Two Steps Forward, One Step Back: Analyzing Michigan's 2018 Passing Attack

Two Steps Forward, One Step Back: Analyzing Michigan's 2018 Passing Attack

Submitted by Ron Utah on November 28th, 2018 at 1:08 PM

38b26a40-be47-47af-8df2-3392c74afc43-11242018_umosu_2ndhalf-22.jpg

Losing to your archrival sucks.  Getting blown out by them in a year when you were favored to win in their stadium?  That causes some fans to lose their shit.  Unfortunately, with their shit often goes their perspective, and heated debates about WHAT WENT WRONG?!?!?!?!! ensue.

Play Calling vs. Scheme

Play Calling vs. Scheme

Submitted by Ron Utah on November 20th, 2017 at 2:01 PM

It finally happened

Lots of folks out there interpreted the "Persistent Underachievement" diaries as a call to fire and replace all or a portion of our offensive staff.  That is not the case.  You can criticize a person or group and still believe they are good at their jobs.

After watching this weekend's game live, I thought the playcalling was terrible.  After a rewatch, I actually believe the playcalling was pretty good.  But the scheme was not.

What I mean by that is that, given the set of plays we came prepared to run, I like that the play calling was more pass-oriented and aggressive.  The deep shot to DPJ felt like a long-delayed relief, like that bathroom trip after too many hours in the car.  McKeon dropped a big-gainer from JOK.  Gentry had a long completion dislodged by a big hit.  DPJ had a TD erroneously called back (though he could have done better with his feet there).

That said, I noticed some fundamental flaws with our scheme.  Before I list them, please know that I understand and sympathize with the predicament we face--very young team with insanely young receiving options, young OL, freshman (or 3rd string) QB, etc.  Even knowing these limitations, and perhaps even because of these limitations, I do not believe our scheme is giving us the best chance to win.

  • The early wrinkle was having McKeon on the outside.  Presumably the coaches saw match-ups they liked with him against the UW CBs.  It had some success, but it was clunky and did not help us score.
  • The reverse to DPJ was a great idea.  Why not more of that?  One jet sweep to McDoom was the only other play I noticed that really moved the ball aggressively horizontally.
  • Their CBs were frequently giving healthy cushions.  Where are the WR screens?
  • The empty wildcat was silly.
  • UW's hyper-aggressive LBs (like ours) shoot gaps early in a play.  They attack quickly and aggressively, often arriving in the gaps before the pulling players can get there.  We did not run away from flow nearly often enough in this game, nor did we use horizontal attacking plays to force their LBs to think twice about blizting gaps.
  • UW's TD on the end around is a prime example of exactly what we need more of: a play that punishes the defense for using discipline and reading their keys.  Both guards move towards the RB's lane while the center and weakside tackle move out into space to wall off the pursuit.  It is a beautifully-designed play, and they have scored TDs off of it at least two weeks in a row.
  • In summary, there is not enough deception in our offense.  Whether we are running or passing, we fail to use many of the staples of successful college offenses to confuse our opponents.  Even 3-star athletes are fast enough and strong enough to disable an offense that's predictable.  We don't use RPOs, QB option (we run these but the QB never keeps--a keeper was open for a big gain this week), and we rarely use jet motion, WR screens, throwback screens, or run away from the line action/pull.

I don't want to fire Harbaugh.  To borrow from Craig Ross, I want Michigan to stop playing like Red Coats and start playing like Revolutionaries.

Persistent Underachievement, Part II

Persistent Underachievement, Part II

Submitted by Ron Utah on October 25th, 2017 at 6:42 PM

This is part two of a diary exploring Michigan's offensive woes.  Part one is here, and is basically identifying that there is a problem that goes beyond youth and inexperience.  Part II endeavors to offer some solutions and a path to sustainable success.  There is a TL;DR at the bottom.

Harbaugh's results in his first two seasons could be accurately viewed as overachievement.  Why would anyone, then, spend so much time researching and writing about underachievement?  As was pointed out in part one, the team's success has been on the back of the defense.  The offense finished 2015 at #38 in S&P+ and ended last season at #40.  As noted in part one, our current rank is #85.  The point is that while the team efficiency was good the first two years and is okay now, that is on the back of the defense.  Our offense was middling (for a P5 team) the first two seasons and is now bad.

