of Heiko "Bubble Screen" Yang. Who needs doctor money anyway?

Another way to say it: the best and most representative player on that unit is Ryan Shazier, who is basically Jonas Mouton with five years of good coaching. Another way to say it: they're exactly as dumb as they are talented, and that's why a group of 5-stars are just an average defense. I am a total jackass homer rival.

The second-best reason, and the best you can say without coming off like a TJHR, is that which Borges himself apparently gave in the pre-game interview with Musberger: "We emptied the drawer." In other words, they finally ran all of those counters to the things they'd been doing all year.

There will be plenty of time in the months ahead to wonder why it took this long to throw paper, especially when that gamble came up just short (and the last play was a rock that OSU allegedly* RPS'ed) of paying off. For the moment, let's look at one of the "third" things they brought out for this game and what that did for the offense.


* Ohio State's players threw out one of those heartbreaking quotes about being uber-prepared for what was coming, but the play also had Gallon about to break open.

It's hard to argue Funchess isn't an "ideal" slot ninja, isn't it? [Upchurch]

The Bubble Package

Yards per attempt; attempts in parentheses:

Opponent Run Bubble Other Total YPP
MSU 2.0 (1) 8.0 (1)   5.0
Northwestern 5.3 (7) 5.7 (3)   5.4
Iowa 3.0 (5) 1.0 (2)   2.4
Ohio State 4.5 (4) 7.7 (3) 18.0 (1) 7.4
TOTALS 4.2 (17) 5.6 (9) 18.0 (1) 5.2

Michigan does the bubble differently than Rich Rod—he made it an automatic check against the slot defender getting too close to his running game—but both work under the same principle: keep your grubby SAM's hands away from my interior running game!

The Borges Bubble game debuted against Michigan State as a bubble screen(!) that got a remarkable-for-that-day eight yards, followed by a fake bubble (out of the shotgun) to inside zone that got unfortunately blown up by a double-a gap blitz. It really came out in the Northwestern game: ten plays for  5.4 YPP. Of those, three were the bubble screen, four were a fake to an inside zone, and three to an iso. Once it was on film, Iowa adapted but Michigan ran the same (basically) two things they had against the Wildcats. The result was 2.4 YPP on seven tries: 2 bubbles and 5 inside zones.

They run it out of different formations, usually with two tight ends opposite the bubble twins (20/27 plays I have charted were from the Ace twins twin TE or I-form twins). They do run other stuff from these formations but twins (two receivers to one side) with Gallon on the line and Funchess in the slot is a good sign the bubble game is in play.

It's a good fit for this team since it: A) de-emphasizes interior blocking by holding the SAM outside and letting his OL play 5-on-5; B) Utilizes the surprising multi-threats of Gallon (as a blocker) and Funchess (as a slot receiver), and C) Lets them get Derrick Green running downhill.


I don't have Iowa video but I can show you how they adapted. The first time Michigan ran it they threatened blitz with the SAM:


Then had that guy back out and attack Funchess. The idea was to lure Michigan into a screen if this was a check, and then blow it all to hell. Like I said, it's on tape. Fortunately Michigan doesn't run checks; they called run:


Iowa got to play their base defense against that basic zone run, and the result was 5-ish yards. That is rock on rock: it's blockers versus the blocked until safeties arrive, however the SAM was kept away from the running game by the threat of Funchess. The thing is, up to then Michigan only had a rock and a scissors, so Iowa could spend all day in this defense, ceding 3-5 yards when Michigan ran it, and blowing up the bubble constraint.

Here's what this looked like when OSU defended it:

Same playcall as Iowa except since they knew it wasn't a check they didn't bother with fake SAM ("Star" in Buckeye terminology) blitz—just lined him up against Funchess. A screen is dead.

But watch Joey Bosa (#97 on the bottom of OSU's line) get way too upfield and try to knock down the screen pass that isn't coming, thus taking himself completely out of the play. He's matched against Lewan instead of Butt, though, so Michigan was probably going to get something out of that block anyway; you still don't want to make it so easy.

The middle linebacker (#14 Curtis Grant) compounded matters by Obi Ezeh-ing his way to the hole, which gave Kerridge enough time to arrive and pop in an advantageous position. Finally, the safety (#3 Corey "City in Pennsylvania" Brown) took a long time to read the play, backing out a few steps before setting up at the 1st down line. He might have been run through if the other safety (#4 C.J. Barnett) hadn't made his way over, got depth with a neat little athletic step, and helped stop it.

So rock on rock nets a big hole and big yards, because Ohio State's defenders are something-something box of rocks. But they're not the only talent-deficient guys on the field. Michigan's OL screwed up rock on the third bubble package play of the game:

That's inside zone. With the Star taken out by the bubble fake, everyone is blocked except the safety coming down (#3 Corey "a Jewish suburb west of Pittsburgh" Brown). And he was set up outside so if Mags and Glasgow can hold their downfield blocks this could bust huge. However Glasgow and Kalis didn't do a very good job on their exchange—or else the DT (#63 Michael Bennett) just did a great job fighting through it—and the Buckeye DT ends the play with a mouthful. Bennett was bent back when Glasgow released so my inclination here is to point at Kalis and call it ten-man football.

In the Iowa play I wish I had video of, that DE threw off Butt, and the middle linebacker, despite drawing Lewan, managed to attack quick enough to cut off escape until everyone else arrived, which didn't take long since Iowa's safeties were playing with their ears back. However Green's momentum vs the size of those guys got an extra two yards. Here his 240 lbs. are irrelevant against a wall like Bennett.

