The Decimated Defense, Part II: the Statisticating

Submitted by Seth on November 2nd, 2009 at 8:18 AM

[Editor's note: holy hell, man. This is like a PhD thesis.]

It hits you like a thousand knives stabbing you all over your body. You can't breathe. You can't think. At least, not about anything but the pain.

titanic1

I'm searching for a metaphor.

Amidst the phantom flags and the Angry Michigan Hating Bounces and the dropping of babies on 3rd down on Saturday you could not possibly have missed a notable lack of competency in the 11 guys tasked with making sure the other guys score less than we do, otherwise referred to as "Michigan's Defense."

This is Part II of the afore-bumped diary "The Decimated Defense," a look at what has happened to turn Michigan's once vaunted defense into..

I don't have a metaphor...

Something that has a lot of really shiny beautiful parts, that ostensibly looks like something grand and wonderful, but like with some major defect or hole in it, from which pours in death and destruction...

titanic2

In the wake of, well, that, I'm sure that you, as I, need to understand what happened to Michigan's defense, how we got here, will it get better, and can it be avoided again?

In the first of this series, I went through Michigan's last five classes to see if we could find where and what went wrong in defensive recruiting to lead us to a day when Jordan Kovacs was all that stood between the program and the bottom of the sea. We looked at the cheap rivets, the lack of safety training, and missing life boats, while Brian UFR'ed a really big iceburg.

It was long, and mostly stuff you already knew, and at one point you had to fix yourself a sandwich, but at the end we identified two factors that were very likely contributors:

  1. Small classes
  2. High attrition

Today we put that in context. I compared the current rosters to the recruited rosters of Michigan, Ohio State, Notre Dame, Penn State, Michigan State and Alabama, to see how each of these teams were built, and what was lost along the way, in order to understand why should so many other luxury liners and loveable tugboats and whatnot stay dry as we face a watery grave?

(Excel spreadsheet lives here.)

Recruiting: Quantity and Quality

PREGAME
Lon Horwedel | Ann Arbor.com

Rose: The fall alone would kill you.

Jack: It would hurt. I'm not saying it wouldn't. Tell you the truth, I'm a lot more concerned about that water being so cold.

Here's how Michigan stacked up in pure defensive recruiting from 2005 through 2009 (Rivals ratings used):
 

  Michigan Alabama MSU Notre Dame Ohio State Penn State
2-stars 2 8 24 1 4 9
3-stars 20 34 30 22 20 18
4-stars 23 37 10 19 32 23
5-stars 3 4 0 1 3 2
TOTAL 48 83 64 43 59 52

This counts every recruit that came in ready to play defense, except athletes who played their entire careers on offense. It also includes offensive recruits later moved to defense. It excludes walk-ons.

Many nuggets here. Let us bullet:

  • Notre Dame fans who blame recruiting for some of their woes have a beef. Their classes have been highly ranked, but even smaller than paltry Michigan's!
  • Bama LOL
  • Michigan and Penn State recruited pretty similarly. The big difference was that PSU brought in 7 more 2-stars.
  • Michigan and Ohio State both recruited 20 players of 3-star caliber, and 3 blue chips, but OSU had 9 more 4-star players during that time.
  • Michigan State clearly isn't in the same recruiting league as these others. They're basically averaging one lower star per recruit
  • ...but out of a respectable class size.
  • Even so, Alabama had more 3-star defensive recruits over this time than Michigan State.

Michigan's closest recruiting analogue here is Penn State, with the high-end (4-5 star) recruiting separated by one more 5-star guy for us. You can call Notre Dame basically a Michigan-light. If anything, the Fighting Irish have been even pickier about talent than U-M, OSU and PSU, except they haven't been as successful at reeling in the big fish as any of the major schools profiled.

For these schools, the distribution seems weighted slightly toward the top, but their bell curves are only slightly ahead of OSU and Bama. However, when placed beside each other, it's easy to see how large amounts of recruits can generate a much more sizeable talent pool from which to draw starters.
Figure 1
So recruiting tells a story, but certainly not the story. Certainly, Alabama and Ohio State recruited the most 4- and 5-star players, and subsequently have great defenses.

Michigan and Penn State should, just going by recruiting, have about the same level of defense, with maybe one more NFL-bound player in Ann Arbor, and maybe a bunch of 2-star guys backing up at Penn State instead of Michigan's walk-ons. Or it would be, if attrition was constant. We will see in the next section that it isn't. But you knew the problem wasn't just recruiting, anyway, since you know that Penn State's defense is legitimately good, and Michigan went into this season steering a pre-WWI luxury liner.

First, though, while we're on pure recruiting, let's look real quick and see if it's actually the age of the recruits that matter. Since they should be theoretically the heart of a great defense, and since the distribution among all schools except Michigan State was fairly equal when it came to 4- versus 5- stars, let's just look at those two groups, and when they came in for each school (MSU left out to spare them the indignation of looking like Antarctica):
 

  Michigan Alabama Notre Dame Ohio State Penn State
2005 4 3 0 6 3
2006 6 5 2 6 8
2007 4 8 4 8 6
2008 7 15 10 6 4
2009 5 10 4 9 4
4-Star+ 26 41 20 35 25

4068065533_9e415636c7_o
[At this point I would ask everyone else to pause for a moment while we give Irish, who has been waiting patiently all this time, an opportunity to assign righteous blame on Ty Willingham. HE did this, precious!]