I suppose that in order for any of this to make sense we have to agree on a premise: Michigan's offense should be elite.  What does that mean?  To me, top 30 every year, top 10 some years, occasionally in the top five.  We are miles away from that, and going from inept (#85) to elite is a leap that will require more than just added experience and talent.  Again, this ground was covered in Part I.

Next we have to agree on some equations for success.  To me, it's a simple formula:

SCHEME + EXECUTION = SUCCESS

Let's break that down even further.  Scheme is the combination of: PLAYBOOK + GAMEPLANNING + PLAYCALLING.  Execution is TECHNIQUE + TALENT + TEAMWORK.  Here are some remedies to Michigan's current challenges, taking these ideas into account:

Masters of None

love this show

Michigan's playbook is vast.  That is probably an understatement.  The number of formations, plays, route combinations, concepts...it feels overwhelming.  As I've said before, it feels an awful lot like Harbaugh's goal at Michigan was to create an offense that had no tendencies and was therefore very hard to predict or gameplan against.  Using formational and play design diversity, the theory goes, the defense has no idea what's coming.

The problem with the unpredictable approach is that your offense can't possibly spend the required time to master those plays.  Successful habits are created through repeated execution of proper technique and teamwork.  This is why teams with lesser talent, like Wisconsin, can execute consistently--they practice the same plays, techniques, concepts every day, week, and year.  What prevents Wisconsin from being an elite program isn't scheme, it's talent.  Their players simply are not good enough to create an elite offense.  But they are damn consistent.

The long-term ramifications of a voluminous playbook, then, are that it stunts the development of mastery of any set of plays.  This means it takes longer for players to get good at the Michigan offense.  Are these limitations offset by defenses not knowing what the offense is about to do?  The clear answer so far is NO.  Last year's offense annihilated lesser opponents who appeared to have no idea what Michigan was going to run.  Against better teams, Michigan's offense stalled, I believe, because they were unable to masterfully execute plays.  Despite scoring more points than we had in over a century, Michigan finished the season ranked #40 in S&P+, a middling Power Five offense that could not move the ball against top competition.  It turns out defenses don't need to know what's coming if they can just react to a play that won't be run especially well.

blocking is hard...can we make it easier?

Every elite college offense has a base set of plays around which they build their playbook.  What are Michigan's?  Many posters have named several plays that we run.  Yes, we run those plays.  The problem is that we also run hundreds of other plays, too, and the most consistest play we've run this year--inside zone--is one that we're not very good at.  Our first play against Penn State was IZ, and Mason Cole--probably our best offensive player--blew his assignment and Higdon was lucky to get close to the LOS. Yes, it probably would have been a nice gain and maybe even a TD if Cole makes his block.  But he didn't. 

Now, this is obviously just one example, but it sure feels representative of our offense's problem: inconsistent execution.  Did Cole screw-up because he's young?  Inexperienced?  Lacking in talent?  Sure, every player is going to get beaten sometimes, but if Mason Cole had been repping IZ for three years and it was an emphasis every week, the chances of him getting beaten so easily decline dramatically.  IZ should, in fact, play to Cole's strengths.

For the purposes of success in this year, and, perhpas more importantly, for repeated success in the years to come, Michigan needs to work on a set of plays that they can master, so that freshmen today are running the same plays when they're seniors.  If we keep changing it up, we'll never master anything.

Jim Harbuagh is far more qualified than I to choose those plays.  That said, he has an obvious knowledge of and affinity for power/gap running.  It's hard to watch Wisconsin and Stanford and not be jealous of their consistent rushing attacks.  But in today's college football landscape, there are plenty of examples of power running games coming out of the shotgun, even utilizing multiple TEs and FBs.  Again, I don't really care which plays are our base set, but I do care how those plays are run.  Adding inside zone this year to an already dizzying array of runs appears to have hurt more than it has helped.

If you're looking for my suggestion, it would be to use power running out of the shotgun.  Having a QB that can run (even someone of O'Korn's athletic level is fine) changes the numbers for the defense.  Running power/option schemes from the shotgun or pistol seems like a way to modernize Hargaugh's bread-and-butter scheme.  Look at Auburn, Ohio State, Clemson...there are plenty of examples of teams using power shotgun schemes.  But if under center is going to continue to be the focus, that's fine too.  Watch a Stanford game and you'll see a scheme I'd love for us run.