[After the jump: other things you can make your fist into]


Here's the second time Michigan ran the package against Iowa:


This was a bubble screen that Gardner threw too low and Funchess dropped. But it wouldn't have gone anywhere because Iowa had its safety moving toward the line at the snap and committed to taking out Funchess while the strongside linebacker blitzed. If this was a run the RB would have eaten SAM in the backfield.

Ohio State used almost this same response:

Linebacker blitzing=run would be dead, but Michigan's coin flip victory there isn't going to get much because the safety (Corey "Monroeville" Brown) is going right for Funchess. This time he gets picked off by Gallon's block and the corner (#12 Doran Grant) is coming up for a free hit after just a few yards.

One of the extra bits of win Rodriguez used to get from his bubble screen game was to have a little jitterbug at slot receiver who could often get extra yards when isolated in space with a defensive back. That's exactly what this play does, except Michigan uses the very not-slot Funchess. You'd think he's not going to shake many guys—maybe plow them—but then on the Ohio side not lunging at a guy twice your size was always an "Area for Improvement":


Grant gets hurdled like a fool, and Michigan is awarded 20 extra yards for having the better football player. That's a thing to keep in mind about the bubble screen, or any play: it's not so much the design of the play but how the design of the play works with the talent of your players and that of thee opponents. When Brian gets around to UFR-ing Iowa (which he will now that his latest attempt to pawn it off on his employees was thwarted), making a play come down to Doran Grant's tackling can be an RPS win itself.


Since Michigan ran ten plays from the bubble bag against Northwestern we'll start there for the third counter:


Oh. Well I guess they didn't need one since Northwestern didn't see this coming and thus didn't prepare a counter. Against Iowa, then, because after putting it on tape of course Ferentz would have his counter prepared. Here's what Borges came up with to defeat that:


Right so Michigan didn't show a third thing and Iowa duly brought out paper. On the fourth bubble look, Ohio State did the same thing. Here's what it looked like:

Drawn up:


Some of this might have been man or half-quarters but it's irrelevant. They rolled the coverage to the bubble side, blitzed the Star, and had the safety (Corey "Just off Rte 22" Brown) high-tailing toward Funchess the minute he saw the bubble; the cornerback as well. It was the screen, Gallon walled off Brown, and Grant arrived quickly enough to not be made a fool: 3 yards.

So on the very next play (3rd and 11) Michigan throws out a wrinkle:


(Funchess and Doran Grant are spread to the wide side)

A whole new formation! And instead of throwing the bubble inside-out to the slot (now Chesson), it's coming outside-in to the split end (Gallon)! And this time the run play is a zone read. To Kerridge! On 3rd and 11! So this is one of those things where Borges thinks he's got a paper and really it's just acting like the defense is completely stupid.

Ohio State has a wrinkle too: Shazier has his hand down on the playside, and Roby (the boundary corner) is right up in Gallon's grill. Notice how Shazier and Roby—their best defensive players in the back seven—haven't been very involved yet? Funny that. Well now they are, as is the nickel back Armani Reeves. These are big upgrades from the dudes we've been involving before, and it helps.

Shazier comes in as the option guy and only makes a quick feint toward Kerridge since lol handoff to that guy in this down and distance. Because Shazier's a ridiculous athlete he is on Gardner before the screen can be set up. Not that it matters since Roby wasn't letting Gallon get loose and Reeves is locked in on Chesson, and even the free safety (#4 C.J. Barnett) is coming way up to blow this up.

THIS is when Michigan should have used their timeout (or gotten to the line early enough to check out), because enough Buckeye talent to empty a car dealership lot was loading up to stop this playcall. OSU got the ball back and tied it, and Michigan got the ball back with a minute left in the half. I won't count this in the stats since it's a conversion down and OSU isn't playing a base thing, but I wanted it mentioned.

Paper-ish Rock

At this point it seems Ohio State has adjusted but Michigan didn't stop running it. Nearing the end of the first half, after getting a first down on a great Green run, Michigan ran the PA handoff, basically to run the clock out. and OSU ran their "rock", giving up three. On the first play after the half, we tried it again, it was the bubble screen this time, and OSU again went with their paper: the blitz/corner attack, except Doran Grant didn't even bother with a courtesy step back into coverage before firing at Funchess like a bullet:

About now we're screaming: THEY'VE ADAPTED, AL!!!! But of course he's not sure the paper's set up yet so:

This time the Buckeyes faked like they were going to do the paper bug-out again but did a different paper: slanting the line to the playside and otherwise doing rock. Michigan still had numbers but Mags didn't recognize Glasgow was going to be in a terrible spot and went to help Lewan wall off the 5-tech. Meanwhile the WDE (#8 Noah Spence) just pwned Jake Butt.

And then again on the last play of the 3rd quarter they lined up in twins and faked the bubble and…

The Real Paper

Jumpin Jehosaphat that's different! On the eighth bubble package look, down two scores, they finally pull out the counter thing. Again OSU was in the fake-paper rock thing so it didn't hit huge but this is what setup play action looks like: all the DL stay home to close running lanes and the quarterback gets to sit and consider how he would divide the land amongst his minions. Funchess actually had Tyvis Powell beaten if they wanted to go over the top but Gallon broke so wide open on his flag route that's where they hit. The linebackers, too, were sitting in against the run.

