This Week’s Obsession: Can We Keep It?

Submitted by Seth on November 8th, 2017 at 10:30 AM

THIS ARTICLE HAS A SPONSOR: If you’re at that point where you’re thinking further out than the Maryland game it’s time to start a conversation with Nick Hopwood, our MGoFinancial Planner from Peak Wealth Management. This conversation doesn’t have to be about your financial future and security; he’s also willing to discuss power/counter run game futures.

Our deal is Nick is the guy I go to for financial strategies, and he gets to ask us Michigan questions on your behalf. Anytime it’s a Nick question, we’ll let you know. Anytime you’ve got a financial question, let Nick know.

-------------------------------

Legal disclosure in tiny font: Calling Nick our official financial planner is not intended as financial advice; Nick is an advertiser who financially supports MGoBlog. MGoBlog is not responsible for any advice or other communication provided to an investor by any financial advisor, and makes no representations or warranties as to the suitability of any particular financial advisor and/or investment for a specific investor.

--------------------------------

image

[Bryan Fuller]

Nick’s Question:

How sustainable is the run game?

The Mathlete: My first thought is who else is running the ball like Michigan has lately and how have they fared against the best defenses. Because that is really the question. There is little doubt that the run game will be sustainable against Maryland. Wisconsin and Ohio State are different questions. So far this year Michigan has 3 games of 200 yards and 6 YPC with sacks and garbage time excluded against Power 5 opponents. Five other teams can say that. Arizona, Notre Dame and Louisville have elite running QBs and aren't really considerations. Oregon and Oklahoma both utilize the QB in running game effectively, so maybe not them. Right there, Michigan is doing something that isn't easy to do without a QB involved running game.

Last year Georgia and Stanford each accomplished the threshold four times and are probably our best comparisons. Against playoff bound Washington, Stanford was completely shut down for <1 yard a carry and McCaffrey didn't crack 50 yards for the day. Georgia barely cracked 1 yard a carry against Florida and survived against Auburn with 3.1 YPC.

This isn't by any means an exhaustive or complete list, but the road is uphill for a rushing attack without a quarterback run threat against top level defenses. (edited)

[Hit THE JUMP for the rest of the responses]


--------------------------------

David: This answer is time dependent. There are many variables.

This week: Very. Maryland's defense is not great. Looking at their advanced stats, they're 101st in total defense, 59th in rush defense, 70th in Adj Line Yards, but 22nd in Rush IsoPP. So, they'll give some stuff up on the ground, but not in huge chunks. I'm guessing this weekend's game will look rather similar to the previous couple of weeks. Hopefully with a little more passing, though?

Rest of season: Ummmm...I think it could still be okay. Michigan moved the ball decently well against MSU and PSU -at least at times- and both of those teams have pretty good rushing defenses. Both Wisconsin and OSU are similar to Penn State and Michigan State, respectively, in their rush defense numbers. The key for these couple of games will be the development of any semblance of a passing game. I do not believe Michigan will need 300+ yards in the air...but they will need a couple of deep shots completed. Or at least a couple of preferable matchups exploited. That has not happened, yet, but we do not know for certain that it cannot, as the Peters passing game has not been revealed in much detail. I think there is still some potential there for November...if pass pro is not complete butt.

Next Season: Definitely. The key for me all season has been the development of the guards. Both Bredeson and Onwenu have grown a lot and gotten better, Onwenu especially. They should be the cornerstones of the line, next season. We're presuming that Ruiz will most likely take over at the center position and he has also looked formidable in run blocking. Michigan will need Newsome to come back from his horrific injury, and they will need to hit early on one of their red-shirting tackle prospects, as well. If those things happen, the run game could be sky high, next season. If they're still searching for a RT and Newsome doesn't make it back or isn't close to what he looked like he was becoming in 2016, expectations could drop a bit. On the plus side, as well, TE blocking from McKeon and Wheatley has also been trending to the good and I would believe that will continue given each's age.

Beyond: I am still in the Very Much Yes camp. Track records speak for themselves, man.

  --------------------------------

Ace: It depends on what you’re hoping to sustain. The blocking is visibly improving on a weekly basis, and that continued against Minnesota despite a true freshman taking the place of one of the line’s two best run-blockers. The running backs have shown that, when given the space to make plays, they can indeed make plays. That all looks quite sustainable. Michigan doesn’t lack for physical talent on the line or in the backfield; it’s just taken time for them to download the system to the point that one or two players biffing a play doesn’t bring the whole thing down.