In the passing game, the base plays, in my opinion, should be simpler versions of what we run now.  For starters, I do not believe we ran a single slant against Penn State.  To me, this is bonkers.  The slant is a route that even high school WRs can run effectively.  We have run it some the past two years, but this season we've seen very little of it.  We run tons of drags and mesh concepts (more on that later) but not many slants.  It's one of the most effective routes in football, and it's easy to run and throw.  PSU gashed us on several slants.  Why aren't we doing it? It should be a core play in any college offense, again, in part because of the extra real estate afforded by the wider hash marks.  From there, I'd like to see more angle routes from backs and TEs, and more swing passes.  In short, we should have a core package of 3-step drop pass plays.  Oh, and did I mention that slant routes are the best set-ups for fades?

time to stop mocking arm punts

And what about the deep passing game?  For years, this blog has mocked the "arm punts" that have become increasingly common in college football (though I doubt Michigan fans were mocking arm punts in the Grbac-to-Howard days).  But watching the PSU game, it really stood out to me that arm punts might actually be a good idea. 

Trace McSorley is a pretty good QB, but his arm is no better JOK's or Wilton Speight's.  It's probably worse.  The picture above is representative of a poor throwing motion. His deep passes do resemble punts.  So what the hell is going on?  Is it luck?

Arm punts actually help overcome many of the limitations of college offenses.  For starters, the pass is released earlier, decreasing the required time to block.  In college, this is essential, since it seems that every elite team has at least one nearly unblockable DL player, often more.  McSorley's ugly deep passes are thrown sooner and higher, giving the receiver time to find, run under, and catch the pass.  The OL doesn't have to hold their blocks as long.  Against cover zero or even a single high safety, an arm punt can still essentially be a throw into single coverage.  And, as you saw on Saturday, it's really hard to defend an inside fade (fade route from the slot) when the receiver knows when the ball is coming and has time to turn around and adjust, while the defender--even if the coverage is great--still has to get his head around and avoid pass interference.

The point of this long explanation is not to suggest that we should be running inside fade arm punts every play, but that this is actually a pretty damn good solution to a deep passing game in today's version of college football.  Heck, JOK's long completion to Crawford was an arm punt, and it worked for all the reasons explained above.  Michigan needs a deep passing game that: 1) Helps reduce the amount of time the OL has to protect 2) Takes advantage of aggressive defenses and 3) Allows athlete to make plays without having to run perfectly precise routes.

I loved the inside fade concept PSU used, and see no reason we could not execute that play with DPJ or Gentry.  Michigan needs to find a deep passing game that doesn't demand so much precision and uses repeatable concepts that can be applied every season.

What about beating OSU?

This season's offensive RPS is...not good

As noted above, any scheme really can work.  That said, this year's offense--and, to a lesser extent, the previous two versions as well--seems to make it much harder than other teams.  Given that it's too late to focus fall camp on mastering a select set of plays and building counters off of those, what can we do this season (and going forward) to make life a little easier?

Penn State's biggest trick is not very tricky: they spread the field with athletes and force you to pick your poison.  But they did several things against us that were very clever, and we were clearly not prepared for.  The first was motioning Barkley to receive the direct snap and running read option from that look.  The rumor is that Michigan had prepared for that.  If true, it sure did not look like it.  Our safeties should have been crashing down on that motion.  As it was, we only had six in the box, which meant PSU could block every defender.  Which is what happened.

The second was much more clever and even more impactful.  They ran speed option but had their OL run zone in the opposite direction of the play.  Our defense read their keys instead of the ball, followed the action of the OL, and McSorley and Barkley were off to the races against blocked DBs.

Another was the way Barkley motioned.  The "Never put McCray on Barkley" crowd must have been furious, but McCray is not a MLB, and he would have had to play that position (which he has not practiced) if Bush vacated to cover the RB.  This is not as simple as some people suggest.  Don't get me wrong--our scheme should not allow McCray to be one-one-one with Barkley, but the fix requires some fairly significant changes to our defense.

