The Decimated Defense, Part III: the Resuscitating

Submitted by Seth on

The Rodfather Part III

Why was I so feared, and you so loved? What was it? I was no less honorable. I wanted to do good. What betrayed me? My mind? My heart? Why do I condemn myself so? I swear, on the lives of my children: Give me a chance to redeem myself, and I will sin, no more.

If you have followed this series, you are now familiar with the message: light but focused recruiting plus really bad attrition equaled a defense with some really great players and some suicidal kittens.

Here's a question: based on recruiting and retention from 2006 to the present, how good will Michigan's defense be this year?

This diary will look at the talent on Michigan's roster in 2010 versus that of 2009, plus that of two rivals in that time, to try to get an idea of what kind of team we will be looking at this coming season.

In the first, we met the family.

Grahambro - image created by Misopogon

In the second, we saw that family destroyed.

correllandkiss - image created by Misopogon

Today, I bring you Part III.

GERGfatherIII - image created by Misopogon

You know, the one that's chronologically incompatible with the other two, with the substantially lower production values, that rehashes old characters and plotlines from the previous flicks in order to squeeze more cachet out of the franchise.

In the third (and final?) installment of the Decimated Defense Trilogy, I will look to the future, comparing Michigan's 2010 defensive roster and attrition numbers against those of last year, and also against 2009 and 2010 for two relevant rivals: Ohio State and Michigan State.

Still interested? [ED: YES YOU ARE.]

Justify Your Existence

Here's an adage: "Defense wins championships."

Here's proof of that adage, thanks to MCalibur:

http://mgoblog.com/sites/mgoblog.com/files/SickChart_0.png

Things go into goodness of defenses. Things like coaching. Like scheme. Like experience. Like training. I cannot qualify these.

It's also about depth, because the weakness of your weakest link matters more than the awesomeness of your best link.

And it's about talent (whether athleticism or intellect). And that is what I shall attempt to quantify and qualify with regard to next season's defense in Ann Arbor.

Again, I used Rivals Ratings as the metric for measuring talent. It's not exact, but a highly ranked recruit is more likely to become a better player and vice versa, and over a large enough sample, e.g. several recruiting classes, we know that recruiting rankings do matter.

Recruiting rankings and hype, of course, go hand-in-hand. This is because recruit rating is a metric of potential, not performance. When a program signs more recruits of higher rating, what the program is getting is a better pool from which it must ultimately draw its level of performance.

Practical application: Michigan State didn't have a better defense than Michigan last year because Greg Jones is better than Brandon Graham, but because their worst player was the best that could be drawn from a large group of 3-stars, while our worst player was the best that could be drawn from L.S.A.

As before, you are welcome to have at my Excel file.

Great News Everybody: We Have … Dudes!

http://www.annarbor.com/assets_c/2010/04/Spring-Game-041710-thumb-590x397-36636.jpg

Attrition didn't end entirely. Here's what we lost from the 2006 (5th year seniors this year) through 2010 classes:

Name Class Pos Stars RR What happened?
Jason Kates 2006 DT **** 5.8 Left team
Cobrani Mixon 2006 LB **** 5.8 Left team
Quintin Patilla 2006 LB *** 5.7 Left team
Quintin Woods 2006 DE *** 5.6 Left team
Brandon Graham 2006 DE ***** 6.1 Graduated
Steve Brown 2006 LB **** 5.9 Graduated
Austin Panter 2007 LB **** 5.8 Graduated
Artis Chambers 2007 S *** 5.6 Left team
Marell Evans 2007 LB ** 5.2 Left team
Donovan Warren 2007 CB ***** 6.1 Early NFL
Boubacar Cissoko 2008 CB **** 6.0 Left team
Marcus Witherspoon 2008 LB **** 5.8 Did not qualify
Taylor Hill 2008 LB **** 5.8 Left team
Brandon Smith 2008 S/LB **** 5.9 Left team
Adrian Witty 2009 CB ** 5.3 Did not qualify (may return)

The bold/italics are the guys lost since last November.

Speaking just for next year, all four were the worst kind of attrition: two thirds of the team's blue chips, and two more high-four stars. Where blame lies for them, well, Graham you can't blame on anyone since he was an important backup for the 2006 team as a freshman and became a mid-1st round NFL draft pick. Steve Brown could have benefited from a redshirt (and also four years of Greg Robinson).

