Hokepoints: Carries for All

Submitted by Seth on August 13th, 2013 at 10:55 AM


Running back carry splits: we went over them in last week's roundtable, argued them in the practice tweet post, and then Brian answered a question about it in his mailbag. This horse has been very bad; perhaps it needs one more lesson.

What I'd like to do is illustrate this point of Brian's from the mailbox:

Even if Fitzgerald Toussaint is 100% healthy, someone else is going to get a lot of carries. Michigan ran 502 times last year, and even carry-magnet LeVeon Bell only scooped up 382 for Michigan State. Meanwhile, Toussaint has 130 and 187 carries the last two years. There are going to be 200 to 300 carries, minimum, handed out to other players, and with the situation at quarterback only a handful will be Devin Gardner's.

Last year, yes, but last year was a weird one for rushing and I'd like to see what's really been out there. And here we go:

RB Sharing

(Blig enclickens)

[Jump for splits by depth and discussion]

So the number of carries fluctuates between 595 (Biakabutuka!) and 428 (senior Tom Brady behind an OL playing hurt), centering around 500 most years. Years with great running fullbacks (Jarrod Bunch & Burnie Legette 1990; Chris Floyd in 1997; B.J. Askew in 2002, though a lot of those last were at tailback when Perry was hurt or struggling)

Askew in '01. Scout

ate into what's left for the RBs. Since Askew graduated the fullback carries have diminished to barely visible slivers. Starting in 2008 the QBs became a major part of the running game (I couldn't take out sacks else it would be even more dramatic), and that dropped the RBs' share to around 300 a season. Chart:

% of Carries by Position:

Coach RBs QBs FBs Other
Moeller 78.2% 5.0% 14.6% 2.1%
Lloyd 80.3% 11.0% 5.7% 2.9%
Rodriguez 58.0% 40.0% 0.2% 1.8%
Hoke 51.7% 46.5% 0.0% 1.8%

Hoke was riding out the Denard years and the expectation is a return to something approximating Carr-ian levels since 1) No Denard, 2) Year of the Gardner Protection Program, 3) Decent RBs: we has them, 4) Fullbackian weapons: not just walk-ons anymore. Anyway this didn't seem to matter much or at all to the % of carries that went to running backs nominally below the "Starter":

RB Sharing2

(Clig Embickens)

What does seem to matter is how good the starter is relative to his backups. Sometimes there is no starter or that changes during the year, for example in 2008 McGuffie led the backs with 118 touches with Brandon Minor close behind at 103, with another 85 carries split between Shaw, Brown, and Grady24. Sometimes two guys split it nearly 50-50 (closest to that is 1996 when Clarence Williams and Chris Howard were a 55-44 division). On the opposite extreme are the years when there's a clear-cut #1 running back. In 1999 Anthony Thomas had 87% of the handoffs to running backs, and 66% of total rushing plays because he was that good.

The chart also shows that 2002-2006 (minus 2005 when Hart was injured and Kevin Grady was the new 5-star toy) glut of starter carries for Perry and Hart. In '02 at least Askew and Perry were a tandem. But the more important factor was the backups in those years were mostly David Underwood, Tim Bracken, Jerome Jackson, and a spattering of freshmen who transferred when it was clear they couldn't compete with the top guy. In '07 Hart was hobbly again, but the Minor/C.Brown tandem (90 and 75 carries respectively) were now sophomores and would have likely garnered a goodly share anyway.

Anyway with all those variable considered, your baseline expectation ought to be something like 200-250 carries for the starter, 60-100 for the backup depending on how close he is in effectiveness to the guys above and below him, and another 100 or so to split among the other RB and FB folk.

This Year

So what's the nearest thing to 2013 among the stables since Moeller? I'm going with 1992 despite the returning starters being two juniors instead of one senior because it shows what Michigan has done in the past when an all-world youngster enters a crowded backfield.


The breakdown that season:

Name Yr Carries Yds YPC
Tyrone Wheatley So 185 1357 7.34
Jesse Johnson Jr 155 792 5.11
Ricky Powers Jr 67 374 5.58
Ed Davis So 61 321 5.26
Burnie Legette (FB) Sr 43 155 3.60
Che' Foster (FB) So 14 112 8.00

It takes a lot of squinting, I know. In 1991 Ricky Powers got the majority of carries with Jesse Johnson (107 carries) and freshman Wheatley (86) getting fed plenty. In 1992 Wheatley took over. Imagine Johnson and Powers combine to become senior Toussaint and Green is Wheatley reborn (this is a pleasant fiction) and you get 222 carries for Fitz and 185 for Green and over 100 carries to spread around the other pieces. Unfortunately Steve Everitt isn't blocking for these guys so depress the yardage totals.