Okay, so other than an '05-'06 "Domer LOL," did we get anything out of this?

Penn State's great defense has a lot of high-rated juniors and seniors on it -- more than any other school. Michigan was kind of even, but actually should have had more upperclassmen than Bama or Notre Dame. Ohio State has been strong all the way through. Alabama is going to be really really good in a few years.

There's nothing here to suggest Michigan should be really bad. Not yet.

Moving on.
 

The Other Shoe, of Which Its Current Gravity Situation You Were Well Aware

312059

Rose : Don't you understand? The water is freezing and there aren't enough boats. Not enough by half. Half the people on this ship are going to die.

Cal: Not the better half.

Cal, if you make it off that ship, and if that whole heir-to-a-robber-baron thing doesn't work out for you, you might make a fine SEC recruiting coordinator.

What I'm talking about is Alabama's over-signing strategy, which has been covered many times on this blog. In short, the Crimson Tide under Saban have recruited more guys than they have scholarships for, expecting enough will find reason to get themselves expelled or booted off the team before the count becomes official. The ultimate effect is that Saban has a strong incentive not to keep troubled players, particularly less talented troubled players, in school.

I bring it up now because:
 

  Def. Recruits On Roster On Roster %
Michigan 48 28 58.33%
Alabama 83 52 62.65%
Michigan State 64 45 70.31%
Penn State 52 40 76.92%
Ohio State 59 46 77.97%
Notre Dame 43 36 83.72%

Bullets?

Bullets.

  • Michigan has had higher attrition from 2005 to 2009 than Alabama.
  • Let's rephrase: Michigan has had higher attrition than a team that has been TRYING TO SHED PLAYERS.
  • If Ohio State is pulling a 'Bama, there is zero evidence for it here. They have a reasonable number of recruits, and very low attrition.
  • Penn State, as I mentioned before, is a much older team, and therefore has had a lot more time to lose guys to graduation and leaving early for the NFL and whatnot. In that light, their retention rate is pretty darn good.
  • Michigan State and OSU ended up with about the same number of recruits on their respective rosters, while Bama was just a bit higher.
  • Notre Dame's team is much, much younger, hence the high retention rate.
  • Attrition has generally been higher for the teams with coaching changes in the last few years.
  • Michigan's 28 scholarship athletes on defense may work on your pre-2005 EA Sports video game (which had a 55-player limit) but is way, way below the competition.
  • 58.33 percent, as it turns out, is in fact quite putrid.

Attrition is going to normally be between 60 and 80 percent, for any team.

The 2005 class isn't expected to leave very much for 2009 -- those are the current 5th year seniors. Guys play early because of their great talent, or because of [Position] Armageddon, or because you're Michigan State and know that 4-star prospects come along so rarely that when you get one you can't wait to rip open the cover and see if you won a Golden Ticket!

For the bigger programs, NFL Early Entry takes its toll from the junior (2006) ranks. Sophomore classes are generally mostly transfers. Freshman attrition has a lot to do with the tough transition from The Hottest Shot in Home Town High to collegiate freshman (class breakdowns will be in another section).

Let's break this down, too, by talent. Because it's not just numbers, right? If you have 28 guys on the roster, but they're all highly recruited athletes who have shown they belong on an NCAA roster, you're still afloat.
 

  4*+ Recruits 4*+ Retained 4*+ Ret %
Michigan 26 16 61.54%
Alabama 41 30 73.17%
Michigan State 10 6 60.00%
Penn State 25 18 72.00%
Ohio State 35 27 77.14%
Notre Dame 20 20 100.00%

And so.

Among primo athletes, Michigan was slightly above Michigan State, which is our small sample outlier (MSU lost three 4-star player to injury or transfer from its 2006 class, and one 2005 guy graduated). Penn State, at a respectable 72 percent, was a victim of graduation and the NFL. Notre Dame still has all 20 of its blue and light-blue chippers, thanks to packing them into later classes. Alabama is about where Penn State is, but remember, the Tide packed their high-level guys in from 2007 on.

Still, for our purposes, the important thing to note is that we still have 16 4-star defensive athletes on the roster, which is just two less than Penn State (but 11, i.e. a whole 'nother defensive unit, behind Ohio State).

While we're at it, let's see the 3-stars:
 

  3*- Recruits 3*+ Retained 3*+ Ret %
Michigan 22 12 54.55%
Alabama 42 22 52.38%
Michigan State 54 39 72.22%
Penn State 27 22 81.48%
Ohio State 24 19 79.17%
Notre Dame 23 16 69.57%

Now we see the difference.

Retention rates are down across the board here, as this group is more likely to be hit harder by the transfer-due-to-playing-time bug.

Michigan State, you can see, built their city on 3-stars, with retention rates similar to that of 4-stars for the big schools.

