How Much Should The D Improve? Not Enough
Let's take some small sample sizes and extrapolate wildly. It will be fun. Here's Bill Connolly breaking down expected improvement from teams that return varying numbers of defensive starters:
So Cincinnati returns 11 defensive starters. That's probably a good thing, right? But how good? And how much can a bad defense improve in one offseason just because of experience? Let's take a look, shall we?
Average Change In Def. F/+, Last Three Years Starters
ReturningN Avg Chg in
Def. F/+1 1 -12.4% 2 4 -10.9% 3 10 -8.4% 4 32 -2.1% 5 53 -1.1% 6 69 -0.5% 7 85 1.1% 8 56 1.5% 9 37 4.2% 10 9 6.0% 11 3 5.4% So basically, if you return between five and eight starters, you are likely not going to change much, but three or fewer is a problem, and nine or more is a good thing.
F/+ is Connolly's advanced metric; it's play-based instead of drive-based like FEI. Don't be fooled by the % symbol—the metric is percentage based and from context it's clear the difference is meant to be added to the score, not multiplied. Since the best defenses are around +17% and the worst around –13%, 6% is about a fifth of the entire scale.
Michigan is, unsurprisingly, right at the bottom of that scale at 115th. They were 12% worse than an average defense down-to-down. The good news is they return 9-ish starters, losing Greg Banks, James Rogers and Jonas Mouton while reacquiring Troy Woolfolk. (They also lose Ray Vinopal and Obi Ezeh, but Ezeh had been replaced and Michigan should get JT Floyd back so let's call it a wash.)
The numbers are thin at both ends of the spectrum but, hey, extrapolating wildly from small sample sizes. Doing so says Michigan's defense will storm forward from 115th nationally to…
99th.
I have no source for this, unfortunately.
But wait! Our sample sizes are not small enough and our extrapolation is not making out with other nubile young extrapolations in front of a television camera. Bill added a second factor, the previous year's defense, and finds that a defense with an F/+ under –10% that returns nine starters should expect (for a given confidence level that is not high at all) to improve by 8.6%, which would see them get to…
82nd.
You might be able to argue that Mike Martin wasn't right and the team was even younger than the average team that returns nine starters and GERG is rubbing stuffed animals on the faces of other stuffed animals at a tearful tea party and for the first time in a long time they'll just run one damn defense per year and that they should expect to improve even more. You're probably setting yourself up for disappointment. Like installing the spread 'n' shred, digging out of a hole this big is a multi-year project.
Never underestimate the power of Hoke's pointing (Yeah, yeah, I already posted this but it got buried within like five minutes).
If he can kill the King of the Devils, he can turn this defense around from 11OhGodWhyth to something I can actually watch.
I will take 99th over 115th any day.
for a little perspective, the 82nd defense last year gave up 6 less points per game than UM, and about 50 less yards. So yeah, I would be fine with that for a 1-year turnaround. Even a sub 100 ranking means 2-3 points per game less.
considering how quickly Michigan scored (or punted) on most possessions, our defense saw a lot of possessions
Flip side is, the defense wasn't generating a whole ton of 3 and outs either, it takes two sides to contribute to a time of possession deficit.
I thought "How many games would we have won if we'd have just held each of our opponents to 6 fewer points than they actually scored?" Then I remembered, "Oh, yeah. Still just seven."
In 2008, Syracuse's last year with Greg Robinson as head coach, the Orange finished 101st in total defense. In 2009, under Doug Marrone and with Scott Shafer as DC, Syracuse finished...37th. Their scoring defense didn't improve nearly as much(101st to 81st), but that was in large part due to Syracuse's awful situation at QB(this was the year of the Greg Paulus transfer), as Syracuse finished 97th in turnover margin.
I think one of the things we're going to learn this year is that the previous defensive staff really was that bad. I can agree that Michigan becoming an elite defensive team again is going to be a multi-year project, but I don't agree that Michigan is necessarily doomed to field a bad defense again in 2011.