UPDATE: The other example pointed out in the comments by whitedawg is 1997, when Chris Howard and C-Will had to compete with freshman A-Train. As dawg mentioned Thomas would have been comparable in recruiting hype (link goes to Sandeep's ancient

'97 A-Train via.

blog) to Green. You've even got Patrick McCall and Ray Jackson to stand in for Deveon Smith and Wyatt Shallman. That breakdown:

Name Yr 2013 Equivalent Carries Yds YPC
Chris Howard Sr Toussaint 180 868 4.8
Anthony Thomas Fr Green 130 529 4.1
Chris Floyd Sr -none-* 59 262 4.4
Clarence Williams Jr Rawls? Euh. 57 264 4.6
Patrick McCall Fr De'Veon 13 41 3.2
Ray Jackson Fr Shallman 11 41 3.7

* If Stephen Hopkins had stuck around he goes here.

Still squinting: Chris Howard was just a decent, non-fumbling north-south guy and nowhere near the home run threat that Toussaint is. Fitz's game is closer to that of original diminutive Cass Tech back Clarence Williams, Michigan's leading ballcarrier in 1996 who had many of his carries taken away by A-Train. Also comparing Smith to McCall isn't fair; Patrick transferred to Denny Erickson's Oregon State (they were pretty good then) and ended up their #2 or 3 guy his junior and senior seasons, respectively.

The point is there are plenty of carries available after the nominal "starter" gets his, even on a team that wants to pass first.



August 13th, 2013 at 3:16 PM ^

Now i am generating $86 an hour or so operating from home. We ended up being shocked when our neighbor informed me she ended up being averaging $95 nevertheless We observe the item is effective right now. Personally i think so much freedom now that Now i am my personal manager. And this I actually do, ~> Bay35.cℴm


August 13th, 2013 at 11:45 AM ^

I have a sneaking suspicion that Norfleet will get more carries now, out of the slot, than he ever would have had burried on the depth chart at HB behind all the moosebacks now on the roster. Motioning him into, or through, the backfield adds another dimension for defenses to adjust to, and I don't think an occasional handoff is out of the question.


August 13th, 2013 at 11:26 AM ^

I think it would be a mistake to not use Garnder's full compliment of skills as an explicit threat to the defense. I'd like to see him get 100 - 150 rushes on the year; call it 20-25% of the rushing share. He's got the ability and making the opponent defend against his legs will open everything else up.

Something like Vince Young, Dennis Dixon, Troy Smith, Donovan McNabb would be appropriate. The pre-RichRod level is way too conservative for what Devin can do.

Also, hasn't there been talk of Colin Kap -- SFO's offensive style? THAT's what I want to se at Michigan this year.


August 13th, 2013 at 11:30 AM ^

You would?  Holy crap man have you seen what's coming into the game if he goes down with an injury?  Pretty much uniformed consensus that he's the most critical piece of our offensive puzzle and not cause he can run.  I agree he could represent that multi-dimensional threat but I'd keep that weapon in the bag this year at least and only ask him to run when he has too cause the pocket broke down.



August 13th, 2013 at 12:23 PM ^

Your point is spot on.  If we lose Gardner, in all likelihood we will lose a bunch of games. Pure running plays from DB should be rare.  I would also hope that in some of the three pre-B1G games not including ND, that Shane Morris gets as much work as possible, although Devin is also in need of playing time.  


August 13th, 2013 at 12:29 PM ^

Oregon (and now Philadelphia) use the QB threat far more than they actually run the QB. Chip Kelly has stated that his offense can move without a running QB, and he proved it by making the national championship game with the leaden-footed Darren Thomas.

So I would like to see read option looks from Michigan with the proviso that Gardner rarely keeps it. I think it will be effective. And then, against OSU, I want Borges to pull the trigger and go for the throat no matter how many times DG runs.


August 13th, 2013 at 1:10 PM ^

Possible? Sure. If you do a pure read every time, the defense has influence, but Chip Kelly had a strong bias toward running the ball with the running back. Read this to get the flavor of it; he rarely wanted the quarterback to pull the ball. Evidence? In their 2011 Rose Bowl winning season, Darron Thomas ran the ball 56 times for a total of 206 yards (sacks included in both totals, I believe). All season long. 

Meanwhile, the team averaged 45 attempts and 300 yards rushing per game. They were okay.

Now, Gardner is a better runner than Thomas, and Fitz isn't exactly LaMichael James, but you can see what Chip was going for. His philosophy, per the above-linked article, was that he wanted the QBs to hand it off until the defense was totally committed to the RB. Even then, he often countered with a different type of play rather than involve Thomas. 