Bama dropped almost half of their 3-stars (evidence of the Saban Stratagem working harder on the less-pedigreed?), ending with the same amount of 3-star guys as Penn State, but of course from a march larger pool.

Of course, our 3-star retention rate sucked too, except ours came out of the more meager Penn State/Ohio State/Notre Dame budget. All told, we lost 10 guys: Quintin Patilla, Johnny Sears, Brandon Logan, Quintin Woods, Artis Chambers, Chris Richards, Carson Butler, Chris McLaurin, Adrian Witty (who may make it onto the 2010 haul) and Marrell Evans. Most, but not all, of those guys transferred when they saw playing time disappearing. All but Witty (who may end up not counting) were pre-Rich Rod guys.
 

But this ship can't sink!

She's made of iron, sir! I assure you, she can... and she will. It is a mathematical certainty.

0000-_D1299500646

To help put Michigan's attrition in perspective, let's look at it by class:
 

  AVERAGE Michigan Alabama MSU Notre Dame Ohio State Penn State
2005 33.33% 0.00% 36.84% 30.77% 37.50% 54.55% 33.33%
2006 67.95% 63.64% 53.85% 61.11% 90.00% 76.92% 69.23%
2007 77.42% 66.67% 60.00% 90.00% 85.71% 72.73% 100.00%
2008 87.50% 62.50% 85.71% 90.91% 100.00% 81.82% 100.00%
2009 88.24% 90.91% 73.33% 91.67% 100.00% 100.00% 81.82%

Here's what the attrition looks like next to two of our three annual rivals:

4068820012_2ac03352c1_o

I postulated in the opening diary that not having any old guys around was hurting us. It seems our competition did much better. Ohio State still has over half of its 2005 on the roster as 5th year seniors. We have none. Other than us outliers, everyone else has about a third of their oldest class around.

Of the seniors, we are blessed with 63 percent, which looks good against Bama's mass exodus (under 54 percent), but is sub-par nonetheless. Notre Dame still has 90 percent of its 2006 class.

The one place where Michigan is up against the average is in true freshmen, the one spot it hurts the least to lose guys. Our 2008 class stands out for premature desolation -- that is something that could likely hurt us down the road.
 

You can be blasé about some things, Rose, but not about Titanic

610x

So before we start ripping into Michigan recruiting and retention, and assigning blame (yes, I will do this before we're done), let's get an important thing out of the way:

There is still talent here, and will still be talent here.

Brandon Graham and Donovan Warren are about as great of players as you could ever ask for, and Mike Martin and Craig Roh are on that same path. Ryan Van Bergen has been everything expected. Steve Brown has come a long way to be effective at a position nobody may have ever imagined for him. 4-star recruits Jonas Mouton and Mike Williams have shown themselves to be prone to mistakes, but are effective players (usually) when protected from having to play defensive quarterback.

The point is, most teams, as we shall see, don't get Grahams and Warrens and Martins and Rohs and Van Bergens. They have to pray their precious few blue chips end up more like Warren and less like Jonas Mouton, while waiting for their legacy 3-stars (Woolfolk) and athlete fliers (Ezeh) and a cascade of JT Floyds to get competent.

Michigan isn't "most teams." We can recruit with the big dogs. When times are good, we can reasonably expect most of the defensive positions to be filled with upperclassman 4-star talent or better, with young 4-stars and maybe a few diamonds in the rough pushing them for playing time.

However, as we shall see, you cannot reasonably expect your 4-stars to all pan out. You need backup plans. You need the dudes.
 

So...What Happened?

A horrible confluence of events:

  1. Michigan had generally small recruiting classes
  2. Those defensive recruiting classes were about on par with Penn State anyway, and well below those of Ohio State
  3. Michigan had exceptionally high attrition from its 2005 to 2008 classes.
  4. Attrition disproportionately attacked our higher-rated players.
  5. Attrition disproportionately attacked our older players
  6. Of Michigan's 3-star recruits, a disproportionately few ended up as contributors

This all resulted in a generally young defense with a good but well-short-of-the-competition retention of high-end talent, and pretty much zero in the way of backup talent.
4068065611_2485d2a0ef_o 

When the attrition is cut out, we are visibly deficient next to our competition. Of the guys we've kept on, every one that is a true freshman, or didn't work out, cuts further into our ability to field a team. In fact, let's run that same thing again with just upperclassmen (classes 2005 to 2007):
 

  MSU Ohio State Alabama Penn State Notre Dame Michigan
6.1 0 1 0 1 1 2
6.0 0 0 2 1 2 1
5.9 1 3 3 6 3 2
5.8 2 9 5 4 6 2
5.7 2 2 3 2 4 2
5.6 3 1 3 2 2 2
5.5 5 5 4 4 0 2
5.4 3 0 0 1 0 0
5.3 0 1 1 0 0 0
5.2 5 1 0 0 0 0
5.1 2 0 0 0 0 0
5.0 1 1 2 1 0 0
Total 24 24 23 22 18 13

4068164727_2f79d2a7db_o
And there you have it: the guys that Michigan has left are good, but there are only 13 recruited upperclassmen left on the roster, spread out evenly between middling 3-star and blue chip. Note that every other comparative school has a great big chunk between 5.7 and 6.0 -- that's high 3-star to just-below-a-blue chip. Everyone but Notre Dame has another hump of 3-star guys as backups. The next closest rival is Notre Dame, which is considered an incredibly young team. Everyone else has over 22, i.e., enough scholarship athletes left on the roster in the junior years or above to fill the two-deep.