The question is more about the passing game, really. If Michigan wants to succeed against Wisconsin and Ohio State, they going to need the opposing secondaries to respect the pass. Before Brandon Peters replaced John O’Korn, a lot of the improvement in the run game was masked by opposing safeties going hell-for-leather after lining up eight yards off the ball; even perfectly blocked plays can only go so far under those circumstances. Peters has shown the competence required to get the Minnesotas of the world to back off a bit. He’ll probably have to show a little more or the Badgers are going to load up against the run and force a redshirt freshman to beat them behind a shaky pass-blocking line.

--------------------------------

Seth: I'm reopening a chart I made last year on Wisconsin's rushing offense and updating it with 2017 data. The Badgers are a year ahead of Michigan in the power running game rebuilding process. Here’s their rushing performances (by sack-adjusted YPC) since 2016 plotted against the opposing rush defenses (by run defense S&P+):

  image

On the macro good run defenses are better against the run (duh!) but there's some HUGE game-to-game variation, enough where you might start to think there’s something to the old adage that an offense that can block power well can get you 3 yards consistently. When you squint at that chart and apply what you know about those teams the thing that stands out is the 3- or 4-YPC line were statistically good pass defenses. You’ve got LSU, Iowa, MSU, Michigan, a couple of Northwesterns, and then some bad defenses who played aggressively (in those games at least—looking at you, Illinois). The teams that liked to keep a safety back were a yard per carry higher.
When you run that against the overall defensive S&P the R-squared almost doubles:

image

It's not scientific but two things are apparent: 1) Single-game performances can vary dramatically, and 2) Run defense isn't necessarily more meaningful than simple defense.

So I agree with Ace: running the ball isn't just about not screwing up your running plays, but also having the ability to keep the secondary the hell away from them. Against Penn State Michigan's blocking was fine, but the safeties were able to play so aggressively that all the runs making it through the front seven were meeting defensive backs at 2-3 yards. Minnesota and Rutgers can't trust their safeties to play that way--both because Peters and because they don't have good safeties—and Michigan ran all over them.

--------------------------------

Adam: I think it's moderately sustainable. As mentioned above, probably not going to have much of an issue running the ball on Maryland. Yeah, they have a pretty good linebacker in Jermaine Carter and corner Antoine Brooks has some gaudy run stop numbers, but a corner with gaudy run stop numbers makes their run defense look that much more pliable. 

This comes down to what they can do against Wisconsin and Ohio State's defenses, which are generally better on a unit-by-unit basis than the ones they've been running roughshod over. Wisconsin's linebackers and secondary are going to pose a challenge even if the line and fullbacks continue to improve their blocking precision. Ohio State's run defense gave up 243 yards on 38 carries against Iowa, but that was behind a 7.6 YPA, 32 throw performance from Nate Stanley. If the sustained rushing success comes down to how they do against Wisconsin and Ohio State, I agree with Ace: it's really a matter of how the passing game performs. Peoples-Jones continuing to build on the progress he's made is big, as is the reliability of McKeon and Gentry. I think what looms largest with regard to the passing game is the injury to Grant Perry.

--------------------------------

Brian: Fairly sustainable. Michigan isn't going to run for ten yards a carry the rest of the season, but they've been an arrow pointed straight up this year since they gradually de-emphasized inside zone and inserted Juwann Bushell-Beatty at right tackle. Michigan moves people, and they've become extremely proficient at power football. UFR grading has spat out good to very good blocking numbers for a month now; this isn't a flash in the pan.

It's only "fairly" sustainable because of the passing game. Michigan gave up three sacks on just 16 dropbacks in the Minnesota game and while Brandon Peters looks like an improvement over John O'Korn the training wheels are very much on. Better defenses will make Michigan pay for that with safeties at eight yards; Rutgers was content to bend and Minnesota's corners so bad that it didn't matter. Against Wisconsin and Ohio State, it will matter.