Smith and Warren are on Rich Rod. From Warren's comments, through family, it seems he still kind of felt like he was part of a Michigan that wasn't there anymore, which, for better or worse, is Rich Rodriguez's fault. If ultimately his regime shift creates a leaner, stronger Michigan, it will be worth alienating some of the better legacy players. Smith was a Lloyd recruit who hung on until he didn't know why he was hanging on anymore.

That is the price you pay for the life you choose.

Still:

Remember how last year we only had 28 recruited players on the defensive roster? Now we are up to 43.

  Def. Recruits On Roster On Roster %
Mich '09 48 28 58.33%
Mich '10 54 43 79.63%
MSU '09 65 46 70.77%
MSU '10 61 43 70.49%
OSU '09 60 47 78.33%
OSU '10 57 46 80.70%

This is just a handful fewer recruited defensive players on the roster than Ohio State and Michigan State had last season. It's the same amount that Dantonio will take into battle (bad choice of words?) this fall.

The notorious 58.33-percent attrition from last year has evaporated; we are now in about the same range as Ohio State for pure retention, leapfrogging Michigan State in the process.

The biggest difference – and this is also good news – comes from the 5th Year Seniors. The 2005 class left zero (0, zilch, nada, personne) players for the 2009 team. Despite the departures of Brown and Graham, the 2010 Michigan Defense will still have four returning 5th Year men. These are:

Name Stars Pos RR
Jonas Mouton **** LB 6.0
Adam Patterson **** DL 5.9
Greg Banks *** DL 5.6
Obi Ezeh *** LB 5.5

None of them are stars. Banks and Patterson will back up the defensive line, while Obi and Mouton are being pushed by others (one a walk-on) for their starting roles. But they're 5th year seniors, playing under the same defensive coordinator for a second year in a row, both luxuries we did not have last year.

Warm body count is simply depth. This year, there will be, on average, just under four scholarship players for each defensive position. If three out of every four of them was a 4-star recruit or better (and half of them were upperclassmen), this would be about as good as good as 2009 Ohio State.

So, are they that much better?

 grahamforthor - image created by Misopogon

  Def. Recruits 4*+ Recruits 4-5* % 4*+ Retained 4*+ Ret %
Mich '09 48 26 54.17% 16 61.54%
Mich '10 54 28 51.85% 18 64.29%
MSU '09 65 10 15.38% 6 60.00%
MSU '10 61 12 19.67% 8 66.67%
OSU '09 60 36 60.00% 28 77.78%
OSU '10 57 32 56.14% 23 71.88%

Kinda. No, not really.

The percentage of 2006-'10 recruits who are highly (4 or 5 stars) rated for Michigan are much better than Michigan State, and a bit under Ohio State. This actually went down from last year, as Rich Rod replaced the outgoing talent with whatever he could get his hands on. Michigan State improved, from about 15 percent to almost 20 percent, but anyone who tells you "Michigan State is just as big as Michigan" are kidding themselves.

Retention rates for the highly talented only went up a bit this year, with the four losses above keeping Michigan from Ohio State's 70- to 80-percent clip.

What this means: we have just two more players rated four stars or above on the roster this year over last. It helps. But we didn't get exponentially more talented. The gains came by filling with young 3-stars.

The recruiting distribution shows a bit of this story:

charts2010-FIGURE1

I figure you can guess which school is which by the colors.

Michigan State hasn't changed all that much in recruiting (their one major difference is Gholston). Ohio State is a bit below last year's high.

Remember last year how Michigan looked like Penn State on this graph? Now they look like Ohio State, i.e. Ohio State Lite (they have a big handful of top recruits that we don't).

charts2010-FIGURE2

See?

Not as nice as Ohio State, because we haven't recruited as many of those really top echelon guys. But not really ugly. Against next year it looks better. And yeah, MSU looks about where it was last year: lots of 3-stars but precious few 4 and 5 stars.

Statement: Michigan has a fairly talented defense in 2010, and not in the "only a few talented guys and that's it" kind of way we were in 2009.