So now they've got the constraint set up and…and they put the whole package on the shelf. For the non-hurry-up drives in the 4th quarter Michigan didn't even line up in a twins formation, rather putting the receivers in a stack and running power and PA off of power. The PA bubble pass to Borges wasn't a constraint to make the defense play the bubble package honest; it was the coup de grace.

And thus we see how Borges sees paper: a thing to save to really nail your opponent, but not worth doing just for the sake of keeping the base thing clean. Ultimately Michigan hit big twice on this package because Ohio State's defenders screwed up. Then those plays went for 2, 3, 3, –1, and 2 yards on the next five attempts because OSU adjusted to stop the only pair of plays the Wolverines had ever shown. Then this burned them for 18 yards. What should have happened next is a string of 5-yard PA runs and bubbles with more passes mixed in.

So give Borges some credit for his OSU game plan. But I think we also have every right to be pissed that he thought he needed to let defenses tee off on the one thing his offense does well for 1.75 games in order to set up an 18-yard gain. They've finally found an offense that they're kind of good at and which they can force the defense to defend with their vanilla. And the second that happened, they scrapped it, because Borges doesn't get constraint theory.



December 3rd, 2013 at 4:05 PM ^

No. And by the way it has a name: the "Transcontinental." It's a play they pulled from the old Lloyd days, back when Charles Woodson or Steve Breaston was the receiver-quarterback.

I looked at it and that was out of a completely different look than they've been using to run the bubble game. Only similarity was that the receiver stepped back like he's getting a bubble screen, but only after taking a step forward to get the defense backing off. Ultimately the action and situation didn't seem to match the downhill bubble game--it was closer to Michigan's shotgun passing game where Borges bunches lots of receivers to one side then runs triangle rouotes and floods with them.


December 3rd, 2013 at 4:03 PM ^

Is Heiko's really gone?  How the Hell did I miss that announcement?  I give that guy SO much credit for having both the brains to understand what was going on and the huge balls to ask questions nobody else had the stones to ask.  I would literally skim pressers looking for his questions cause they were usually the only ones that made any sense and required the coach to actually think about the response and not give some bullshit canned response.


December 3rd, 2013 at 4:34 PM ^

I think he was always in Med School but he was on a MD/PhD track. So maybe his research portion is up and now he's on to clinicals (the last two years)... I can't imagine he'd have time to attend pressers and whatnot during clinicals. Just a guess though haven't seen an announcement or anything.


December 3rd, 2013 at 4:18 PM ^

This just confirms more about what we know of Borges, which scares me. He wants to hammer a D over the head with a few plays, just to set up once big counter later in the game. Its like when he's calling a play, he's thinking about one play 8 plays from now. Personally, the only way this really works is when you have a Caddilac Williams, Jason Cambel, and a great O-line, because that allows you to run plays the D knows are coming, and at least not get stuffed for 0 yards. That forces the D to over compensate the way Borges wants. They don't over compensate when their base set stuffs us, then they just wait all day for the counter to come. Its like he's the opposite of Bill Bellicheck. I know its the pro's, so its not the best analogy, but with BB its always next man up. Doesn't matter matter as much which players are in there, he's always a step ahead of the defense and finds ways to get his guys open. Rather, Borges needs a team that can run 6-10 base plays so well, the D is forced to over compensate to stop. I just don't like the thought of having to wait until we have the players capable of that  

snarling wolverine

December 3rd, 2013 at 4:54 PM ^

So give Borges some credit for his OSU game plan. But I think we also have every right to be pissed that he thought he needed to let defenses tee off on the one thing his offense does well for 1.75 games in order to set up an 18-yard gain. They've finally found an offense that they're kind of good at and which they can force the defense to defend with their vanilla. And the second that happened, they scrapped it, because Borges doesn't get constraint theory.

Given that we finished the game with three consecutive TD drives (and the drive before that went 69 yards before a turnover on downs), how exactly did we scrap the only thing we're good at?

This article has a definite whiff of confirmation bias.  It sure sounds like you guys made up your minds about Borges before this game happened, because you're basically making the guy's gameplan sound like crap when it produced 603 yards and 41 points.  If you want to pick apart the Iowa gameplan, OK, sure.  But this game?  Really?  




December 3rd, 2013 at 5:33 PM ^

if you go back and watch every offensive snap (its on the tubez) you'll see at least 1/3 of these yards come from unforced errors (blown coverage, needless over pursuit, misalignment) another 1/3 from guys who just seemed to MAKE PLAYS (broken tackles, leaping over a corner) and the last 1/3 from good play calls. You don't get kudos for the first when you don't do any line checks. I won't tear Al to shreds, this was pretty good all in all...But you can't analyze this game in a vacuum...Al has sucked hard all year long; one decent game does not overcome the eye-bleach inducing four game run we just had. It was above average for Al below for most other OCs

Space Coyote

December 3rd, 2013 at 5:41 PM ^

"Needless over pursuit", so taking advantage of that isn't because of him. "Blown coverage" does that have to do with the play design, or maybe because it looked like something run previously. "Misalignment" happens, but sometimes that's due to offensive formation, doing something different out of that formation (breaking tendency) and other things. Also, most teams have guys "make plays". That's part of any good offense.