The one thing we have on everyone else: two 5-stars. That's Graham and Warren. Indeed, we spared no expense in a luxurious 1st class cabin, but our ship is visibly lacking in the hull department.

Who's to Blame?

4068218831_57afbecaa0_o 
If you want a short way of saying it, we had Alabama-level attrition from Penn State-sized classes. We lost the bulk of our upper classmen to injuries and Clearing House snafus and sad lack of talent, and couldn't recruit or hold onto anybody for depth. We missed opportunities to find diamonds in the rough while looking for new coaches. We didn't recruit a safety for a decade.

No one solitary factor caused this. A small recruiting class is fine if you still get highly rated players who stick around and pan out (Notre Dame, Penn State), or if you have a lot of 3-stars behind them from which you can draw le creme (Michigan State, Alabama).

And (except for musing along the way) I failed to even touch on several other very important factors, such as the fact that Michigan had four different defensive coordinators over this period, or that it seems an inordinate of our recruits ended up overrated.

But that doesn't mean everyone's absolved. There were a number of points along the way when, without hindsight, a better job probably should have been done:

On Lloyd, and his staff:

  • Someone should have recognized the weak depth at safety back when Marlin Jackson had to move there for a season. Lloyd let the positition go unaddressed in recruiting for years, which killed us in 2005 (burning Brandon Harrison's redshirt in the process).
  • And then he came back with Jonas Mouton and Stevie Brown, two outside linebackers, meaning in the recruiting year immediatly after "Safety Armageddon" we got ZERO defensive backs.
  • Ron English had (still has) a strange belief that Johnny Sears was a great future cornerback, even after The Horror.

On Rich Rodriguez, and his staff:

  • 2008's defensive recruiting was a page right out of Notre Dame: a spattering of 4-stars, but no "diamonds in the rough" discovered, no guys brought in for a certain scheme. Two never made it to campus, and a third has been kicked off the team. I have to believe there were enough 3-star speedy defensive backs who would leap at a Michigan offer if we put in the research. The coaching change, and then RR's focus on the (even more depleted) offense didn't allow that to happen.
  • Shafer. We spent a year with a defensive coordinator who was dismissed at the end of it, meaning we were right back where we started, minus a (rather paltry) graduating class, at the start of 2009

Is There Any Hope?

Michigan's Obi Ezeh, top, and J.T. Floyd, right, team up to bring down  Michigan State University wide receiver Mark Dell after a short gain during first quarter action of Saturday afternoon, October 3rd's clash between the in-state rivals at Spartan Stadium in East Lansing. 
Lon Horwedel | AnnArbor.com

Not for this year, no.

And not for next year either.

That is to say that there's no hope in this study for Michigan's current defense to suddenly turn into a Top 15 defense that Michigan has traditionally thrived on. There isn't even an endless cupboard of 3-star guys from whom one or two could emerge to fill holes (a la MSU). We're basically down to walk-ons who can tackle, and praying nobody runs by them at this point.

Basically what this exercise shows is that we got triple-whammied by relying on too few highly rated guys not to bust, relying on a normal number of recruits to stick around to graduation, and having some upperclassmen around who could surprise late in their careers as contributors. You can't plug iceburg-sized holes and fix years of mistakes and bad luck with one slick move.

4069119184_274566da31_o
Based on what's left on the roster, in 2011 we may end up being better (but not deeper) than Penn State and Michigan State, but not by much, and almost certainly not better than Ohio State, Notre Dame, or Alabama. However, if our attrition rates come down from their ludicrous highs (which can be expected), and we get ourselves some nice classes, there is great hope for 2012. Also, I expect Notre Dame's silly-high retention rate to drop once people give up on Weiss, or players next year decide they don't fit into Tenuta's schemes, though I can't deny they have the basis for a pretty solid team.

For the rest of this year, I recommend, no matter who the opponent may be, the strategy we take into every game starts with a burned ship and ends with Tate Forcier throwing jump balls deep down-field. The offense has shown it is quite capable of awesome. No more Lloyd ball -- not in 2009, not in 2010.

We are a team that can get torched by Illinois and Indiana and Eastern Michigan, thus our strategy must be to always torch more.

As for the future, well, another class full of defensive backs and linebackers can't hurt. This kind of bad is the kind that seeps deep into next year, but at least in 2010 we should have what's left of the 2008 class and some of the 2009 class ready to contribute, and also that rare second year under the same coordinator. Roh and Martin and Van Bergen are a defensive line we can bank on. The defense will lag behind the offense by a year, but the situation isn't actually as desperate as it was there. On the other hand, there's no Rich Rodriguez/Calvin Magee schematic/coaching advantage for the defense to coax greater early returns, or promise future greatness.