Your enemies always get strong on what you leave behind; Michigan State, you are Joey Zasa.

But talent's not the only story.

 

When I'm Dead, I'm Gonna Be Really Smart

threemiguelos - Image created by Misopogon

"Maybe you should come with me for a few weeks. See what happens. See how much you learn. Then, we'll talk about your future."

Remember: experience matters (just ask the Mathlete!)

Any prediction for this year's defensive performance is thus predicated on how much time the guys have played. The Mathlete looked only at starts, but the experience metric that coaches talk most often about is "time in the system," not starts.

First, a review of the go-to chart from last year: talent distribution of upperclassmen.

charts2010-FIGURE3

…and the flipside:

charts2010-FIGURE4

Figure 3 is more relevant to next year's performance. As we can see, last year was very thin – one or two guys recruited at each level. All told, 11 recruits, meaning if everybody played up to their hype (which never ever happens), we would have had an upperclassman team with some really good players and some really mediocre players. This year, there's a little more play but it's not all that different. Specifically, the tradeoff in upperclass talent is a likely Brandon Graham (6.1) and Renaldo Sagesse (5.6) for two likely Ryan Van Bergens (5.8) and an Obi Ezeh (5.5).

Straight-up, it's probably not a difference, meaning the performance level that Michigan's defense gets from its upperclassmen in 2010 will probably be about what it got from its upperclassmen in 2009. It is still well below that of Ohio State, and like last year, is drawing from a significantly smaller but significantly more talented pool than Michigan State.

Figure 4 tells the other side: we now have a more-than-respectable pool of low 4- to high 3-star talent of young depth players. The 5.6 to 5.8 pool is where you find guys like Mike Martin, Troy Woolfolk, Michael Williams, Brandon Herron, Cobrani Mixon, et al. What I mean is they're a crapshoot: some might become NFL-potential stars right away (Martin), some might become good players as upperclassmen, a good many become mediocre or depth players, and others end up being unremarkable linebackers at Division II schools. For the future, this is a good sign, although the lack of underclassman blue chips – what happens when you win just 8 games in two years – makes a truly elite defense before 2013 unlikely. For this year, it seems there is a large enough pool of young guys to present, conservatively, one or two more solid-to-very-good starters.

Also: Michigan State's youth = Ohio State Lite.

Then again, that's only if the talent is spread out evenly between the various sub-units: line, linebackers, and backfield.

A tough guy? I don't need tough guys. I need more lawyers.

warrenandfredo

Sometime before this offseason mercifully ends, an MGoBlogger will write a Diary in which he runs down the expected starters and applies recruiting/spring hype to replacements in order to demonstrate why this team should be "solid." Whether or not the word "stud" is deployed in this hypothetical diary, it will assuredly include something like the following:

"Defensive Line: Brandon Graham's loss will be felt, but Roh, Martin and Van Bergen all return, with [prize horse metaphor] recruit William Campbell, a sophomore, ready to step in."

From there, in order to maintain this "solid" conclusion, I predict the optimism-to-realism of expectations ratio will rise exponentially as the position gets further from the line of scrimmage.

The units, we know, are not even, but they could perhaps be better than they were last year. And because this is MGoBlog, we way we look at this is…

Charts?

YES.

How to read: The warmer the color, the better. For the Talent charts, Red is a 5-star, Orange is a 4-star, etc. For the Class charts, Red is a 5th Year Senior, Orange a Senior or RS Junior, Yellow a Junior or RS Soph, Green a Sophomore or RS Freshman, Blue an Incoming Freshman.

This year:

01 - Mich10 02 - Mich10

Last year:

03 - Mich09 04 - Mich09

Defensive line loses Graham, yes, but it still looks mighty good, with very good talent, depth and experience. A dropoff from last year could be minimal.

Linebacker, which has one more scholarship player than it did in 2009, is now the least deep position (where it was the most). The talent level is okay-ish, a bit below last year, though bolstered with a few more low 4-stars it didn't have before. The experience level is better at the top, with two 5th year seniors (Obi and Mouton).