This offense has had some very rough games, but essentially saying 2/3 of the performance against OSU was because of luck? Come on. How can you honestly say what Michigan just did to OSU was "below [average] for most other OCs"?

Ron Utah

December 3rd, 2013 at 5:43 PM ^

AB's offense produced the best yards/play against Ohio since...ever.

I have watched every snap.  I see plays that are working and players that are executing.

If you're going to play the "1/3 off defensive mistakes" card, what percentage of AB's play-calls were ruined by our own team's screw-ups?  And isn't it the OC's job to call plays that force the defense to make mistakes?

Ohio's defense is hardly 'Bama, or even Sparty.  But they were the #13 defense in the country when we faced them, and we did far better against them then any of their other opponents this year.

I'm not going to defend Borges' season--it wasn't good enough.  But to call ths game "above average for Al below for most other OCs" is ridiculous.


December 3rd, 2013 at 5:57 PM ^

I really enjoyed the analysis.  Best one I've read this year.  However, I agree with you on the commentary part.  This was a great game plan, and great playcalling.  Consider that we're dealing with three freshman linemen and throw that into the mix as well.  One other think that made us look as bad as we did against Nebraska, Northwestern and Iowa is that Gardner played poorly in each of those games, but against ohio decreased his propensity to look at rushers and the hold the ball.  It still happened.  But not as much.


December 3rd, 2013 at 4:55 PM ^

This is my biggest issue with MANBALL. Its not necessarily that spread is better than manball its that most manball coaches go back to the 3 yards and a cloud of dust "I know what I'm running, you know what I'm running but I'll force my will and you won't stop it."

Such a neaderthalic (yea, that's not a word, I don't think) way to play football. Guys who run the spread are more likely to build constraints into their system (or so it seems)...maybe its a generational thing? 

Anyone more well versed in a functional manball team have good examples of manball with constraints built in? I can't say I'm well enough versed in playcalling theory to be able to see it myself. 

I'd be okay with an offense that wears manball clothes (2TEs, I form, lots of under center) if it thought like Art Briles/Holgo/C. Kelly/RR.

Running an offense that is dependent on out-talenting your opponent is not practical anymore.

Space Coyote

December 3rd, 2013 at 5:37 PM ^

And there is no "spread" verse "manball". Manball isn't an offense, it's a made-up phrase.

But back on topic, every system has contraints and contraints on top of that, it's simply that spread does it in a much more visable way. In fact, many of the contraints used by most spread coaches stem from pro-style schemes. Many of Meyer's contraints stem from pro-style schemes. Dating back to UCLA Borges has had bubble screens, tunnel screens, middle screens, flare screens, long screens, throw back screens, slip screens, etc.

Power O -> Boot -> Throw back screen. Three iterations.

Power O -> FB Flat -> FB wheel. Three iterations.

Iso -> Iso PA -> Iso Hitch and Go -> TE seam (when the safety sneaks over the top)

Iso -> Iso PA -> Iso PA Tunnel Screen (it requires the outside CB backs off on PA action)

The reason things like bubble action come into play more often for spread teams is because spread teams tend to utilize WR screens as extended runs. It's different theory. They require that extended run to have run game diversity, because they don't have the blockers inside to account for extra defenders playside (as in, you can't block someone coming from the slot because you don't have the lead blocker or TE available to do so), and as such they need to build in a constraint off of that screen action.

A pro-style system tends to utilize the screen as a constraint itself. It's a third iteration itself. Spread does this too, FWIW. In Indiana's screen package the utilize bubble/flare action (a quick screen) one direction and a tunnel on the backside. It's just that the flare/bubble are runs in their system because of what I described above, and their third iteration is based on that and is more in the open. Pro-style is less visable in it's adjustments, because it's more compact, and a lot of the iterations come down to blocking schemes and less obvious things inside, but they are there.


December 3rd, 2013 at 6:09 PM ^

As a non X's/O's guy, I love it when you post.

The last part of your comment here reminds me if something Bo said in his book w/Bacon, about how it only looks like they are running it into the line over and over from the fan's perspective; in fact, they are changing up blocking schemes, testing the opposing DL, etc, things that the casual fan (like me) would not notice


December 3rd, 2013 at 7:59 PM ^

This is a fascinating debate, and a crucial one to the degree that the future of the offense depends on its outcome. It seems to me that Seth and Brian are generally correct about the spread and especially Borges' take on constraint theory. But I have a fraction of the knowledge Seth and Brian do. Space Coyote seems at least as knowledgeable, and I desperately hope that he is right. Everyone else should too because we are moving ahead with Borges for better or worse. Space Coyote is the only thing giving me hope.



December 3rd, 2013 at 8:22 PM ^

hand avocational aficionados, on the other an aficionado with vocational experience. 

While not in the sports arena, my background is also in small unit and team training so can relate to the objective evaluation of what is going on, in what context, and what it indicates for the future.

Space Coyote

December 3rd, 2013 at 5:25 PM ^

The bubble/hand off is a read. You reference the low pass to Iowa, that was a play in which DG second guessed his read, and that messed up his footwork causing the ball to go low. The read is not presnap though, it's something I talked about with the extended hand off, it's a post snap read.

