The 2011 Defense: The Pace of Evolutionary Change

The graphs make it easy to compare 90th place to 91st place, but that's not what I want to look at.  I want to quickly compare year 1 results to year 2 results for all of those spots.  A y=x line would make that a little easier.

It's harder to go from 51 to 1 than 110 to 60, maybe there is a better way to compare than on an absolute change in ranking, i.e. 25 or 50 positions in your analysis. This is maybe overly simply but you could use a ratio: change in ranking/original ranking.

It's just a matter of if you'd like to know what's likely or what's possible!

Just: nice job, thanks for posting.

Averaging rankings has an inherent normalizing effect.  Unless somebody is number 1 every year, nobody comes out number one after you average.  You re-rank after averaging the base year data (by grouping in tens), it seems like you should re-rank after averaging the year 2 data also.  Worst year 2 performance is #120, best is #1.  Good idea to take a look at this...thanks for doing it.

Macro change is coming for a cute, cuddly, fuzzy beaver in one's face employed in 2010 as an edible tool publicly used by a Michigan coach (used to inspire a defense?). The beaver will become extinct before the start of this season (Thank God),and will be absented from the food chain. This macro change which will take place this season, should result in a higher than expected degree of tangible defensive success for 2011 as compared to 2010 resulting in a top 50 finish.

Even the memories of a bad taste in the mouth will be long forgotten. Yes!!

An appetite for victory also will be restored by a new (some say old species of) highly efficient coaching staff; potentially resulting in unexpectedly unprecedented change for the better.

I am surprised by the fact that 13.3% made a 50 spot leap.  I would have expected much less.  I find that encouraging for the upcoming season.

It has been a long time since I have been in any evolution classes, but I seem to recall that fossil evidence for gradual evolutionary change -ie a bird's small beak becoming larger, then medium size, then medium large, then large size- is lacking.  Some postulate that this is because fossils are infrequently formed and then more infrequently found.  Others believe that evolution is more fits and starts- either the mutation works and a new species propagates(the Large beak) or does not work and it quickly dies off.  I thought that this latter theory was now more accepted, but perhaps I am out of the loop. 

As an aside, I heard Stephen Jay Gould give a talk at Michigan, and he was great.  I miss opportunities such as that.  We don't get that much here in Hawaii.  Of course, there are other aspects of paradise that make up for it.

What it reveals is obvious.  What it conceals is the real mystery.

I like that so many of the educated readers of the blog have set out to find data-driven reasons for optimism with this defense, even when just about all of them result in disappointment.  I think the bottom line in my mind are these facts:

• Realistically we shouldn't expect more than a 20-30 spot improvement in the rankings this year.
• If there were ever a team to fall into line as an outlier for a larger jump it would be this Michigan team for all the factors mentioned
• It really will come down to health.  Last year would have been different with a healthy Wolfolk and Martin the entire year.  Same thing goes for this year - aside from DE, SLB, and maybe CB there is little depth and experience beyond the starters.

To me, this is the best line in the entire piece, which is an interesting examination overall. It's also my main reason for hopein 2011: What are the statistical odds that one head coach, in three seasons out of 131, would oversee the three worst defenses in school history? He not only sustained ineptitude, he was able to achieve greater ineptitude with each new season. One would think that, perhaps due to nothing more than statistical norms, that at least one of those three defenses would be superior to some other horrible defense in the other 128 seasons. But that was not the case.

The message is that the evaluation, schemes, positioning of players and game-planning were so bad that RichRod and his staff were able to fail well below what could be expected statistically. And it is why I continue to believe the defense, now that it has proven defensive coaching, will revert to the norm and make a substantial improvement. And this gives us a reason to hope that the current players may have more talent than we've seen translate to the field, and the improvement this season may be greater than what could typically be expected, simply because the sustained ineptitude defied all norms.

of D's that jumped 50+ spots because Michigan has everything needed, i.e.,

1. elite D coaching

2. lots of returning starters

3. lots of *experienced* returning starters

4. above average 4-year recruiting

IMO teams with 1-4 are set up for maximal progress.  In '11 I expect Michigan's D to be somewhere around 30; and then 15-ish in '12.

btw, great read.  enjoyed it.