Defensive back is radically changed. The talent level is, on average, lower, but out of double the pool. Experience, even with a boatload of true freshmen and a junior starter leaving for the NFL, only went down slightly, though it did this partly by the squad acquiring players from the offensive side (Teric Jones and Cam Gordon). There is a bit more talent here than in the linebacking corps, but it is so darn-tootin' inexperienced, their 2010 season looks likely to elicit a certain, common pre-teen malady from sportswriters on multiple occasions.

Just so you know that all the blue on the right isn't normal, and all the orange on the left is pretty okay but not awesome, here's this year and last year for a couple of land grant state schools that we play every year:

Ohio State 2010:

05 - OSU10 06 - OSU10

 

Ohio State 2009:

07 - OSU09 08 - OSU09

Michigan State 2010:

09 - MSU10 10 - MSU10

 

Michigan State 2009:

11 - MSU09 12 - MSU09

Bullets:

  • See what I mean by young? The 2009 and 2010 classes make up about half of each unit for our rivals; for us it's about 75 percent.
  • Again, Michigan State's talent pool is way below that of Michigan and Ohio State.
  • Ohio State last year should very well have had an exceptional defense (and did). They had plenty of both talent and experience.
  • tOSU should experience a very small dropoff this year, but is still a little more talented than Michgian and much, much more experienced.
  • Holy hell does Dantonio love defensive linemen or what?
  • MSU needs a big linebacker recruiting haul this year. They are moderately talented, but not very deep.
  • Michigan is relatively very young.
  • Michigan's 2009 team was an outlier, especially at defensive back.
  • Unless a true freshman cornerback is a once-in-a-decade wunderkind, the Warren loss is going to haunt us all year.

Your Life Could be Redeemed, But I Know You Don't Believe That

Lamberto

We haven't learned much you didn't already know. The defense in 2010 is likely to be better, but only a little bit better than its 2009 disaster version. The high attrition that wrecked the last three Carr classes has been better to players recruited under Rich Rodriguez and Greg Robinson. We are not Michigan State; we are a very young, slightly worse Ohio State.

There is potential here. One of the kids could emerge. The defensive line actually looks like a really good defensive line, with the caveat that no unit that loses a Brandon Graham can actually get better.

Best guess, this year's defense is not going to live up to the classic defenses of 2006, 1997-98. If you are waiting for the trilogy to end with the family finally legit, then you underestimate how hard it really is to find legitimacy.

940full-the-gher-part-iii-screenshot

Especially with the young secondary coming in on the heels of a unit that played a walk-on extensively, there is bound to be at least one Sophia Coppola ruining most every scene she appears in.

izvl6p

And that means there will be moments of excruciating pain.

940he-godfather-part-iii-screenshot

But with this great big pool to choose from all of a sudden, there's got to be at least one player, maybe two or three, who will emerge and really, really entertain us. Many, if not all of our favorite, characters will return so that we can witness their ends. So for those already familiar with the franchise, for those looking for good drama rather than another installment of the greatest story on celluloid, Michigan's 2010 defense will be, dare I say it, watchable.

And that's progress.

940full-the-godfather-part-iii-screenshot

fin.

Comments

aaamichfan

May 9th, 2010 at 11:04 PM ^

If Jahvid Best YouTube videos are considered "Schwartz Porn", this is what I would consider, "aaamichfan Porn".

 

Great work as always.

Seth

May 10th, 2010 at 12:29 AM ^

credit Misopogal.

I was trying to describe to her how I see the 2010 Michigan defense -- and I said it's a disappointment because we have had some epic defenses in the last 20 years, but that it will at least be "watchable."

And she said "like Godfather III"

And I was like YEAH! and went off to her about Sophia Coppola as an allegory for Obi Ezeh, and she stopped listening entirely, so I figured I'd just pelt all of you guys with my thoughts instead because you listen to me...I mean you really really listen!

(had the charts all made since last week but needed motivation).

Dan Man

May 10th, 2010 at 12:42 AM ^

You mention the significance of this being the second year under Robinson, but it seems your prediction that the defense will be "slightly better" is based primarily on number of recruits, experience, and star ratings.  So my question is, do you think that the year of continuity under Robinson actually gives the defense a good chance to be more than just slightly better?  Have you ever analyzed the effects of consecutive years with the same coordinator?

Thanks, and great stuff as always!

Seth

May 10th, 2010 at 7:25 AM ^

I made passing reference to this earlier.