And it's predicated on numerous things that you have to work on reading. What's the initial alignment, how is the coverage rotating, where is the key reacting? If the key is 10 yards off but his initial step is forward then the bubble isn't necessarily open. If his first step is neutral it is. If the key is 5 yards off but he's bailing hard at the snap, the bubble may be open. If the LB is inside far enough and the first step is to the QB, then the read is likely bubble. This read isn't something that is easy, it takes practice. People complained about the lack of bubble or the lack of extended hand off, but DG struggled making the correct read earlier in the year and only started to get better at it later (and still made a couple mistakes). But this is almost certainly a read. In fact, you see Gardner reading the slot defender every time (this is why the play didn't work as well when the safety walked up, because that wasn't his key).

2nd, the tunnel screen that you point out as a paper-ish thing is very different than a bubble. It's certainly not a bubble from the outside-in. This is the same play that sprung Gallon early, it's just from a shotgun look. The PA is similar to Indiana pairing the tunnel with a screen to the other side, it's just something to get motion away from it to help form the tunnel. There are two key blocks, the WR and the EMOL (most often Lewan) (some teams run it differently, but it's still two blockers: support #1 and alley). It's a play Borges has run since ND in 2011 at least. For whatever reason people have insisted on calling it a "throw back" screen, but it isn't that at all. Run action is away, blitz comes up the middle, a tunnel is formed between the WR blocking inside-out on the CB lined up over #1 and Lewan leads in the alley and picks off the next guy. This forms the tunnel between the outside guys and the box players diagonally across the field (or if the backside safety runs through straight up, the WR cuts it up). This was Borges trying to catch OSU in a blitz and get the ball out fast.

Now, for the Paper play. The paper play has nothing to do with the player defending the slot. This is different than when Rich Rod ran bubble because Borges is having them block it differently. Most spread teams block the outside CB with the #1 and try to seal him inside, effectively blocking that guy and the slot defender in the process, giving a lane on the edge. Borges runs the #1 down to the slot defender and blocks him, attempting to block the outside defender with the route. The former depends on WR blocking more and is a bit more dangerous because the slot guy can jump it (meaning it must be practiced more), but if executed it's generally more effective. The latter is less dangerous as far as turnovers but more prone to get blown up, it gives a larger alley, but a cleaner shot at maintaining leverage.

But either way, the key to getting to the screen action is the outside defender. In Rich Rod's system it's about the slot defender chasing and the outside defender jumping outside to maintain leverage. In Borges's system it's about the slot defender blitzing or chasing and the outside defender coming off his man quickly. If you watch the outside defender, up until the actual last bubble screen, the outside CB doesn't react until after the ball is thrown, he's still bailing in coverage. Then on the last bubble pass, the outside CB is shooting up immediately on bubble action, now the third iteration is open.

It turns out the CB wasn't consistently jumping the out route, but he was jumping routes, as was the rotating safety that was filling the seam so that Michigan could run the slip iteration. This is the same exact thing Rich Rod looked for when running this, because there are two different third iterations: the pop pass in the seam or the seam-and-out. This is why they went with the wheel and with the seam-to-out concept, the safety was defending the seam against the pop pass. Clear everything underneath with the wheel, if the outside CB bites throw the wheel. If the he doesn't and the safety fills the seam, throw to the out cut.

Anyway, there is a lot more to calling it than just calling it because the first two iterations aren't working. It needs to be repped to get the keys correct, it needs to be run to see how the defense is adjusting. Borges ran exactly one more time after OSU adjusted in a way to take advantage with a third iteration. I'm not saying that it isn't something that should have been prepared to the point of bringing out against Iowa, but they may just not have been ready. They may not have been comfortable with the line holding up or how the blitzer was reacting or the coverage on the back end. It's not that he doesn't get contraint theory, it's not that he likes running things to keep killed at them then abandoning it, it's much more likely that the team wasn't prepared or didn't see something they liked in the other games to keep running it or move on to the third iteration. Sometimes the third iteration isn't set up in a way to contrain the defense one step further, so you go a different direction, particularly when that something is installed half-way through the year. The last line, while likely said jokingly, is quite false.


December 3rd, 2013 at 6:06 PM ^

"I'm not saying that it isn't something that should have been prepared to the point of bringing out against Iowa, but they may just not have been ready."

I get that if they're not ready because a bad post-snap read on bubble action and you're hoping for a PBU because at least that's not a pick-six or a TFL.  But shouldn't these plays ideally be installed as a set?  If you're a constraint short, you don't have a sustainable offense.

It's hard for me to understand what you're saying in fine detail without some diagrams, but at the end of the day, what I want to see is when the offense aligns a certain way, the defense can't deviate much from their base alignment.  You might see a safety shuffle or a linebacker show blitz, but the idea of constraint theory, at the most basic level, is to eliminate pre-snap reads and force the defense to react -- or, if they insist on going SPARTAAA on your Rock, burn 'em for 20 yards with Paper.  If they've got Rock and Paper covered, the OC had darn well better have a third play from the same look ready to justify his salary because RPS has been around forever.  I won't pretend I understand football well enough to know what a safety's read is on a play just by watching them move, but I can see how they move.  When defenders tear turf at the snap for the ball even before it gets there, something's wrong.  I guess that's where I share everyone's frustration.

This has jack to do with pro set vs. spread; as you said, all functional offenses utilize constraint theory.  But I don't see an acceptable explanation for the accusation that Borges ran these plays for two games with no answer for the inevitable defensive adjustment if the players weren't ready.  If he doesn't know, well, he shouldn't be an OC.  But if he knew they were short on weapons and had them go over the top anyway, how's that any better?  That does beg the question what the hell else would he run considering just about nothing was working, but the answer to that question still falls on the OC to prepare a scheme the players can execute, constraints included.