Exactly! Brian's fixations with how bad he expects this team to be and his use of stats to prove his point is getting as tiresome as his use of stats to prove how great out team was under RR.
While Brian certainly has/had a hell of a lot more faith in RR than I did, the remaining problems on D are a talent issue...that makes them, at least partially, a RR issue. Brian can and maybe will point out through the coming season how much better the O would have been with RR, but if the D is bad it really is starting to look more like an RR deal than a Lloyd deal. It's gonna be tough to blame too much on Mattison given his track record and the complete absence of players from the current (and hopefully long term) regime. Brian is suggesting, with evidence, that D improvement will be slower than hoped. We all (Brain included) know that if improvement is slow on D it won't be because Hoke/Mattison replaced RR/Gerg.
Brian just doesn't want his heart broken again, so he's afraid to fall in love.
Apparently jdberkley hasn't heard that yards are meaningless - only points matter. It doesn't matter that a lot of those points were due to the lousy offense, it's the job of the defense to stop scores no matter what the circumstances. Just like the offense should be expected to score on every possession even if they have no kicker and they're constantly getting the ball on the 20. Therefore, the Syracuse improvement was minimal.
These facts will not deter my belief that our new defensive coaching staff will spur an above average improvement in the defense. It's all about the intangibles!
proving that coaching does matter.
I refuse to even contemplate anything else, it makes my eyes bleed.
Here are some of the variables that favor Michigan significantly improving, that are completely ignored or not accounted for by that Cincinnati "statistical analysis:"
(1) GERG was coaching multiple schemes with success at none of them;
(2) the defense had 4 coaches total last year - one was a first year coach, another was a coordinator coaching the linebackers, and the other routinely caused his position group's players to transfer. This year, there are 3 position coaches along with a D-coordinator and a head coach (and GA Aubrey Pleasant) handling the coaching, all with significant experience.
(3) Last year - all those freshman. This year, not so much;
(4) Way more bodies near the point of attack in Mattison's offense;
(5) By this fall the restrictions on practice time will be gone, permitting the coaches more time to counsel their players;
(6) I presume the coaches will not switch schemes during the bye week;
(7) there are simply way more bodies on the defensive side of the ball;
(8) there are at least 6 players with at least one years' starting experience....
I could keep going.
Technically, there are NO bodies in Mattison's offense.
There are so many more variables going into the equation here than just previous starters and previous defensive rank. For one, we go from having a bumbling moron as DC to someone who has been a shit stomper everywhere he's been. Also, we're going from a system that took our D-linemen (you know, our best unit) off the field in favor of freshman secondary players. Furthermore, that system was run by the same bumbling idiot I mentioned above, who just so happened to know nothing about said system.
but didn't Brian point out many times last year that the extraordinary youth and inexperience of our D was one reason it was so bad? And that we simply had to be patient, at least with regard to that aspect, and it was unreasonable to expect much more from such young players?
It strikes me as counterintuitive to now proclaim that we shouldn't expect much improvement. Isn't it an accepted view among coaches that there are generally huge increases in playing ability between a player's first and second season, especially if we're talking about pure or red-shirt freshmen?
When you add to that basic fact that we now have a defensive coordinator who actually knows what the hell he's doing both in terms of scheme and in terms of managing HIS assistants, saying that the highest we'll rise is 82nd in the country is purposefully looking at the black side of things.
Back last season before Hoke was hired, I was so sure Harbaugh would never be hired that I promised I would take Brian and his GF out to dinner at the Chop House if he was. I will now make a similar offer: if Michigan isn't better than 82nd in the country in total defense after the end of the 2011 season, I will take Brian and his new bride out to the Chop House for dinner, cigars and dessert included.
Don't be too harsh on our boy. Brian is just being overly conservative. That whole "don't expect too much and you won't be disappointed." I get it.
Under-promise and Over-deliver. That's what pappy always said.