Comments

wayneandgarth

November 8th, 2017 at 10:22 AM ^

Obvious to most but bears repeating (as was stated in more than one comment above) that sustaining the run is also dependant on growing a competent pass game - one which is dependant probably more on the OL and WR than Peters.

tnixon16

November 8th, 2017 at 10:46 AM ^

But you also have to factor in a young (and depleted) receiving corps that has to know its route trees...and get open. That’s three iffy factors for our pass game — iffy line, frosh QB, inexperienced WRs — which adds up to inconsistent pass game.

Which makes this run production more sustainable NEXT year than it does this, IMHO.

grumbler

November 8th, 2017 at 12:18 PM ^

What you say is true, but I'd point out that the passing game only needs to be threatening; it doesn't need to be consistent.  The key is making it too risky for teams to bring up the safeties, because a 1/3 chance of passing success (i.e. to WR on deep routes) will still mean lots of TDs if the Corners are 1-on-1 and left there.  Scheme might buy Michigan that much, if Peters is for real.

BassDude138

November 8th, 2017 at 11:42 AM ^

I know we usually say this type of thing during the out of conference schedule, but I feel like there may actually be some things, passing wise, that they have for Peters that we will see in the last two games. All they've asked him to do was complete a few easy throws in his first two games, then let the running game carry the way. He hasn't really had to show anything yet, but there has been a few well designed pass plays that have worked really well. Rutgers, Minny, and MD are all games that are winnable by just wearing them out on the ground.

I am holding out hope that they are developing some passing schemes in practice with BP, as I'm sure they aren't just practicing him handing the ball off. By the time we see Wisconsin, Peters will have had a solid four weeks of practice as the first string, so I really do expect to see some new things in the last two weeks. Not expecting to see a completely different offense, but at least open things up a bit and have some passing plays that will keep the UW and OSU defenses off balance a bit in order to continue pounding the rock.

Fezzik

November 8th, 2017 at 12:30 PM ^

I think it will be hard to scheme much without good pass blocking. If we are working on something for Peters I hope its west coast style quick passing. Being able to sprint out rather than sit in a pocket could help a lot when we are not good at forming a pocket for our QB to sit in. 

dragonchild

November 8th, 2017 at 12:45 PM ^

You don't get a safety at 8 yards to go farther out with a 5-yard pass.

The way to do it is PSU-style high arcing arm-punts.  Those go out with the speed of a slant to mitigate the pass blocking issues, but they're still long passes that punish Cover Zero.  The problem is you gotta complete those, and they're much more refined than chuck it up and pray.  Can we do that with a new starter and freshman receivers?  PSU, we ain't.

Fezzik

November 8th, 2017 at 10:20 PM ^

This season we are not able to throw/catch jump balls. Therefore this is currently not an option. The fade route is essentially a jump ball route depending where the defender is playing you and where your QB decides to place the ball. It is a short drop, deep ball which should negate pass blocking. But how many of these have we completed?

You can totally still throw quick passes against safeties playing in the box, especially the rollout passes swing passes west coast offenses heavily incorporate. Slants would be difficult but we don't seem to run those anyway. A safety at 8 yards deep has a more difficult angle sprinting towards the sideline to make a tackle than a safety at 12 yards. At this point of the year I'm not as worried about taking the top off the defense as I am being able to have any passing attack. I can't think of a better idea than high percentage, fast developing pass plays.

jgoblue11

November 8th, 2017 at 10:37 AM ^

I do believe that Peters was limited against the Gophers due to the weather. The running game got going really quickly and as this was discussed before, there was really no need to pass. We were breaking 50 plus yards with no real threat to stop us.

I think, if the weather permits, we will see BP open up and have a real solid game  through the air against Maryland. I don't think Jim is going to flash all of his cards though. He will limit BP if the run game is strong again. It will be interesting to see what coach Harbaugh has prepped for Wisconsin. If we do not develope a pass game, teams such as Ohio State and Wisconsin will soon shut down our run game.

Maryland will be a great 2nd start for Peters. I hope Wilton is ready to go at least by the Ohio State game just in case.

jmblue

November 8th, 2017 at 1:07 PM ^

I was in attendance and really didn't think the weather was a factor.  The temperature was about the same as in the Rutgers game, and the mist (which is all it really was) didn't seem to affect anyone's hold on the ball, nor did people seem to have trouble with their footing.  I guess our field drains pretty well.