December 3rd, 2013 at 6:29 PM ^

The Space Coyote, but like you I'm no X's and O's guy.  This is what I took from what he wrote.  Given that the entire offense, save three seniors, is inexperienced, one must bear in mind games still must be played, even if it means crawling, walking, and only then running during the season.  I think much like what one experiences when learning to walk, there were times the offense simply fell down.  

As for the OSU game, I'm not prepared to say the offense was running against OSU, rather it was shuffling forward at a brisk walk. 


December 3rd, 2013 at 6:33 PM ^

They'd help, actually.  What I can't follow is jargon, like this:

"Run action is away, blitz comes up the middle, a tunnel is formed between the WR blocking inside-out on the CB lined up over #1 and Lewan leads in the alley and picks off the next guy."

I got parts of that, but football is a formation sport so positioning is key.  If I can't see where the X's and O's are, I can't follow this stuff.

Space Coyote

December 3rd, 2013 at 7:04 PM ^


Now, this is different for reasons I described above, but if you scroll down to the combo screens, you'll see a bubble one direction with a tunnel on the backside. In those diagrams, Indiana has trips to bubble, and a single WR to the tunnel side. They tend to run a guard to the support player and the OC to the alley. It's slightly different, but in general the same thing. Michigan more often than not uses a slot to block support and the EMOL to block the alley.

FWIW, almost all screens require three blocks in general (bubble is an exception): Support (outside), Alley (safety), and Seal (inside). A tunnel screen untizes the run fake (or bubble fake, or flare fake) as the seal block most of the time.

Space Coyote

December 3rd, 2013 at 7:10 PM ^

The bubble itself if a constraint off their base play. They likely are trying to install the entire package, but it takes time and reps along with doing the rest of their offense. In the mean time, they likely installed a portion of it to the point of use so that they could find some success with it. Like GoBLU said, you still have to play a game every Saturday. Even that one constraint helped move the ball against Iowa. When they couldn't take the next step they went to something else.

But yes, particularly if something was installed in the off-season, it would be installed as a package.

FWIW, I've said this elsewhere and people have scoffed at the idea, but a bubble screen is a play that needs to be repped a lot. Once it is, it is a very easy play to run, but before that it isn't necessarily. The footwork is awkward (less so in gun), the angle of the pass is awkward (more so in gun, as is the actual route for the WR), the ball handling is difficult (equally in both, but must be even faster from gun, though has the added benefit of looking similar to a weak run fake), the read in this case isn't easy. So it's not simply throwing it out there, if you don't practice it you will look very bad doing something that looks very easy, but it isn't until you practice it, if that makes sense.


December 3rd, 2013 at 8:02 PM ^

As an old ballplayer, I often get frustrated with people who just don't understand how hard the game is. When this whole bubble screen obsession started on here, I pointed out how bad Denard was at throwing it. People called me a fool and other unkind things. After all, the bubble screen is an "easy" play. Its just a pitch and catch. Anyone can do that.

But it isn't. Nothing is. Denard skipped quite a few of them off the turf, and certainly wasn't consistent in getting it to a receiver in stride and in position to do anything with it. I'm sure he would have improved, with practice reps, just as he improved with his reads on the option and in other areas. That's the nature of the game. They don't practice because it's fun.

Football is hard. Running an ISO is hard, even if its the easiest play to draw up on a whiteboard. I've never played QB, but the amount of moving parts that go into playing the position is staggering. If your footwork is off, incomplete. If your ball-handling is off, you're late and probably incomplete. And on and on.

This is why I generally defend coaches and tend to give players more blame/credit than most. Devin Gardner barely resembled a football player two springs ago. He was as raw as they come. He took snaps from the gun and literally did nothing with his feet. Just plant and throw. He is now a QB, and a pretty good one. He has improved tremendously since then, and the improvement has been consistent and constant. The results have not been perfect or even ideal, but they almost never are. Peyton Manning and Tom Brady have bad days. But people keep telling me Borges has ruined Gardner. It's nonsense.

Space Coyote, I love you. You've yet to be an out and out asshole to people. I haven't been able to do that, unfortunately. And you are the only guy on here who's criticisms are based on anything other than results. You try to understand the underlying ideas and see if they are sound, or if they make sense. Seth has a great post up here, and there are 30-some comments. Someone takes a quote from a presser and we get 300. People don't seem to want to understand how the game really works. They just want to know enough to be able to complain. Its sad and frustrating.


December 3rd, 2013 at 8:59 PM ^

"After all, the bubble screen is an 'easy' play. Its just a pitch and catch. Anyone can do that. But it isn't. Nothing is."

I'm sorry but I'm going to have to draw a line here.

I won't speak for or defend the derpheads who trivialize everything, but I have never, EVER heard "it's hard" accepted as an excuse for anything that has been executed with regularity.  You can argue it's acceptable, but in reality it's never accepted.  I mean, can you give me an example of anyone out of HS who gets away with that?  My profs sure didn't buy it; in hindsight I have to admit they were right.  The excuse didn't work; I had to grow up instead.