First, nice job.  I think that problem that you are encountering is that there are just way too many variables to take into account to accurately predict anything (as you noted).

So, my way less scientific approach:\

Baseline - we were around 110 last year, depending on which metric you used, but bottom line is that were were as bad as a defense can be:

1)  Returning 8 defensive starters on a VERY young team - no more true freshman filling the ENTIRE secondary.  Experience is worth a ton.  Worth 15 spots

2)  Return of T-Wolf, a 5th year senior who may be our best secondary player on a team that was not so good in the secondary - worth 5 spots, minimum (because he will be able to at least cover a team's #1 receiver)

3)  Mattison / Hoke replacing GERG - Generally, a coaching change is a cause to expect some difficult periods of learning the new system.  Here, we are replacing a man who rubbed stuffed animals on the faces of 18-22 year old men during football games with one of the top DCs in the country.  From comments by Mattison, the team did not know how to do such basic things as tackle, cover, choose proper pursuit angles, etc.  GERG also played players in positions that they should never have played (Roh, for example).  Simply learning how to cover, tackle and be in the right place is worth 10 spots, and playing guys like Roh and BWC in their natural spots (while also getting a full season of Demens) is worth another 5 spots.

4)  No 3-3-5 - the 3-3-5 was an epic failure - it is simply not suited for B10 play.  Returning to a 4-3 instantly improve our pressure on opposing QBs, result in these things called "sacks" and take away some of the pass game from opposing QBs.  Worth 10 spots.

5)  MANBALL Offense - Disclaimer - this is NOT a criticism of the spread and shred that RR run, or a commentary on the effectiveness of last year's offense.  RR ran a VERY fast paced offense.  When it worked, we scored quickly.  Whe it didn't we gave up the ball quickly.  Either way, the defense was back on the field quickly.  Under the new system ,there will probably be less possessions for both teams, and time of possession should be closer to even.  Simply being on the field for less time will reduce the # of yards, points, etc. that our D gives up.  Worth 10 spots.

6)  Turnovers - we turned the ball over a lot last year.  This put the D back on the field and in a position to give up yards/points (usually both).  Denard was, for all practical purposes, a freshman last year.  In his second full year as a starter, I expect massive improvement in his turnover #s, as is common for second year starters.  Result - less situations where our D comes on the field with the ball well into Michigan's redzone.  Worth 5 spots.

7)  Schedule.  I do not expect another 67 point, triple overtime game this year (Illinois).  We will not be facing an OSU team led by TP.  These two factors alone are going to improve our final D #s by a few spots - 2 spots, maybe?

Total:  Improvement of 52 spots, putting us at somewhere around 58.  Right around average, which would be a massive improvement from where we were before.   

This whole diary jibes with overall common sense - mostly, good D's stay good, bad D's stay bad. Outliers correct harder to the middle than those in the middle of the pack. Fine.

I do think that all the studies that use Rankings and not yardage or points tend to be misleading. For instance - there's about 10 yards a game difference between the #75 defense in the nation and the #93 defense in the nation. An 18 spot disparity in rankings suggests there's a meaningful difference where there really isn't much of one.

Good read, thanks.  There are so many variables involved that trying to extrapolate one team's performance year-to-year based on what other teams have done is almost meaningless.  Assuming the schedule is more or less the same, for me there are three main factors that can influence a movement up or down.

- Quality of player

- Quality of coaching

- Experience

We stand to have significant improvements on all three this year.  That's what I'm banking on.

My research is in an evolutionary field so I appreciated this. I'm certainly hoping that our relative equilibrium on D is punctuated fiercely this season.

One correction, in case anyone else was confused: "salutatory" should be "saltatory". The latter refers to a sudden change between generations, whereas the former refers to a welcoming gesture or statement.

would I be forced to remember my Bio 172 class from last year and relate it to football. God I love this site.

And the next time I need to create a new account somewhere, my screen name will definitely be "Hopeful Monster".

Michigan grad with a paleontology degree, this might be my favorite diary ever.  

I was watching Hoke's first press conference, yes, the whole thing, again... and anyway he mentioned that SDSU went from #113 ranked defense to #43 while he was there.   So, NCAA.org to the rescue with stats...