But I think Michigan can use Gardner as a threat while rarely exposing him to tough runs. All it takes is a few, effective designed plays where he keeps and runs per game to keep the defense occupied. Beyond that, the risk is unseemly.


August 13th, 2013 at 12:07 PM ^

I can't remember if Washington was one of the offenses that Borges said he watched, but an interesting thing about RGIII last year is that he only had a couple of designed runs per game.  Just the threat of him taking off around the end seemed to really help his RBs by making the defense cover the whole field.  Devin Gardner doesn't have the speed of RGIII, of course. 

Zone Left

August 13th, 2013 at 12:24 PM ^

You may see that for Ohio State, Notre Dame, and whenever Michigan is in trouble in the second half. Anything else is asking for a doomed season after a bad tackle against UConn. Shane Morris will be really good, but right now he's a stick figure who doesn't really know the offense and needs some coaching work.

The Vince Young / Dennis Dixon offense isn't really possible now absent another huge innovation in offense. Young was one or the first really awesome quarterback with fantastic talent to begin running the zone read concept as a base play. The zone read still works, but not at that level.


August 13th, 2013 at 12:25 PM ^

I hope you actually read these responses. Nobody, not even the coaching staff, is suggesting either that Devin is not capable nor the coaches uninterested in using his running ability to add a dimension to the offense.

As my colleagues have so elegantly stated, the problem is depth. Michigan cannot afford to get Gardner hurt this season. At all.

I would like to see him be used as a running threat, and rest assured that he will not be restrained from the effective scrambles that he did so much damage with last season. But running a midline veer is asking for trouble--one blown knee and the season is basically over.

Also, you used Troy Smith as an example. By his Heisman season Smith almost never ran the ball.


August 13th, 2013 at 1:42 PM ^

I guess I'm just not that concerned about him getting hurt. My concern level is the same as it was when Henne, Navarre, Henson, Brady, Griese were back there. I dont think QBs that run are any more susceptible to injuries than QBs that don't run. I don't propose mine as the definitive study, but it is a study and I am not aware of another that attempts to objectively answer the question re: QB Fragiltiy.


To be clear my position is that the likelihood that Devin will get injured has little (probably nothing) to do with how often he runs the ball. We all agree it would be bad if lost him, I just dont think there's anything we can do to help avoid it. How do we know he won't get lit up while the interior of the o-line develops? We don't. ND's d-line is pretty dope...

I threw out 10-12 runs/sacks/scrambles a game as a target based on what I think is a reasonable split assuming 500-600 total carries by all rushers. In fact, in the 5 games Devin started at QB last year he accounted for at least 23% of the teams rushes. That is exactly what I'm talking about. I really dont think that is a wreckless amount of exposure but ultimately the precise calibration isn't the main thrust of my argument. I'm not suggesting a spread option, but rather FORCING the opposing defense to guard against his legs by deliberately running with him with whatever frequency is necessary to make them honor the threat.

The offenses lead by the QBs in my original comment weren't Spread Option offenses (not sure about Dixon), they were Prostyle offenses with a gazelle under center.That's what we have. They were potent offenses, thats what we want. The notion of playing it safe out of fear that Devin Gardner can't take a hit is laughable to me. Actually, it seems like the Notre Dame thing to do.I bet Al Borges is not afraid to deploy Devin Gardner with his full arsenal of abilities. Michigan is not good enough to win the B1G and beyond by playing it safe; we haven't been for a LONG time.

Ali G Bomaye

August 13th, 2013 at 11:30 AM ^

You used 1992 as your example, but in 1992, Wheatley was a sophomore and the unquestioned starter.  He was a freshman in 1991, behind sophomores Powers and Johnson.  Here's the carry breakdown in 1991:

  • Powers: 240-1197, 5.0 avg., 9 TD
  • Johnson: 107-634, 5.9 avg., 6 TD
  • Wheatley: 86-548, 6.4 avg., 9 TD
  • Legette (FB): 43-213, 5.0 avg., 1 TD
  • All other RB/FB: 25-103, 0 TD

So in 1992, Wheatley got 86 of the 433 carries for tailbacks who played in non-garbage time, or about 20 percent, despite having the highest per-carry production of the three.