Naw, its more than that. Brian has been down on the Hoke hire and now seems to be fixated on looking for negatives to the same degree that he looked for points of optimism in regards to RR. One almost gets the sense that he will be disapointed if Hoke et. al. succeed where his boy RR failed.
Sorry, but I could not disagree more.
To each their own, I suppose.
Saying that someone's underlying and well documented prejudice against someone (Brian ridiculed the idea of Hoke being a serious candidate for the Michigan) is now colouring and affecting his thought processes in regards to the new administration is not an ad hominum attack. Saying that someone's basic biases are affecting his ability to think objectively is a perfectly valid line of argument. Its an argument that goes behind the surface reasons being presented as a valid argument to deal with issues of prolegomena. The line of my counter argument is this: Brian, you are no longer credible when criticizing the current staff because you have not overcome you inherent bias against them. It is the same reason most around here don't see the Free Press as credible. Wrapping those arguments in statisitics simply masks the underlying bias.
I don't expect him to embrace Hoke with a big love fest; but when you look at the level of excitement he showed when RR was hired and extensive analysis and educaitonal efforts in regards to the spread offence the contrast is startling. Now we get negativity, fears over his expectations that the offence will be screwed up and Denard's talents squandered (MOAR SHOTGUN PLZ) and now a "Don't Expect the Defense to Improve that Much" post. It bears a spooky resemblance to the Free Press' handling of RR.
How dare he use stats to try to explain what our expectations for the defense should be.
<br>
<br>It must be nice to live in a sunshine, lollipop rainbow filled wonderland.
Just because he brings in facts doesn't mean that THOSE particular facts are applicable. And just because we think Brian is being all negative nancy and that maybe we have reason for optimism doesn't mean we all live in sunshine, lollipip rainbow filled wonderland. Sweet Jebus.
Yes, Brian had a nice statistical analysis linked. But the results of the analysis are in an AVERAGE SENSE. You cannot do meaningful analysis based on simple averages. I expect Brian to delve deeper, look at similar teams that may have gotten the "Mattison bump" or other teams that parellel our situation to evaluate the chances of improving beyond what the average 10% improvement says of this analysis (or however it works). This analysis is way oversimplified and does not take into account extremely viable and important variables such as a new coaching staff.
One other thing...extrapolation is one of the single most dangerous things to do in statistics. While not impossible, it must be considered very carefully...and here it is not.
Don't crap on my optimism...there is suffecient reason for it.
"Saying that someone's underlying and well documented prejudice against someone (Brian ridiculed the idea of Hoke being a serious candidate for the Michigan) is now colouring and affecting his thought processes in regards to the new administration is not an ad hominum attack."
I'm not sure you know what ad hominem means.
It means "against the person." An ad hominum attack completely disregards the argument being made and attacks the person rather than dealing with the substance of the article. Its like saying, "Did you know that he is gay/cheats on his wife/beats children/is racist/etc. therefore you can't trust anything he says." I simply have taken his wel documented opinions of the past and say that they are colouring his present outlook and arguments. It is in fact the opposite of an ad hominum attack. The exposing of methodological bias is a perfectly valid form of argumentation.
Because your argument is against the person (his bias) rather than the substance of his argument. And it's still a logical fallacy even if you're right about his bias. Whether Brian is biased or not doesn't change whether this particular post presents a valid case for reigning in optimism about the possiblity that the defense will show marked improvement.
That's the issue at hand, not his bias. Your argument that he is biased would be relevant if you want to talk about whether he is being objective or not.
An ad hominum attack completely disregards the argument being made and attacks the person rather than dealing with the substance of the article
I simply have taken his wel documented opinions of the past and say that they are colouring his present outlook and arguments
So what you are doing is disregarding the argument being made and attacking Brian for being biased rather than dealing with the substance of the article?
You were doing a good job commenting on the article earlier, you should stick with that.
No, dumbass (that is an ad hominum), bias is part of any argument. Part of picking apart an argument is to tease out the bias and see if the argument is weakened or falls apart. Claiming an opinion is biased, does not nullify the argument; rather it should give you cause to go back and review the conclusions drawn from the data and see if it changes the results once bias is accounted for. Sometimes bias can affect which subjects are dealt with and how. Bias and uncovering bias is a perfectly valid line of argumentation and a necessary part of the pursuit of the truth of any matter.