 

 

Blue In NC

November 8th, 2017 at 2:34 PM ^

I think it also sets up an interesting dynamic in the OSU game.  For years and years, it has been OSU's dominant running game against Michigan's terrible running attached with some hope of pulling off an upset with a passing frenzy.  For games in late Nov, that's not a great receipe, Bob.

This year that story seems to be different.  We are certainly less explosive on offense but we may actually have the better power running game.  That could be an advantage in a late Nov matchup.  

BayWolves

November 8th, 2017 at 10:46 AM ^

You're probably right about not wanting to show too much of the passing game if it can be avoided yet harbaugh himself admitted that you can't keep Peters cold and out of rhythm by not throwing any passes in the game.  He absolutely has to pass more against Maryland.

Also, I don't see anyone being successful by hiding their entire playbook until they meet an arch rival.  Teams get better by playing their playbook and doing it often enough to perfect it.  I am just not down with this hide 90% of everything until we meet OSU in late November.

jgoblue11

November 8th, 2017 at 10:54 AM ^

Yeah you are absolutley right. I am maybe just saying that Harbaugh is possibly going to want to keep Peters "safe" against Maryland and not try to over due it. If the run game is on point against Maryland, do you think we will see another game like Minnesota? Or will Jim have Peters prove that he can throw? I'm not sure. I think the Wisconsin game is going to be a very telling game for Brandon Peters.

buddha

November 8th, 2017 at 1:35 PM ^

I like your line of thinking mostly because it aligns with what I am hoping is going on too! With the run game the past few weeks- and, predictably - with Maryland this weekend, I think the offense has been very vanilla. My hope is that the offense has been vanilla because it really hasn't had to be anything else! Meaning, our run game has eaten up opposing defenses, and we haven't had to "show" much.

If we see some new looks and a couple wrinkles in the offensive and - specifically - the passing game against Wisky and OSU, I'll be pretty stoked. I just have to believe Harbaugh and the offensive staff are coming up with something special for those games...and - if not - sorry liver.

Kevin13

November 8th, 2017 at 11:11 AM ^

is going as well as it is, there is no need to pass the ball. Just keep running it down their throat and moving the chains for an easy win. I too don't think Harbaugh is going to show a lot when we are simply eating up yards on the ground. Peters is getting more confidence with every game and more comfortable with the offense. I would expect a little more passing this game and another good performance by the offense will have us in good shape to face Wisky.

jdemille9

November 8th, 2017 at 1:22 PM ^

I think Peters gives us a better chance to beat OSU than Speight (though I don't think we have much of a chance at all).. Speight's been out for a while and is likely very rusty. Not to mention the games he did play he wasn't looking very good anyway. Unless Speight comes back to mid-2016 form then I don't see why they'd go away from Peters.

Icehole Woody

November 8th, 2017 at 10:52 AM ^

We've seen Brandon Peters hand off the ball, throw short, scramble and throw short, and pull it down and run for the 1st down.   The weather in College Park on Saturday should be fine so lets hope we see him throw deep with the same success he's had everywhere else.

Go Blue!

Old_TBone

November 8th, 2017 at 10:52 AM ^

is so forward looking that the posting time is an hour ahead of real time.

So, do you work on passing downfield at Maryland or do you pound them into the ground and reveal little on tape... hoping to sucker in WISCY or OSU into overloading the box?

jgoblue11

November 8th, 2017 at 10:58 AM ^

That is what I am wondering. Do we pound Maryland, and try to keep our guys healthy in time for Wiscy on the road? If we have no reason to put peters in a lot of "sack" situations, do we just ground and pound to get to Wiscy healthy, and then really open it up? Man, I don't know, but it feels good to have this game right before a big one, to get Peters and the team confident going into the next few weeks!

Blue Mike

November 8th, 2017 at 11:15 AM ^

It seems like it would be more beneficial to show off to Wisconsin and Ohio State that we do, in fact, have a passing game they have to prepare for than to hide it. Also, it seems like it would be better to rep the passing game in the real world before trying it against a tough defense.

I'd love to see a game go somewhat like last year's UCF game; throw early, throw often and keep throwing even when up big. Peters needs reps, the line needs reps, the receivers need reps. Go get em!