In the long term, defense is solved by lots of recruiting. I suggest going for lots and lots of fast guys, using the carrot of early playing time.

Screw the look of the deck. Screw the extra 3rd class cabins. I want bulkheads that go all the way up, and close up tight, and reinforced steel second hulls, and lifeboats, lots of lifeboats.

We need the dudes.

animalhouse19

As Delta Tau Chi men Lawrence Kroger, John Blutarsky and Kent Dorfman illustrate above, there are two ways of making sure your House always has enough good brothers to stay strong:

  1. Bring in lots of fellas. "Pinto" and "Flounder" may look like pretty useless guys, but you never know: one of them might have access to a car that, when modified and matured, will save your whole fraternity.
  2. Keep 'em there. "Bluto" here is a 7th year senior. While the NCAA frowns upon 7th years for guys who aren't Mormon quarterbacks, you can get away with being choosy in your class selection if you can stay true to your mission and keep "Otter" from chasing tail outside your doors, or "Boon" from listening to his girlfriend, or "D-Day" from riding off into the sunset. The young guys are the future of your program but it's the oldest guys who make the house.

In Michigan terms, this means if you find yourself with only a few quality guys and not much else, the entropy of the recruiting-to-player process will kill you. You don't just need enough 4-stars to fill the roster spots, but enough also to cover injuries to good players, and the inevitability that some guys won't pan out, or will get injured, or will transfer.

I said there would be hope. So here it is: Rich Rod seems to be way ahead of me on this. We are in on a lot of defensive back prospects, including 4-stars aplenty. Not only this, but starting last summer, RR packed a cannon with scholarship offers and has been shooting them into every high school defensive unit in America. The result has been an odd recruiting year, with guys jumping onto the board with high interest at crazy times through the year. If it's dudes we need, it seems RR and GERG are already on it.
 

Advice for Further Study

I posted a copy of the Excel spreadsheet above. I would love it if someone would add more teams to the study, or qualify the recruits by creating a new category for later-career ranking. In that, I mean find some way to reassess each player based on his performance thus far against what we should expect from a player of any given Rivals Rating. I'd like to see how Michigan stacked up in picking up guys who would come above versus below expectations.
titanic_photo-761580
fin.

Comments

WanderingWolve

November 3rd, 2009 at 11:31 AM ^

Thanks Misopogon. I'm really enjoying your posts. It's a helpful reminder that there is a legitimate reason behind the madness. Go Blue, beat Purdue! If we win Saturday (please God, let it be!) I think it'll help the boys loosen up and could make for better games vs. Wisky & tu0$.

MechEng97

November 3rd, 2009 at 12:03 PM ^

This is exactly what I was wondering about, but unfortunately have not had the time to research. Thank you for such a great post. As Rodriguez said many times on Monday, I'll take a 4th-5th year man over a 1st-2nd year boy almost every time. Rankings matter based on percentages and so does retention rate. Although this depresses me somewhat for the 2010 outlook at least I understand more why we are playing walk-ons and teams like MSU/PSU have better D's.
Hopefully someone takes your advice and quantifies the performance of recruits we have versus others. Another bit of information that I'd be curious on. Hopefully we get the #'s this year and next to start turning this around.
I agree that we need to out score our opponents for the next year or two. It doesn't help that our QBs are so young to accomplish this, but it's the right strategy. If we could have scored that TD in the 3rd Quarter and answered one of IL scoring drives we'd be a lot happier now. It just snowballed..

markusr2007

November 3rd, 2009 at 12:45 PM ^

I tried to cover some of the attrition issues under the Nehlen-to-Rodriguez change on my blog - Amateur work compared to the masterpiece(s) you've posted here.

What struck me most was the Alabama business.

Presumably the Saban regime culls the herd of recruits that also includes several players who are talented and don't have massive personality and academic issues.

I'm wondering where those kids end up?

DoubleMs

November 3rd, 2009 at 1:36 PM ^

I crunched a few numbers, made a magic formula using the season stat lines (I'll put a full post up with more info later), and came up with this for defensive performance 2005-2008 by player:

Name Year PTS PPG PPG/RR*10
Brandon Graham 2008 157.25 14.29545 23.43516393
Steve Brown 2008 151.5 12.625 21.39830508
Obi Ezeh 2008 102.25 8.52083 15.49241818
Brandon Graham 2007 103.5 7.96153 13.05168852
Brandon Harrison 2005 88.25 7.35417 12.67960345
Brandon Harrison 2008 77 6.4375 11.09913793
Terrance Taylor 2007 75.75 5.82692 9.876135593
Donovan Warren 2008 64 5.81818 9.538
Obi Ezeh 2007 68.25 5.25 9.545454545
Jonas Mouton 2008 57.5 4.79167 7.986116667
Brandon Harrison 2007 59.25 4.5769 7.891206897
Adam Patterson 2008 4.5 4.5 7.627118644
Steve Brown 2007 56.75 4.36538 7.398949153
Donovan Warren 2007 51.25 3.94231 6.462803279
Terrance Taylor 2008 38 3.1875 5.402542373
Mike Martin 2008 35.25 2.9375 5.064655172
Terrance Taylor 2006 35.25 2.71154 4.595830508
Michael Williams 2008 28 2.54545 4.388706897
Brandon Harrison 2006 25.75 1.98077 3.41512069
Austin Panter 2008 7.25 1.20833 2.083327586
Brandon Graham 2006 9.5 0.95 1.557377049
Austin Panter 2007 6.5 0.92857 1.600982759
Boubacar Cissoko 2008 9.5 0.79167 1.31945
Adam Patterson 2007 9.5 0.79167 1.341813559
Marell Evans 2008 4.75 0.67857 1.304942308
Johnny Sears 2006 6.25 0.56818 1.014607143
Greg Banks 2007 3.5 0.5 0.892857143
Steve Brown 2006 5.75 0.44231 0.749677966
Ryan Van Bergen 2008 5.25 0.4375 0.754310345
Troy Woolfolk 2008 4 0.4 0.727272727
Carson Butler 2008 3.5 0.35 0.636363636
John Ferrara 2007 3.75 0.28846 0.515107143
Troy Woolfolk 2007 2 0.28571 0.519472727
Brandon Herron 2008 2.75 0.25 0.438596491
J.B. Fitzgerald 2008 2.75 0.22917 0.388423729
Chris Richards 2006 2 0.2 0.363636364
Greg Banks 2006 2.75 0.19643 0.350767857
Jonas Mouton 2007 1.75 0.19444 0.324066667
Artis Chambers 2008 1.5 0.16667 0.297625
Chris McLaurin 2007 2 0.15385 0.279727273
Renaldo Sagesse 2008 1 0.125 0.219298246
Eugene Germany 2006 1 0.125 0.208333333
Adam Patterson 2006 0.75 0.125 0.211864407
Carson Butler 2006 1.25 0.10417 0.1894
Kevin Leach 2008 0.5 0.1 0.204081633
Kenny Demens 2008 0.25 0.08333 0.143672414
Johnny Sears 2007 0.25 0.0625 0.111607143
Jason Kates 2007 0.25 0.0625 0.107758621
Artis Chambers 2007 3 0.06 0.107142857
Terrance Taylor 2005 0.5 0.05556 0.094169492
Chris McLaurin 2006 0.5 0.05556 0.101018182

Bullets (without bullets):

Brandon Graham is a Beast.
Brandon Harrison's best year was his Freshman year (!?).
Stevie Brown really wasn't that bad last year, statistically... I diminished the importance of tackles and increased the importance of TFL's and Sacks and he still came out with a great season, implying that it isn't just because everything got through to him.
Brandon Graham is responsible for 2 of the top 5 years represented.
Terrance Taylor peaked in his Junior year. Part of this was probably Shafer's system.
Ezeh was really quite dominant last year... where has the Ezeh of last year gone?

msoccer10

November 3rd, 2009 at 2:49 PM ^

It is great to see the analysis of our defense. A lot of what you showed was discussed by Brian and others but with this research it is undeniable that our talent level on defense isn't good.

I think you pointing to 2012 as the time to expect an elite defense is right on, but I fully expect our defense to perform better next year than this year for a three reasons (by that I mean less points allowed).

First, we are only losing two (three) starters in Brandon Graham and Stevie Brown (and maybe/probably Warren). Lets look at it by position group.

I don't think anyone can replace Brandon Graham, but just like Mike Martin is doing an adequate job filling in for Terrance Taylor, I think Roh will do the same for Graham with another year of strength and condition and coaching under his belt.

So, our defensive line will be young, but three of the four will have a year starting under their belt. We have seen that Sageese is adequate and have reason to hope Campbell can play at a reasonable level by his second year if for nothing else other than his recruiting hype. Herron also has game experience and we have backups in Patterson, Banks, and Lolata.

Linebackers are an issue, but that is actually an area where we have decent retention and a lot of lottery tickets. Ezeh, Mouton, Leach, Kovacs(?), Jones, Smith, Herron, Fitzgerald, Demens, Hawthorne, Bell, not to mention Marvin Robinson, and any other true freshman. I have got to believe that after a year Robinson can find/coach three of these guys to competency.

Defensive back play will live or die with Warren. If he leaves we are praying that Turner can become a stud his redshirt freshman year and that both he and Woolfolk remain healthy. Woolfolk is decent and will get better I believe. If Turner can come out and play better than Cissoko, we are no worse off than when we started this year and if Warren stays we are better.

Safety is a serious problem and I don't see a bright side other than the fact that we don't lose anyone(not hard when you only have one scholarship player available, who is a sophomore). Still, since this is Williams' and Kovacs' first year starting, you hope Robinson can get improved play from them and develop Emelien, Gordon and maybe Floyd or Turner depending on what Warren does. It won't be pretty, but I think it will be better than this year and can't see it getting worse, barring injury.