Of course it's hard.  No shit.  That's why it's guys like DG and Gallon up there instead of a balding pencil-necked geek like me, doing these things on scholarships worth more than what I make.  But you know what?  The crap I do isn't easy either.  It took six months of daily training before I was ready for OJT; throw Peyton effin' Manning into the seat of my job and he'll crash & burn in half an hour.  Of course he would!  Who can possibly do anything that takes months of practice in a day??  But I don't expect him to do my job.  Everyone I work with expects ME to.  They don't care if I'm ready or not; if I'm not, my boss gets the heat.  That can get rough at times but it also limits my sympathy for Borges' plight.  That guy makes money I could only dream of.

I mean, point taken, I'm not one to believe ANY play can be installed in a day.  It takes reps until the body can do what the brain wants it to.  But lack of preparation is on the OC.  That changes the argument a bit -- it's "why couldn't he get the players ready" instead of "why doesn't he call a play they haven't even repped" but the responsibility is the same.  I contrast with Mattison, who also had a ridiculously young squad to work with but implemented a scheme that was, while maddening to watch at times, something they could execute well enough to hold their own.

Football is hard?  Really!  Whether it's college football or a desk job, if what you take pride in can be mastered by anyone in a day, you're not ready to be a grown-up.


December 4th, 2013 at 12:37 AM ^

I'm no saying, "football is hard, lets give everyone medals for trying." I am saying that football is hard, and it takes time and reps to get something as simple as a bubble screen right enough to call it in a game. Stuff takes time.

This isn't even a defense of Borges, so there's no need to get all up in arms. Its related, in fact, to my biggest problem with Borges.

We ran a grab bag offense that tried to exploit each defense's weakness because we are incapable of just playing to our strengths. This is due to our poor line play. But this is bound to fail. You can't ask a team that cannot run 4-5 base plays well against all looks to run a bunch of other stuff well. You can, but you can't expect it to work with any sort of consistency. Desperate times call for desperate measures, I suppose, and a bad line is certainly a desperate situation.


December 4th, 2013 at 7:54 AM ^

Football is hard. But I think Brian's biggest beef with Borges isn't so much that Borges didn't call bubble screens, but that he dedicated two weeks of practice time to developing another constraint (Tackle Over) that had no logical basis to succeed. If Michigan had dedicated that time to developing the current bubble offense, then maybe we're talking about 9-3.


December 4th, 2013 at 10:29 AM ^

Yeah that's kind of what what I said above. I actually had a few lines about just that but erased it because the post was super long. Tackle over was an effort to get our two best lineman at the point of attack against Minny's smallish ends. It worked. But against better defenses (and better prepared ones from having our tackle over on tape) it didn't. But we ran it because we had repped it whereas we probably hadn't repped the bubble much. This is a criticism of Borges, and a legit one.

Space Coyote

December 4th, 2013 at 11:28 AM ^

But disagree that it's a "gimmick" (not your words). It isn't a gimmick, it's another formation. It isn't necessarily a base formation, but it's not like throwing a HB pass either. 

The point, like you said, is to get the two best blockers at the point of attack. For this to be effective, the backside (the G-C-G combo) needs to be "good enough". Now, because it's further to the backside, "good enough" isn't as hard to achieve, it just needs to be enough to keep the defense honest for the few times you run that direction. The problem is, not meeting the "good enough" criteria leads to worse results than it would in a standard formation.

So what you saw was a game against Minnesota, where they didn't have it on tape and weren't as good on the DL, where the backside performed "good enough". There wasn't a ton of success to the weakside, but it was enough to keep the defense from going ham. Against PSU, with the formation on tape and with a better DL, the weakside was no longer enough to keep the defense honest, meaning that the whole formation no longer worked, which is why it was subsequently scrapped.

Now, you could argue that Borges should have had a feel that his interior guys couldn't meet that "good enough" criteria, and you would have an argument. The fact of the matter is that they weren't. And they weren't to the point that they couldn't even run some of the staple constraints of the formation that Borges probably would have liked to get to (the TE throw back screen is a big one, some of the playside boot plays, trap plays to the weakside, etc).


December 4th, 2013 at 12:59 PM ^

Yeah I do t consider it a gimmick. Its a formation, just lik3 any other. And I get the idea behind it. And I'm sure there were constraint plays in the package that we couldn't run in games (any pass, for example, as Williams was a total liability). My complaint is that the price of installing that package may have come at the expense of repping our base package. And I think we needed a lot more of those, as our guys are young up front.

Again, I get that Borges is handicapped by the bad line. MGoBlog thinks he should spread the formation and install the bubble as a primary constraint. I have no real preference, but my method might be simpler. Fewer formations, fewer packages, and just rep the shit out of a few plays until our guys can execute it with some modicum of success. I'm not saying its the answer, but it would be my first attempt. Just differing philosophies.


December 4th, 2013 at 1:46 PM ^

You (Reader71) ascribe Michigan's "grab bag offense" to their inability to "just play to [their] strengths" due to "poor line play."  But my problem with Borges is that he has a predetermined idea of what the "strengths" of his offense are (or should be), regardless of the reality.  The reality, for whatever reason (dating back even to last year, as you've pointed out before), is a line that doesn't block well as a unit.  The advantage of a play like a bubble screen is that it minimizes the effect of poor line play, but as Borges explained to Heiko over the summer, he doesn't like bubble screens because he'd rather block and run guys over instead.  Fine, but if that's not your "strength" (and it isn't for this Michigan offense), then you'd better identify what you're good at and avoid what you're bad at.