All of these are examples of misdirection.  Using pre-snap motions and hard action that actually took advantage of Michigan's great eye discipline was genius, but not complex.  There is not reason Michigan can't do that.  Harbaugh did more of it at Stanford.  Heck, we did more of it the last two years.  I'm puzzled why we don't do it more.  And I didn't even get into the RPOs, which were also hugely successful.

This is not our savior

I like Jedd Fisch.  The "Good shit, Jedd" gif is one of my favorites.  Interestingly, that was just a play he stole form the Patriots, and was a simple rub route.  But even Jedd Fisch did not use enough misdirection for my taste (and for today's college football).

While many posters argue that I can't possibly know what I'm talking about since the Michigan coaches are so much smarter (the second half of that statement is true), the reality is that the Michigan offense has been moving more towards my suggestions in recent weeks.  Backs are protecting less, and running hot routes more (though it's clear they aren't very good at it, so they probably did not practice it enough in camp).  Our most productive package against PSU was the McDoom pre-snap motion package.  We tired a TE screen (though I hated the play design).

Here's an open question: Why do we use slow-developing mesh concepts instead of the instant rub routes we've seen deployed against us by teams like Cincinnati and Indiana?  Does someone on Michigan's staff believe it's dishonorable to use the blatant disregard of offensive pass interference to our advantage?

My point in this section is to say that all is not lost.  We can make subtle adjustments this season (and keep doing them in the future) that can take better advantage of our talent.  Rub routes, pre-snap motion that isolates LBs on our RBs (remember the smoke screen to Evans against MSU?), and hard action (moving the OL away from the play) are not hard to execute.  Mesh concepts, triangle patterns, 7-step drops--the core of our passing playbook--require much more precision to execute.  Do they work?  Of course.  But only if you're playing 11-man football, which is tough in today's college game, and even tougher when the 11 men have to know 250 plays each.

In the running game, if we're going to stay under center and use FBs every play, let's use more wham, trap, and counter concepts that keep defenders guessing on where the block is coming from (this used to be Harbaugh's calling card).  More deception.  And let's use pre-snap motion there, too, to get the defense to declare their match-ups and help make sure we're in the right play, if nothing else.

The Coaching

Lots of folks are calling for the jobs of Drevno, Pep, and Jay Harbaugh.  While I believe Drevno's OL development can certainly be called into question, and that all of the offensive staff (and Head Coach) deserve some criticism for this year's offense, all of these guys have track records and did not suddenly become idiots.  That said, having so many accomplished coaches might not be a good thing--it may just be adding to the diversity of concepts and plays.  Greg Frey's addition has conincided with the emergence of inside zone--a play that has been a loser for us and certainly took precious practice time from other concepts.

Let me be crystal clear: I'm not calling for anyone's head.  If you asked me to fire someone, it would definitely be Drevno--as the OL and playcalling have both been suspect--but Drev has coached some great OLs to great success (he may not be a good playcaller).

Honestly, it's up to Jim Harbaugh to steer this ship.  As noted above, Jedd Fisch's offense was better, but still not good enough.  Since Rich Rod, how often do we have WRs completely uncovered?  How often to we have running lanes you can drive a truck through?  I'm definitely not calling for the Rod's return, but there is no reason our current brain trust cannot scheme us into a couple of big plays each game, and those plays should not require all 11 players to execute perfectly.  Those QB power fakes certainly didn't, and PSU's RPOs, inside fades, hard action, and pre-snap adjustments weren't new ideas with high degrees of difficulty either.  They were largely the same plays PSU has been running over and over for two years now.  And that's one reason they're so effective.

I have no doubt Harbaugh can pull this off.  But I do doubt our offense's long-term success if the continued approach is to run so many diverse concepts and plays, especially packages that require all 11 to execute precise assignments in order to find success.