It just wasn't in the scope of this study. All I sought to show was the strength of the talent pool at Michigan, for the defensive unit as a whole, and for each subunit. If our players vastly outperform the expectations of a good d-line, meh linebacking unit and super duper young defensive backfield, then it follows that coaching made at least part of that difference. Probably luck as well.

That is a fantastic idea for future study, though.

TESOE

May 11th, 2010 at 2:16 AM ^

to make this D hold water.

It's possible.  Campbell pushing Martin to DT would be a start (I like the Thor for BG trade - but no Nordic God can replace that hole alone.)  I saw signs of Campbell moving in this spring, but Magnus has moved him back to #2 behind Martin at NT on the unofficial 2 deep. 

Injury made the data collection pretty poor as to what the DL plan might be.  I'm hoping it will literally hold the line.  (Ideally Martin dominates and RVB's experience rules.)  If Banks is truly the best option then I would like to see him get it done.

Clearly one CB position is up for grabs.   Hopefully we won't wait for a Horror show before starting a Freshman - if that is what the situation calls for.  Michigan would have done better with Donovan from day one.

We definitely have the horses - new and old.  Ezeh and Patterson (who is physically gifted) could make this D much more potent by stepping up.

Thanks for the analysis!  This is good.  MCalibur gets kudos for the graphic as well.  It is a classic and bulletin board worthy in every locker room.   Defense wins championships and will make or break this team.

M-Wolverine

May 10th, 2010 at 1:10 AM ^

I'm completely changing my predictions for the season outcome now...(him AND the Hulk on the D-line? C'mon!)

And though certainly accurate, I'm very afraid that we're headed for Godfather 3.  If I have to put up with Sofia, I'm hoping someone gives me some Andy Garcia too to make up for it.

Which I guess means we're going to have to wait for Godfather 4...unless they reboot the franchise.

hausoian

May 10th, 2010 at 1:48 AM ^

This is an awesome post. I was hoping for a better conclusion, but I guess I'll settle for "watchable" at this point.

But to get 7-8 wins, we don't need the '06 defense. I think just from looking at the talent we have on the roster (albeit raw as can be) we won't have to worry about a Darius Willis turning the corner on us. 

I'd be interested to see how the multiple looks on defense will help us out. For example, moving around Roh and putting Kovacs in good situations will hopefully get the most out of what we have. 

uminks

May 10th, 2010 at 2:51 AM ^

I agree, championship caliber teams always have an excellent defense.  We are far from that but hopefully we'll see some improvement on defense this season. Enough to get us 7 wins and a bowl bid. If we start winning then the blue chip defensive players will start to choose M. My hope is that the offense will be a  bit more mature and may help us to win some of those  close b10 games this season.  And there is a trend in place with RR. In 2008 only 3 wins, 2009 we had 5 wins...may be 2010 we'll have 7 wins...I hope!  A NC will be at least 4 to 5 years down the road, that is assuming we continue to improve and we attract the top talent on defense once again.

jg2112

May 10th, 2010 at 7:21 AM ^

I hope to God Brian is paying you for these columns. There's a William Shatner "Star Trek Convention" SNL skit that comes to mind when I read your columns.

Well done, but by God, nothing better to do?

jg2112

May 10th, 2010 at 8:09 AM ^

Yeah, you're putting a lot more thought into a question I think is properly answered by, "I think it probably will, there's a good chance of it, but then again, it might not. Hell, why don't we just wait until September and find out together?"

bluewave720

May 11th, 2010 at 12:39 PM ^

 this upcoming season has forced me to start routinely lying to my wife. 

"Hey, what are you so serious/upset/concerned/excited about?"

I honestly think she'd rather hear me mention another woman than listen to another synopsis on the state of Michigan football.  I know for a fact that it concerns her, in regards to my mental health, the amount of time spent thinking about something I have no control over. 

Excellent work, Misopogon. 

Rasmus

May 10th, 2010 at 8:53 AM ^

Is that right? In 2009, there were Campbell, Emilien, and Hawthorne, all of whom saw some playing time (the latter two mostly on special teams, I believe). That might help a bit from the perspective of coaching continuity -- these three basically came in with Gerg that Spring. That's good.