Second, our offense loses only four starters. We will get Molk back and Tate/Denard will have another year to improve. Another year to learn Rodriguez's offense and put in new plays for the youngsters and another year for o-line maturation and competition. I expect our offense to be a lot better next year and to give our defense a much better chance at success.

Third, Robinson will be here again.

With improvement from both sides of the ball, we can realistically challenge for the big ten next year imhe. I am not saying we'll be the favorite, but I am thinking 2 losses in big ten play is a reasonable goal.

DHerrick

November 3rd, 2009 at 8:03 PM ^

One of the best things I have read in a while. Outstanding!

For as long as I have been following recruiting rankings online (7-8 years), I remember discounting Michigan's ranking and focusing instead on "average stars." My theory was that the rankings overrated numbers and underrated quality. Your work makes me think that Rivals et al. had it right all along -- numbers matter too ("we need the dudes"). Anyway, from memory, our numbers always seemed relatively low, though I was not focusing on defense specifically.

Big10Defender

November 4th, 2009 at 9:11 AM ^

...are better than any OSU blogger out there. Kind of amazing that as a Buckeye, the best recruiting breakdown I've seen on OSU comes from a UM blog's Commenter/Diarist.

Also, reason #3,287 why newspapers are dying - with all the MSM out there PAID to write about sports for a living, have you seen anything so well researched and clearly laid out as this?

MIdocHI

November 8th, 2009 at 3:51 PM ^

The statistics are good, but your whole premise is that it is only the players, their quality and depth, that matter. Coaching matters. RR and GERG are not doing it.

Ron Zook, great recruiter, can't coach. Urban Meyer wins national championship with Zook's players.
Stoops took over from Blake(a nobody) wins national championship in second year at Okla.
The players were obviously there based on your premise, but the previous coaches could not get it done.

Just because RR had good teams before does not mean he will have good teams again. Spurrier has been at So Carolina for 5 years, so all players are his recruits. His overall record is 2 wins for every loss which average to an 8-4 season. Nothing special.

There is more to it than just players. These coaches are not getting it done. You can't just say the players on defense aren't there. Otherwise, why spend money on a coach? What you are saying is that once you have statistically enough depth of 2, 3 and 4 star players, then the defense will succeed. It has already been proven that having talented players is not enough, or Ron Zook would have won a national championship. You need a good coach also, and it is not clear that we have that.

Seth

November 9th, 2009 at 10:13 AM ^

In Anthropology and Paleontology, a lot of times you'll get the best complaints about your papers from the strangest sources. They'll point out something you completely took as a given that you totally shouldn't have and you'll be flabbergasted to have your entire premise shaken so easily. Then they'll go into a long counter-assertion that argues for the morphological changes in question to be the work of a flying spaghetti monster or some such over an impossibly short period (like 5,000 years or something crazy like that). And because you know exactly where the Flying Spaghetti Monster comes from, you apply your carpals to the glabella of your frontal bone..

...and forget that there was actually a flaw they pointed out.

Before launching into the sports commentary equivalent of this:

, you said this:

The statistics are good, but your whole premise is that it is only the players, their quality and depth, that matter. Coaching matters.

Absolutely, coaching matters. The thing is, it's not the only thing that matters, and that's what I think this project was able to show.

I don't think you believe coaching is the only thing, either, even though you don't say it in your post. I think that can be a given, just as it's a given that you can't fault a defense that is always put in horrible field position for giving up the same amount of points a defense that gets great field position gives up.

There is nothing in history or the present that suggests to me that Rich Rodriguez is a good defensive coach. He seems to think that too, and gives control of the D to his DC, so really if there's any coaching effect it's GERG's that we have to assess.

Thing is, that's really hard to do, especially right now. GERG has been here one year. That one year has been an utter disaster on defense -- the worst defense I've ever seen at Michigan. But to pin that on GERG alone is a bit of a Post Hoc, Ergo Propter Hoc argument (GERG arrived, we sucked, ergo GERG made us suck). You have to show why.

And so do I.

The closest I think I came to that in this study was this:

Visually, the teams that seem to be doing very well right now are the ones with lots of high-level recruits in the junior, senior, and 5th year senior ranks.

The other thing I did was the chart at the very beginning of the series, that had a lot of s all over the place on the two-deep beside starters with descriptions like "Current Infinite Safety Disaster, who is worse than the walk-on."

The question I did not prove, and you pointed out, was "Will a good coach win with that type of talent deficiency?"

I point you again to Figure 5. Note how closely these defenses correlate with that chart. I think that's fairly good evidence -- though certainly not iron -- that it's the talent of the players, more than the abilities of any coach, that is most important in dictating the ultimate effectiveness of a defense.

I personally think that with defensive coaching in college, player development and on-field adjustment are infinitely more important than scheme. This is outside the scope of this study, but there are things that I think we know about various defensive coaches, like that Mark Dantonio is fantastic at developing linebackers, or that cornerbacks who spent time on Lloyd Carr's field can damn well tackle somebody.