TL;DR

The proposed solutions to our offensive morass: 1) Build a playbook around a core set of plays/concepts that are repeated with much greater frequency than anything we are currently running, and provide daily, weekly, and year-to-year continuity that maximizes each players understanding of and ability to execute said plays.  2) Make better use of deception and match-up isolation the way virtually every other program in CFB does.  3) Incorporate some simple concepts that are ubiquitous in CFB to get some more success this season.  Faster-developing plays that require less pass pro are an obvious bonus (ie, Indiana and Cincinnati's rub routes, PSU arm punts, slant patterns)  4) The coaches themselves are all brilliant...but perhaps this staff needs some focus and reorganization (too many cooks theory).  5) You maximize your talent when they can run plays with mastery...this takes time, which means you cannot run as many concepts.  Once some plays/concepts are mastered, defenses have to respect those, opening up counters and wrinkles that are not complex but can result in big plays.  6) Please, for the love of God, don't run PA on 4th and long, or any obvious passing down.  Ever.  (yep, just had to add that)

Persistent Underachievement, Part I

Persistent Underachievement, Part I

Submitted by Ron Utah on October 23rd, 2017 at 7:06 PM

This is a two-part diary.  Part I explore the problem.  Part II explores solutions, because my college professor taught me that it's okay to complain if you offer ideas and effort to improve the situation.  There is a TL;DR at the bottom.

We live in a world that seems to be dominated by absolutists.  Many folks out there brand you a traitor if you criticize the person/group you supposedly support, and others believe that once something goes wrong, it's okay to abandon your alliances.  Any mistake is worthy of permanent banishment and disassociation.  Any constructive commentary is tantamount to an attempted coup.  To all these people I say:

This next part is important, because I'm going to be attacked by both sides of the ridiculous spectrum.  So please, read this even if you ignore the rest: I support Jim Harbaugh as our Head Coach.  I support the University of Michigan and its football team.  I am a fan, and I am not abandoning my team, my coaches, or my players.  I am fully aware of the challenges and limitations our roster presents, and don't expect this year's version to be Michigan's most successful team.

I'm also a rational human being.  There are deep, systemic problems with our offense that go way beyond youth and inexperience.  For those of you saying, "just wait until we have experienced players in our offense, we'll be fine," you are ignoring the data, the eye test, and the reality that if we are not careful, medicore is going to start to look pretty damn good.  Here's your picture:

Misery Comparison

When things are bad, perspective is important.  Are we really the bowler-hat-dog sitting in hell, or are we just irrational fans?  To this end, hard data helps us uncover bias.  Data is never perfect, but it is a good place to start to see if we can trust our eyes.

Michigan's offense is currently ranked #85 in S&P+.  For comparison's sake, Buffalo is one spot ahead of us.  Nebraska is #73.  Purdue #69.  And here's my favorite: Flordia is #81.  Our offense is worse than Florida's.  Please let that sink in.

Don't like fancystats?  I can sympathize.  Those damn things never seem to predict accurately.  So how about this statistic: our yards per play (5.17) is Michigan's worst performance since 2008.  Yep, the 2014 sludgefart mustered 5.32.  Borges cranked out 5.44 YPP in his final season.  And we have yet to face the #6 (Wisconsin), #7 (Ohio State) S&P+ defenses.  Granted, we've already done battle with #3 (Michigan Staee) and Penn State was #9, but we also had the pleasure of facing #93 (Cincinnati) and #109 (Air Force) defenses--the lowest ranked future opponent is Maryland at #77, and everyone else is #33 or higher.  It's not going to get easier.

Our rushing game is bad.  The current 4.08 YPC is our lowest since 2013--the 27 for 27 year (which wasn't even our worst rushing performance).  We are ranked #41 in S&P+ in rushing.

Our passing game is much, much worse.  We are #74 (which feels high) in S&P+ and convert only 32.4% of our third downs.  Guess when we last had a lower conversion rate?  Yep.  2008.

But what about the youth argument?  Well, we are ranked 99th in returning offensive production.  Let's give ourselves even more handicap since we lost our starting QB, and let's compare with our peers.  West Virginia is ranked #123 in returning production.  Their offense is #5 in S&P+.  Ohio State was ranked #122 last year, their offense finished the season at #23.  Clemson is #121 this year.  Heck, even Nebraska is ranked #127 in returning production, and is out-performing our offense.  Getting the picture?

It's impossible to argue that Michigan performing up to the level of their talent, and that is what is troubling.  It's not that our offense is "not great," we are horrible.  And no, saying that does not mean I'm not a Michigan fan.

If we want to build a strong, successful program, we need to make some radical changes on offense.