I think it's enough to hope for one of the 2010 class to be able to start. Maybe two if we count Justin Turner in the 2010 class even though he's a redshirt, since he basically had no chance last year after his NCAA Clearinghouse problems caused him to miss camp.

MGoShoe

May 10th, 2010 at 9:11 AM ^

...but the data in Figures 1 - 4 is displayed inconsistently when viewed from top to bottom.  Specifically in some cases, the data is 2009 then 2010 by team while in other cases it's 2010 then 2009 by team.  There are inconsistencies within each chart and among the charts. As I said, these are nitpicks, but correcting them would help with chart comprehension, IME. 

[edit - I see now why you did it the way you did -- data would be hidden in some cases if the order was chronological.  I'm sure you struggled with this issue and came up with the best solution.]

Since Rittenberg has linked to Misopogon's previous work in this series, I recommend that the MGoBlogerati alert him to this diary so that he can consider providing additional context to his posts; I just did.  For instance, reading this diary would help Rittenberg to fill out this post 

MCalibur

May 10th, 2010 at 9:12 AM ^

Good stuff, Miso.

Reading your stuff always makes me feel like I need to step up my game in content and writing.

The GERG photoshop job cracks me up.

2014

May 10th, 2010 at 10:11 AM ^

Really good stuff. So if our offense gets a B+ this year (just spitballing), and our defense gets a C (versus last year's D-), does that give the team a B- overall? Wouldn't a B- be 8 wins? 8 wins really feels right. Well not RIGHT, but correct.

Another thought for a future study that I've been wanting to tackle but lack the dedication, the true statistical advantages of over-recruiting (i.e. bringing in 28 recruits a year vs. 20ish).

BiSB

May 10th, 2010 at 10:26 AM ^

Thanks for that.  Now, you are required to continue this series at least once per week for the next 16 weeks.  By mid-summer we will be in desparate need of real content.

Plus, I'm really curious what Godfather XIX would have looked like.

steve sharik

May 10th, 2010 at 10:56 AM ^

Great job, Misopogon.

I fear, however, that this defense may not be "watchable."  Unstable ILB play + baby secondary is not a recipe for success.  In the spring game, the 1st defense only forced the 2nd offense off the field on 3rd downs 3 out of 13 times. (Technically it was 6/13, which is still poor, but one was Tate's pooch punt, another was Tay's dropped TD pass, and another was followed by a 4th down conversion.)  A major correction for this defense needs to be protecting vertical seams.

This will take some time.  The question is, will the right people be patient enough? 

Maize and Blue…

May 10th, 2010 at 11:31 AM ^

that the best CB and quite possibly the best DL didn't see the field.  Throw in the fact, that they played vanilla, didn't blitz, and couldn't hit the QB and the "spring scimmage" was set up for the offense.

I think they may surprise this year.  D line should be as good and the LB and safety play can't get any worse than last year.  They will need a find a way to get pressure on the QB and hopefully Gerg is working on different blitz packages now.  Being with the same coordinator for the second year in a row (!!) should allow the D to react first instead of think then react allowing them to play faster.

Nothsa

May 10th, 2010 at 2:41 PM ^

A couple of guys go down in the defensive backfield this fall and our defense will drop from 'might be watchable' to 'probably toast'. Misopogan ably demonstrates the depth the 2010 Wolverines have, but unless all those children are above average, we really don't want to see the second team dbs playing substantial minutes in B10 games.

Unless we're up by 3 TDs, of course!

leftrare

May 10th, 2010 at 11:14 AM ^

For those who find this exercise a naval-gazing waste of time, well then you shouldn't have bothered to read it, since you should already know what great posters like mathlete and misopogon do and don't do for us.

That said, here's an executive summary prediction that is broad and not particularly daring, but I think useful:

Both offense and defense will move up one quintile in FBS statistical rankings* -- the offense from 2nd to 1st and defense from the 3rd to the 2nd.

Five wins in 2009 confirmed the statistical ranking.  7 or 8 wins in 2010 would confirm my prediction. 

http://web1.ncaa.org/football/exec/rankingSummary?year=2009&org=418

* i.e., top 30 teams in total yardage and scoring are 1st quintile, found in the ncaa.org stats page (linked above).