I guess I agree with you and Space Coyote, to an extent -- I don't think the problem is with Borges' lack of understanding of contraint theory, it's his failure (whether from stubbornness or for some other reason) to appraise and work with what he's got.  The game plan against OSU was an improvement because it worked better with the team's strengths -- e.g., getting the ball to Gallon and Funchess in space and relying on them to beat a man or two, jump ball to Funchess, etc.


December 4th, 2013 at 4:02 PM ^

That's about it. I take issue with people saying a coach is a moron. That's basically the extend of my Borges defense. He gets it. He told this blog he just has a different philosophy. Just because he doesn't agree with ours, though, doesn't make him an idiot.

Your take is pretty much right on. I'm all about taking out TE/FB and replacing them with receivers. This is especially so because our TE/FB aren't great blockers. But I can see why Borges wants to keep TE in our base packages: Butt will be a matchup nightmare one day, and Funchess even more so (should he have developed as a blocker). So, Borges's goal might be worthwhile. But it has hurt us this year. The reason it has hurt us is because our line and TE are bad blockers.

I love when people have other opinions and argue them. Thus is the best discussion I've had about football since UConn.


December 3rd, 2013 at 7:48 PM ^

Where do you talk about the "extended handoff"?

When I watch the bubble screens to Funchess, the one where he hurdles Grant and the two where he gets stopped for a short gain and a loss, the footwork looks different from the snap compared to that first clip where he hands off to Green.


December 3rd, 2013 at 7:49 PM ^

I'd like to get a deeper response when I got the time but wanted to mention quickly that I had to narrow the scope of this by taking out a lot of how the bubble game affects the regular running game. They've found running room from the same formations.

So I don't mean this to be an indictment of all the things they do. Just that it shouldn't take 20 times doing a thing and getting bashed before bringing out a constraint. I love that they are repping the base play because it is a great fit for them. But a constraint to it is really easy for them since they've done more pass patterns than anything.

I did wonder if this is a post-snap read.


December 3rd, 2013 at 7:13 PM ^

Brian and his acolytes made up their mind about MaNBAlL GRrrRR 3 years ago.  NOTHING Borges can do will satisfy them.  We might as well resign ourselves to that fact.  Brian has no interest in a fair, neutral "analysis."  He has OregonDuck on the brain, with a pinch of RichRod.  It's a shame that that monomania corrupts an otherwise decent blog.


December 3rd, 2013 at 7:58 PM ^

Seth, I love this blog and the analysis in these types of posts are incredibly helpful to my understanding of football, but to say something like "borges doesn't get constraint theory" strikes me as so implausible that I can't believe it's so frequently said around here by guys as smart as you, Brian, and the rest.

Analysis like this, while insightful to a bunch of amateurs like us, is surely football 101 at every level from JV on up.  Borges absolutely gets constraint theory.  Now, maybe he doesn't want to run his constraints the same way a spread team does.  Or maybe he wants to focus on repping variety more than base plays with simple, clear constraints.  Maybe these preferences lead to a less efficient offense, or maybe there are one of a million other factors occuring that are affecting the quality of the offense.  But aren't these things much more likely than a highly-paid, incredibly experience OC simply not understanding basic concepts of football?

I'm not a Borges fan, and I'm concerned about his upside, but it's incredible how much we try to simplify things around here.  I think we all need to take a step back and realize that we have no real clue what the offense's problems are, or why Borges chooses to do the things he does.  To suggest so confidently that our problems are because Borges doesn't understand these incredibly basic concepts is so far out there, I can't belive it's constantly parroted on this blog.


December 3rd, 2013 at 8:10 PM ^

Its a blog. It deals in snark. That makes it a fun read.

I had the same thought last year when Brian had a Picture Page that was just positive that Borges designed a play that allowed a free running LB in the hole. It was obviously an ISO with a bad line call that allowed the free runner, which I pointed out in some detail. I was worried that Brian, who carries a lot of cache, was sabotaging our OC in the fan base with such irresponsible claims.

Then I remembered it was just a blog. A good one, but a blog nonetheless. No biggie.


December 3rd, 2013 at 8:20 PM ^

I agree, and I think one of the main reasons for the blog is to speculate as to why things turn out the way they do.  There's just an incredible amount of chutzpah, though, and it's often based on pretty implausible and overly simplistic concepts.  For an author group as sophisticated as this one, who spend most of their time trying to get to the bottom of why the team is good or bad, I'm just surprised people are so willing to think Borges is just plain dumb or incredibly stubborn.


December 3rd, 2013 at 8:22 PM ^

I'm not sure its fair to call out Kalis for "10-man football" on that second one.

On both, Michael Bennett is shaded weakside (playside), engages Glasgow and pushes towards the gap the play is going to.

However, on the first play, Glasgow and Magnuson - who is the playside guard - pushes him back before Glasgow moves on to the second level.  Magnuson is left with Hall two yards up field and perfect position to wall him off.

On the second play, Glasgow and Kalis - who is the backside guard - pushes Bennett back before Glasgow moves on to the second level.  Kalis is left with Bennett two yards up field but Bennett has playside position on him.

So to me, the difference in the two plays comes down to Magnuson having a very easy assignment as far as keeping Bennett from collapsing the hole.

I assume that having Butt as an H in the former play, rather than Kerridge in the I explains the change in assignments.

The question I have (throwing up the Space Coyote signal here) is: is that an assignment you can expect Kalis, or any lineman, to be able to handle.  It seemed formationally doomed.