You seem to think that just because an ad hominem serves some purpose, it must not be an ad hominem attack. Demonstrating that someone is a liar, biased or fat can certainly help win arguments (particularly if those arguments hinge on credibility or fitness) but they're still ad hominem attacks and logical fallacies.
The reason for this is because you'll never be able to say "you're biased, ergo you're wrong." At most, you can challenge Brian's motives for reaching his conclusions, but doing so does nothing to demonstrate why Brian's arguments are incorrect.
The posters above were merely pointing out that if you truly believe Brian to be mistaken, focus on demonstrating why his position is erroneous and not on the motivating factors for a few conclusions that don't rely on Brian's credibility.
I respectfully disagree, Captian. What separates an ad hominem attack from other lines of argumentation is that the ad hominem attacks the person separate from any reference to the argument at hand as a way of discrediting the speaker such that the substance of the argument does not have to be dealt with. An extreme example might be to say something like, Person X cheats on his taxes, why whould anyone listen to a guy who cheets on his takes. If overall character is a part of credibility, as in the example of a courtroom setting where the credibility of a witness can have great importance, then such an ad hominum attack might have some validity. Because person X cheets on his taxes we cannot trust what he is saying here today.
This is very different from the process of uncovering bias. One might argue (quite credibly too), that because of Brian's overall "downer" attitude on the Hoke hire that I am biased against giving his arguments a fair hearing. I can accept that my reaction to Brian's reaction to the Hoke hire colours how I percieve his argumentations as of late. You have to take that bias into account when I make the argument that I find his "cautions" about what we can expect with the Hoke regime to be weakly researched and inadequately developed, especially when compared to the depth of material he prepared when Rodriguez was hired. To say that he has a bias against Hoke and this is playing out in his articles is a perfectly fair line of argumentation and completely germane to the topic at hand. Bias deals directly with the substance and content of the material at hand and the process by which it became a finished product.
Uncovering methedological bias in research is an intergral part of the scientific process and the quest for knowledge. Rather than shying away from dealing with our biases, when someone confronts us with out fundamental biases we need to deal with the issue with honest self reflection. We might come to the conlusion after such reflection and realize that our bias has in fact skewed the results. We might go back to the data and after taking into account our bias towards a particular conclusion, realize that the data still supports what we have to say, bias and all. It about fully owning and taking responsibility for our ideas, begining with all the issues of prolegomena through to conclusion.
Saying, "You conclusion is affected by your biases" is not an ad hominem, but rather meant to sharpen the research and conclusions of the presenter. It should be seen as a valuable counterpoint that shapens the saw and not an unwelcome interloper meant to sidetrack the debate.
I actually think Brian is pretty realistic or even optimistic about the season. In the Maple Street Press preview (formerly HTTV), he predicted the season record to be 8-4 but was wavering between that and 9-3. I don't find this to be negative about the new regime. It's probably one of the most optimistic season predictions you'll find at this point. Perhaps, you have a narrative in mind and you're shaping whatever he posts to fit that.
Saying our defense won't be a ton better isn't an indictment of Hoke, and Brian seems pretty confident in Hoke's overall ability to predict 8-4 or 9-3 given his thoughts that our defense might suck.
Alternatively, our defense could be a ton better and still only be thoroughly average. That doesn't scream "Michigan Defense." An average defense with a very good offense puts you right in the realm of Brian's prediction, which I think is very fair. I'm hoping that the offense either stays excellent or the defense improves to slightly above average, putting us in 10-2 territory, but that's more hope than reality.
You mean the one that Brian plugged on a main page post recently. Seems like Brian should be teaching Dave Brandon about self promotion rather than accusing him of being too corporate.