A better example might be 1997, with senior Chris Howard and junior Clarence Williams competing with all-everything freshman Anthony Thomas for carries (this was before recruiting stars, but I remember Thomas being mentioned as one of the country's top recruits).  That carry breakdown was as follows:

  • Howard: 180-868, 4.8 avg., 7 TD
  • Thomas: 130-529, 4.1 avg., 5 TD
  • Chris Floyd (FB): 59-262, 4.4 avg., 2 TD
  • Williams: 57-264, 4.6 avg., 1 TD
  • All other RB/FB: 32-124, 1 TD

So in 1997, Thomas got 130 of the 367 non-garbage-time tailback carries, or 35 percent.  I could see this being a similar situation to 2013: Howard was a proven player but not a superstar, and Williams was a shifty third-down type, like maybe Justice Hayes or Drake Johnson is now.  Kerridge/Shallman/Houma might steal 4-5 carries per game, and I would still expect Fitz to get roughly 50 percent of the remainder, just like Howard did in 1997.  That leaves Green with about 2/3 of the non-Fitz tailback carries, which sounds about right.

Plus, 1997 ended up pretty good.


August 13th, 2013 at 11:42 AM ^

Yeah I goofed; fixed. Trying to find a good example is hard and I avoided 1997 specifically because I didn't want to equate Howard (a decent back) to Toussaint. Fitz actually might make a better comparison to C-Will.

Zone Left

August 13th, 2013 at 12:29 PM ^

That's an interesting example, because by the Ohio State game Thomas was getting the lion's share of the carries, if I remember right. I've got a mental picture of him slamming into the OSU line over and over again in the fourth quarter as Carr executed the "I've got the 1997 Michigan defense here, punting is winning" strategy. He kept at it for nine more years, but it never worked quite as well.


August 13th, 2013 at 11:30 AM ^

Obviously, we want to see Fitz return to Fitz of two years ago. We also want to see others emerge from the pack to help defray the cost of toting the rock, effectively,  of course (meaning elusiveness or toughness by max-plus yardage after avoidance or contact). 

The angst  or uncertainty of the season is seeing that process unwind. And we all have favorites for that player emergence. And so this is why this issue is so significant, because right now it seem like Fitz and a big if, and that's largely because, while better, Fitz himself is still an if. 

Truth is, there is real opportunity here and the players know it. That also makes it part of the angst for us, not knowing who that guy will be, and how that competition is going. I don't think we are going to see any first game carries going for six like we did a few years ago when No. 16 blew the doors off Michigan Stadium before Mike Patrick could finish his story on why he was even in the lineup before becoming a zig zag blur in the endzone. 


August 13th, 2013 at 11:50 AM ^

In 1998 we had A-Train as a sophomore and proven back, with big-time recruits Walter Cross and Justin Fargas coming in. Seems relevant.


August 13th, 2013 at 11:55 AM ^

Where in the hell do you get your raw data to make those charts?

The "% of RB Carries by Depth" chart requires detail down to individual and their place on the depth chart each year.  That's way more than standard box score detail.

Don't get me wrong ... I love this stuff ... I'm just curious where this level of detail is coming from.


August 13th, 2013 at 12:17 PM ^

1. Bentley Database.

2. Using my memory I went through each guy and added his position.

3. Realized I could do a vlookup against Bentley roster database to make #2 much faster. Doh.

4. Sorted by position and # of carries.

5. Went through each year and made any necessary changes based on injury to starters, e.g. Mike Hart in 2005.


August 13th, 2013 at 12:21 PM ^

and this happened in the championship season, the big thing for any starter was the ability to handle blocking assignments. So, even though he was a freshman that year, and got some carries, Anthony Thomas wasn't a primary threat. In fact, he was the last guy off the bench in one of the greatest wins ever at Michigan Stadium against Ohio State, because the upper classmen Howard and Williams were both nicked up and Floyd scored the lone offensive touchdown. Thomas played because he had to, and he performed well enough. 

That year, Chris Howard, Floyd and Clarence Williams took the bulk of the carries. And each played primarily to give the team a different running option, speed, power, and  first down yardage producer. But carries came with caveats: a no-fumble pledge and first priority blocking duties. The ability to keep Grese upright was paramount even with a stellar Oline, which Michigan had. 

That same situation exists now, a potential leading ballcarrier with younger guys in reserve whose carries will be dictated by their ability to move the chains, but not cough up the football or know all the rb blocking assignments. Keeping Gardner upright and untouched is again paramount.

There is Fitz and Green who we envy as his backup but is a freshman, and subject to all the foibles of first-year play, which under the Schembechler-molded thinking of Michigan coaching meant, you play as a necessary evil, not a weapon. Freshmen are best when mostly seen on the sideline and not played until they prove it. Hoke is from that school of thought. Just saying.Because he hasn't played freshmen in the backfield with regularity while here. Largely, obviously, because there was no need.