Icehole Woody

November 8th, 2017 at 11:06 AM ^

I was wondering about that as well.  A lot of stuff has to happen to complete a long pass.  I have to believe they will want to try it out with the young QB, WRs, and sometimes scary pass blocking.   Wisconsin and the Buckeyes are going to load the box and go after Peters at any rate.

SteveInPhilly

November 8th, 2017 at 11:16 AM ^

I don't think the vanilla passing offense can be attributed to holding back any of the playbook to preserve the element of surprise. I think it has more to do with some combination of preventing sacks to keep Peters healthy and limit mistakes (of Peters, WRs, OL) so that we can win games. 

I think the injury prevention is a reasonable goal, but I wonder whether the potential growth of our passing game is being limited by not pushing its bounderies, which could harm us against Wiscy and OSU. Tough to say that definitively though, when every game counts and you don't want to risk dropping one to Minnesota with a bunch of turnovers. 

Glennsta

November 8th, 2017 at 12:55 PM ^

At some point, you have to work on perfecting enough vanilla passing plays at game speed against MD to convince UW and OSU that they can't load the box on every down.  If you want to get fancier, it depends on confident the coaching staff is that the offense can execute well enough to be successful.  I'd save some wrinkles for the last 2 games but obviously, the more you can show to UW and OSU for them to be concerned with, the better.

Blue in PA

November 8th, 2017 at 11:17 AM ^

We aren't going to run wild against WS or osu, that's just not going to happen.  I hope we can have enough balance to move the chains and put up a few points.  The O has looked much better lately, against pretty sorry D.

 

GO BLUE

Double-D

November 8th, 2017 at 3:50 PM ^

Our run blocking right now is clicking and defensive lineman seem a bit confused against us. It would not surprise me to see us run well vs Wisconsin and OSU. Our play calling seems to have much better flow right now.

It’s interesting how we are struggling with pass pro although I think the improved run blocking will assist in that regard.

MSU is just the opposite. Their young guys on the line can’t push anyone off the ball but they have moved their feet well keeping rushers in front of them. (With a little grabbing being overlooked).

MGlobules

November 8th, 2017 at 11:33 AM ^

are what is going to be on most people's minds. (The days of Big Eight Little Two are over, but there's no reason to think that Harbaugh isn't fielding highly competitive squads going forward.)

The coaches have to be plotting the Maryland experiment with the run-pass balance of the last two games firmly in mind; it will be interesting to see what we get on Saturday. With a presumptive win, I'd be tempted to fling it all over the field. Unless we really looked bad doing it, that would hopefully strike some fear--and require a lot more planning--over in Madison.

UW is an eminently beatable ball club, but we are playing them there, and I think most college teams tend to play down or up to the moment. Some calculation is also required if you feel you have to save something for OSU, but I'd just play lights out for that UW win and let the devil take the hindmost with OSU at home. The boyos will be pumped for OSU at home anyway. 

Should be some fun encounters, and the last two should be nail-biters. It might be too subtle for some fans, but I'd say that if Harbaugh has the kids in both of those games at the end I'll be happy with how he handled this year, losing Speight, the youth, etc.

linnymtu

November 8th, 2017 at 12:14 PM ^

You do realize that both of those show there is virtually no relation between the YPC and defense rank right?  Whether it is .04 or .07, both are saying there was next to zero correlation.

dragonchild

November 8th, 2017 at 1:16 PM ^

This is one of those cases where Mathlete and Seth were duly diligent with gathering and analyzing data, and then concluding there wasn't enough to make a meaningful prediction.  Sometimes, that by itself is saying something.

For Michigan it means there's hope, if no certainty.

Michigan Arrogance

November 8th, 2017 at 4:06 PM ^

yeah, I was about to say - I'm not a stats guy by trade, but I can tell you this: had I produced that graph in graduate school (astrophysics and then secondary ed), the thesis committee/prof would have asked why I bothered to even draw a line thru that clearly random pattern of dots.