If Michigan's D ranks 99th or lower I'll donate $100 to Beveled Guilt. If they rank 98th or better Brian agrees to meet Nick Sheridan one on one and call him Sherideath to his face.
was a convo starter, a la sports talk radio (albeit better, and classier with statistics and charts).
With such a radical departure in coaching and overall defensive philosophy, nothing about the 2011 M defense seems to be quantifiable, imho.
Quantify this, my friend.
82nd could be about right, but that would be a BIG help. Howeva...
As other people point out there are more variables at work. More than just the loss of GERG (who was a clown) and the 3-3-5 (and all its permutations) and RichRod (who meddled in the D, right or wrong), and the AC's (who were meh) you have...additional experience across the board; additional depth; the omission of the no-huddle offense, which forced our players back onto the field faster; blah blah blah, everybody gets it. I also think incoming freshmen will be a big help on special teams, helping in the poistion battle. I don't see many redshirts for our DE's and LB's since we have so much in the pipe behind them.
There are a lot of things going on here. I would never bet on this becoming more than a middling D in one year, but people are forgetting something...we only need it to be so much better. Competent would be a huge stride over laughable.
Hey Don.
If you don't think moving from one of the five worst defenses in football to slightly below average is a huge improvement, I don't know what to tell you.
There's plenty of reasons to believe that more than just returning experience will contribute to an improved defense (by the way, some of the teams in those crappy defenses with returning starters presumably also replaced crappy coordinators with better ones), and Brian has highlighted them multiple times. Nobody in this tread has said anything that hasn't appeared on the front page. So go ahead and say that we'll show better than average improvement, pencil us in as the 60th best defense in the country. I'd take it in a hearbeat.
Goodness, people. How did highlighting the fact that returning experience doesn't translate to as much instant improvement as previously thought cause everyone to freak out?
I think Brian's point is that anything significantly better than 82nd would be a big statistical outlier. It might happen, but if you expect it to happen, you're arguing against history. This is not "aagh Hoke sucks" but instead "don't get your hopes up too much". He's not being pessimistic, per se, but just attempting to calibrate where the line bewtween pessimism and optimism lies.
Besides, I think your first paragraph is basically answered by Brian's admission that the hole really was that deep - the general Brian tone started last year at "these guys are young give 'em time" but by the end of the year I think "GERG is TURRIBLE" was the conclusion.
Unless I misunderstand, this analysis has applied the average expected improvement. Which is fine and all but I suspect there are many reasons that Michigans situation is not normal and therefore averages may not apply.
However, I am inclined against this type of analysis because it tends to mute expectations by taking averages that may.... revert to the mean. I prefer to be more optimistic and just let whatever happens happen. The defense need only be good enough to win for me. A defense ranked 82nd may actually look pretty good to Michigan fans after what we've been through.
I would also wager, without looking, that Michigans offensive improvement from 2009 to 2010 blew the average expectation out of the water.
always contained an absolute requirement by everyone .... GERG had to go.
So GERG is gone, but this "statistical analysis" is only based on returning starters. Not strength of schedule, not which returning starters (D line vs. corners, etc.) but simply returning starters.
With the mass changes this football team will see from the coaching perspective alone, I believe this analysis is of no value whatsoever (down from next to nothing). Please Lord - no more analysis based on pure specualtion of the upcoming season.
Go Blue
Please Lord - no more analysis based on pure specualtion of the upcoming season.
Um... what other modes of analysis of the future would you prefer? Because unless you have a time machine or you subscribe to the teachings of Nostradamus, speculation is pretty much your only option.
If Gerg was as clueless as our fearless leader (and our eyeballs) would have us believe, exchanging him for a ham sandwich should give us a gain of about 10-20 spots in def. eff. just by itself.
Is that ham sandwich on rye or whole wheat?
That will have to mean we actually stop an important drive once in awhile versus never right? That we stop the run occasionally versus never? That instead of getting beat three straight times on wheel routes in one game, we may stop it the third time? If so, and we have a few less turnovers on offense, sign me up. Sounds like progress to me.
Comments