As far as assessing Rich Rod, the most I think we can do is look at how he picks his defensive coordinators. He already fucked this decision up once. And in this he also seems to be at a disadvantage from most other coaches because everything he knows from offense is that scheme is very important, and this seems to have translated into his going for schematically technical defensive coordinators here and at WVU. It's too early to say for sure, but my natural inclination is so far that he sucks at this.

Back to point (did I warn you of Misopogonal logorrhoea?), we haven't had GERG long enough to see how he develops players. We haven't seen how well his scheme does with any group except this suicidal kitten-laden one. We know he has recruited one class and it was this one:

Name Class Pos Stars RR What happened? Here?
William Campbell 2009 DT ***** 6.1 DT (backup) yes
Justin Turner 2009 CB **** 6.0 redshirt yes
Craig Roh 2009 DE **** 5.9 WLB yes
Anthony LaLota 2009 DE **** 5.8 redshirt yes
Vladimir Emilien 2009 S **** 5.8 FS (backup) yes
Brandin Hawthorne 2009 LB *** 5.7 MLB (backup) yes
Isaiah Bell 2009 LB *** 5.7 SAM (backup) yes
Teric Jones 2009 RB *** 5.7 CB (backup) yes
Mike Jones 2009 S *** 5.7 MLB (backup) yes
Thomas Gordon 2009 ATH *** 5.5 redshirt yes
Adrian Witty 2009 CB ** 5.3 DNQ (may return)
no

Which looks like a damn good class.

If you can prove to me that (most) every coach who will ever be great has come in and won with incredibly inferior talent, I will recant. Otherwise, I'm telling you that we don't have enough information yet to make an accurate assessment of GERG as a defensive coordinator. All we know is that our team sucks, and that our depth chart is probably the biggest culprit, and that the 2005 to 2008 recruiting classes and their massive attrition is probably to blame for that.

fin

P.S. Make sure you don't miss the start of the second season of Dollhouse on Fox!!!

--Summer Glau

Believe in RR

November 12th, 2009 at 1:42 PM ^

Excellent Post!

I've been a viewer of the blog for a long time, but felt with all the negative comments on RR, I needed to register with the site and give me two cents.

Question for you and I apologize if someone has already asked this, but of those that left the team, how many were before RR was even hired? I think this would be an interesting stat.

Again thanks for putting in the time to get us this info!

Go blue!

Seth

November 12th, 2009 at 2:48 PM ^

Left under Carr (includes burned redshirt graduations):

Name Class Pos Stars RR
Eugene Germany 2005 DE 4 6.0
James McKinney 2005 DT 4 5.9
Terrance Taylor 2005 DT 4 5.9
Brandon Harrison 2005 CB 4 5.8
Cobrani Mixon 2006 LB 4 5.8
Austin Panter 2007 LB 4 5.8
Quintin Patilla 2006 LB 3 5.7
Johnny Sears 2005 CB 3 5.6
Brandon Logan 2005 LB 3 5.6
Quintin Woods 2006 DE 3 5.6
Chris Richards 2005 ATH 3 5.5
Chris McLaurin 2005 DE 3 5.5

Left under RR (includes early NFL):

Name Class Pos Stars RR
Boubacar Cissoko 2008 CB 4 6.0
Jason Kates 2006 DT 4 5.8
Marcus Witherspoon 2008 LB 4 5.8
Taylor Hill 2008 LB 4 5.8
Artis Chambers 2007 S 3 5.6
Carson Butler 2005 DE 3 5.5
Adrian Witty 2009 CB 2 5.3
Marell Evans 2007 LB 2 5.2

I guess Witty shouldn't be held against RR, because he will probably join the 2010 class. Of those whose redshirts were burned under Carr, only Harrison sticks out. His shirt was burned during 2005's Safety Armageddon. He was a 5'9 DB who wasn't fast, but how badly would you take a 5'9 DB who isn't fast but is a 5th year senior with years of experience right now? I don't know if Brandon Logan could have helped -- he never moved up the depth chart.

Keep in mind, Carr brought in the 2005 through 2008 classes. 2009 was the first that RR did some real defensive recruiting. That said, Cissoko, Kates, Witherspoon, Hill, Chambers, Butler, and Marrell Evans are all guys we lost on his watch.

Considering the stark difference in tenures charted here, I think it's pretty clear that attrition has been much higher since RR's arrival, to which I reply "biggest program culture shift since 1969."

To shorten the blame game, Carr is mostly responsible for not getting the dudes here. His late-career recruiting left gaping holes at linebacker and defensive back. RR can be blamed for losing some of what small depth was there.

The 2008 recruiting class seems headed toward a repeat of the 2005 class, which sent disaster waves through the program that we are experiencing still. My next project has been working toward a definitive "when will we be back?" question, and the early returns are not good; barring an influx of 5-star talent very soon, by 2012 we will still not be back to the talent/seasoning level of the 2007 defense, which wasn't even that great of a defense.

Sneak peak:

Year Defensive Rating Pred. Wins (+/-3)
2007 3.77 8.42
2008 3.70 6.01
2009 3.42 6.72
2010 3.48 6.90
2011 3.76 8.11
2012 3.71 8.01