 

 

yossarians tree

November 8th, 2017 at 12:27 PM ^

What happens against Maryland is somewhat immaterial because Wiscy and OSU are both going to load up the box with run and pass blitzes and force Peters and his receivers to beat them over the top. That is etched in stone. So I think they are going to try to get the pass game a bit of confidence this week as long as they feel comfortable with a (presumed) lead. But let's not forget that Harbaugh is from the old-school "defense and run the ball" philosophy that leads to many games that are closer in score than the fans are comfortable with. In those two big games at the end of the season, as long as the score is close, we are going to see a lot of the run game and lots of punts.

stephenrjking

November 8th, 2017 at 12:29 PM ^

I am confident that the OL is gaining steam and improving. I believe the RBs, particularly Higdon, are coming into their own.

But the OL won't be able to move those OSU DEs nearly as far, and if the safeties are charging into what holes that exist, the improving quality of the OL and the RBs won't matter.

Michigan's passing is an issue, of course. But the passes that matter are the passes that take place on plausible running downs. Peters hitting or not hitting a receiver on a 3rd-and-15 pass isn't going to affect the safeties when it is time to run the ball. The ability of Peters (or, as looks quite possible, Speight) to punish safeties for lining up close on 1st and 2nd down plays is what is going to open things up for the running game. 

Pass protection makes four vertical type plays less likely to succeed, which means play action will be the order of the day. I don't know exactly what plays will best hurt teams with safeties close to the LOS, but hopefully the offensive coaches do. Execute those and either we move down the field picking on safeties, or they adjust and there is room to run.

*BTW, we've totally understated injuries as a factor in the passing game. Comparing last week with the beginning of the season, we were missing, by my count:

  • Our top three wide receivers (Black, Crawford, and Perry)
  • Two of our top four TEs (Wheatley and Eubanks)
  • Our starting QB
  • Our right guard (and if you want to count Newsome, one of our starting tackles as well).

That's a lot of personnel to be missing. Receiver was a deep position, yeah, but we're missing a ton of guys that catch passes right now. That's not nothing.

taistreetsmyhero

November 8th, 2017 at 1:00 PM ^

The Mathlete was the only one who made a convincing argument. It also jives with our run game performance (or lack thereof) last year against Wisconsin and OSU.

My prediction is pain

wolfman81

November 8th, 2017 at 9:11 PM ^

Linear regression is the wrong tool here.  The dependant data (defensive rank of any sort) is not a continuous variable, or even an interval variable.  This is not true about defensive rank data.  For linear regression to be appropriate, these things would need to be true:

  • The difference between the #1 and the #2 defense must be the same as the difference between the #10 and #11 defense.
  • The #20 defense must be twice as bad as the #10 defense.  (Or a similar statement given the way that you've inverted your scale).

As I see it there are two ways to repair this analysis:

  1. Use the raw S&P+ score instead of the rank.  Of course all of the rules of linear regression should be followed when performing this sort of analysis.  (For example, what do the residuals look like?)
  2. Use an ordinal logistic regression.  (I don't believe that Excel can do this without additional software, though I have never tried.  This is a tutorial about ordinal logistic regression using the free statistical software R.)

PurpleBeaverEater

November 8th, 2017 at 1:16 PM ^

Yeah, we definitely need to be able to complete some solid passes and utilize the passing game to move the football a little more against the more elite defenses like Wisc and OSU. That being said, we can do some good for ourselves with some trick plays, but also mostly just keeping it simple. Short passes, quick outs and reads, some screens. Just enough to keep the defense honest. Really, if we were able to complete just one or two balls over the top in either game, I think it would go a long way to opening up the run.

Also, we need to have a Zach Gentry double pass in the books for OSU.

The production will obviously go down against teams like Wisconsin and OSU, but I think the run will still be a serious weapon. That DL of OSU is pretty scary though.

kehnonymous

November 8th, 2017 at 1:28 PM ^

You'll have to squint pretty hard to find scenarios where we move the ball consistently against the cooler poopers, but FWIW their LB corp didn't seem particularly inclined to cover fullbacks or tight ends last Saturday so that's a possible point of attack.  I'm sure the OSU team we get won't look much like the one we saw in Iowa but I'm betting almost every single knowledgable Buckeye would tell you there's a pretty huge difference between Raekwon MacMillan and Not Raekwon.

Ziff72

November 8th, 2017 at 1:29 PM ^

I know people don't want to hear this, but next years RT is going to be JBB.  The guy is playing like a beast in the run game.   They are not going to throw that away.   If he's not your RT you can be very happy because the guy that replaces him will have to be a stud.