Specific Objections

Hopefully, the previous section has at least convinced you that there is a problem that goes beyond the simple and lazy explanation of youth.  But what about other quips from the lemmings that say our problems are all the result of inexperience?

  1. Harbaugh has a proven track record.  This is 100% true, and it's why I still want him to be Michigan's coach.  That said, he's never been this bad.  His third year at Stanford--with considerably less talent--his offense was #6 in S&P+.  Did he get lucky with some recruiting hits?  Yes.  But it's a long, long, long way from #6 to #85. In fact, he was even better in his first year at Stanford--#83.  That was with a roster that only compares to this year's Michigan roster in that they had the same number of players.  At the 49ers he led the #18 most efficient offense his first season, #5 his second season, and #8 his third.  Even the year that got him fired, 2014 (LOL 49ers), his offense was #16.  Middle-of-the-pack.  And that's with the parity of the NFL--he has huge advantages over most competitors at Michigan.  It's also worth pointing out that his offenses have looked very different as his OCs have changed--the argument that it's "his offense" is only very partially true.  Stanford played the same brand of manball they're currently playing, and the 49ers ran a different version and adapted to their QB.  Michigan's offense that past two years was very different from this season's version.
  2. We are a pro-style team.  This is simply false.  If you mean that we operate largely from under center and use TEs and FBs, I guess that's true.  But there is not one team in the NFL running an offense that even remotely resembles Michigan's constipated turd.  If you're making this claim, do you even watch the NFL?  It's largely a spread league now, and even the more manbally teams use more misdirection, more spread/match-up concepts, and more quick passes than Michigan does.  No, we are not a pro-style offense, unless your definition stopped keeping up with the league in 1997.
  3. There is nothing we can do with this roster.  This is another absurd assertion.  No one was expecting Michigan to win the CFP this year (well, almost no one) but everyone should expect a competitive offense, no matter how young we are.  We are making basic strategic blunders.  Our pass protection is bad.  It is known.  And yet, we persist with 7-step drop passing plays with deep and slow-developing routes as our only options.  We leave RBs that can't block in the backfield to block, instead of turning them into hot reads and safety valves.  We max protect with our TEs--our best match-ups in the passing game.  We hardly ever run slants.  We don't isolate our athletes in space.  Our version of misdirection is 1980's play action--we don't use motion, deception, or gadgetry even as much as we did the past two seasons.  Yes, the roster is limited.  No, it's not nearly as bad as its #85 ranking.  And let's remember that last year's loaded roster only produced the #40 offense.  And it's scheme was miles ahead of this year's version.  Which brings me to...
  4. Our scheme is fine.  First question: what scheme?  Try to tell me this team's identity.  Try to identify the carryover plays from the previous two years.  College football is a vastly different game than the NFL.  The hash marks allow offenses to create mismatches in an entirely different manner, and misdirection and trickery are far more effective.  Our offense seems to ignore those principles, opting for plays that require 11-man execution instead of match-up plays.  Penn State did not run an innovative offense on Saturday--they ran their offense with minor tweaks.  They ran the same plays over and over and over again, with repeated success.  And their plays created mismatches, allowed the QB to get rid of the ball quickly (even the fade routes were thrown early), and maximized their talent.  This year's Michigan team has no scheme.  We run zone and power in the ground game, and neither looks polished.  The passing game makes no sense.  Our constraint plays are 20-year-old play action fakes.  There is no imagination, and there is nothing to hang our hat on.  

TL;DR

This offense is dramatically underachieving.  Having better players would help, but being a year better will not take us from inept to elite, and that's the jump we need to make.  Part II will explore ways out of this mess.

Updates on Pep Hamilton?

Updates on Pep Hamilton?

Submitted by Dan Man on January 11th, 2017 at 3:14 PM

Slow day on the board, so I'm wondering if there has been any chatter on what's going on with Pep Hamilton.  Is he still considered a done deal?  If so, why do you think there has been no announcement yet?  Should I be worried?  As a Michigan fan, I feel the need to be worried whenever possible.

EDIT: I thought it goes without saying that I saw the hello post.  There still has been no official word from the school or from Pep Hamilton, which seems like an odd delay.  But, I understand, people enjoy piling on whenever possible, so enjoy!