Mailbag: Stats Love Us, Saban Manball Canary, Substitution Style, Cole Absence, Playcalling Approach
Number 3? For the statistically challenged, what do you think of this methodology?
S&P+ is as good as any other ranking system that drills into play-by-play data to get a clearer picture of a football game than scoring margin alone can give you. Bill Connelly, the guy behind it, also runs Football Study Hall. He does a lot of smart things. S&P+ is a valuable look at who is playing the best.
Unfortunately, it can only go on the data that exists and in early-season college football that's always going to be sparse. Meanwhile some folks will dispute lot of the assumptions S&P+ makes, primarily that turnovers are super random and not major factors in the rankings. It also values all games evenly in ways that humans aren't always big fans of. Utah is significantly below Michigan because:
- the Michigan-Utah game was about even down to down and turned on turnovers
- Utah did not significantly outgain Utah State or Fresno State
- Michigan yardage-murdered everyone other than Utah
S&P+ is not trying to be a descriptive ranking (ie: these teams have had the best season so far) but rather a predictive one (ie: if these teams were to meet who would win). Michigan has performed like an elite team so far according to S&P+, and I can see why it thinks that.
FEI, the other major ranking that takes more than score into account*, is more skeptical than S&P, but I think that's because that still bakes some preseason assumptions into the ranking.
*[AFAIK Sagarin only uses the final score.]
Can we manball it when even Saban flees to spread-type behavior?
It seems that Nick Saban has recently admitted that his current style is a bit outdated, that he needs to adjust to the recent trends in college football. It is pretty obvious that teams like OSU, Oregon, TCU, Baylor, even BGSU are seeing a lot of success by utilizing both up-tempo and featuring quick guys in space.
Can you speak to offensive philosophies such as Alabama and Stanford and how this may or may not be a concern for us going forward? I understand that "smashmouth" football is not mutually exclusive with up-tempo and quick guys in space. But it just seems to me that Harbaugh's style doesn't seem to emphasize either of these current successful trends.
Given how the season has gone so far I actually think Michigan might occasionally run into the opposite problem. They've been absolutely lights out against six consecutive spread offenses. (Not very good spread offenses, sure, but Michigan isn't holding these guys to 20 points and high-fiving afterwards. They are crushing opponents.) Meanwhile the Harbauffense is winning plays against teams that aren't always comfortable putting heavy D packages on the field or filling all the gaps Harbaugh creates.
Saban's move to a more spread and tempo oriented offense is a reaction to the many times his defense has been blown out of the water by those kind of attacks over the past few years. When the Tide get to line up against one of the remaining "pro style" offenses, the results are generally ugly. Ask Georgia.
Michigan might not have that issue. Durkin seems very comfortable devising ways to neutralize spreads. I will have trepidation when and if Michigan does come up against… well, pretty much just Alabama.
On and off and on and off
Brian or Ace-
Do you know, or, if not, could you ask someone, why Dan Liesman (I think that is who it is, at least according to my Mini-Program; it is #54) comes out a few yards onto the field between plays almost every time when we are on defense. It is as if he is not sure whether he is going in or not, but since he NEVER goes in, it is obviously for some other reason. Is there some rule about substitutions that this relates to, are we trying to confuse the opposition, or does he just like to pretend he might be going in? There has to be a reason, and I would think most MGoBloggers would love to hear it. Thanks
We've seen Ross and Gant also do this. It's just a substitution strategy. After the play Michigan sends guys who may or may not be in the defensive package, depending on what the offense does, to about the numbers. (Any farther could get you an illegal substitution penalty.)
If opponents send in two or more blocky-catchy types, the linebacker will stay in and a DB will be removed. Since every team Michigan has played almost never uses two or more blocky-catchy types the LB heads back to the sideline almost all the time.
Liesman specifically is interesting because Michigan usually has Ross available; I haven't noticed if sometimes he is poking his head on the field when Michigan's already in a 4-3. That would imply Michigan has a heavy package in case someone tries to manball them.
Someone was confused.
I wanted you to know how much I appreciate and enjoy your broadcasts of Notre Dame football. Your kind deference to Our Lady's University is a beautiful expression of the christian love that infuses your broadcast persona. Thank you so much! You are a good man.
May God bless you and yours.
I did flip over to the Notre Dame-UMass game when it was interesting for a minute and heard Hammond's dulcet tones. He's missed.
I assume that guy who made the Tom Hammond tie is in Congress by now.
[After THE JUMP: early drives allowed, Harbaugh's playcalling system, a search for superclusters.]
Adding to the list of Adidas wrongs. What really irks me is that the only word appearing on the front of the current Michigan football uniform is Adidas. Unlike most schools, U-M jerseys were famously clean of any identifying words- the signature maize & blue color scheme was all that was needed. The simplistic Nike swoosh, though an identifying trademark, is far less noticeable.
Hoarding disease is a problem with a lot of uniforms these days. In an effort to brand brand brand everything they've cluttered the front of the jerseys with a series of logos: Adidas's clunky stripes, the Big Ten logo, legends patches, bowl patches, a block M or three. There is too much stuff on these uniforms:
They're probably dropping the Legends patches and replacing the Adidas Triangle Of Tiny Text with the swoosh will help; they can ease back on the block Ms.
The Big Ten logo is going to remain a grating presence until the end of time. Because you need to be reminded who is in the Big Ten these days. And that Michigan—surprise!—is in it. But some guy gave a presentation where he muttered something about brand equity, so we're stuck with it. The best they could do is something like they did at Crisler:
Michigan technically complies with the league mandate to have the Big Ten logo on the floor… very technically. If Michigan could get away with a blue-on-blue Big Ten logo that would improve things. I bet some clever person in the league office has already put in a regulation against it, unfortunately.
Maybe a step too far.
Am I crazy for thinking that this is the best look for the away uniform?? Obviously the jersey will be Nike but I love the simple all white jersey and blue numbers. There is enough maize on the helmet and the pants. Maybe put the Block M or number each shoulder pad. I just think simple is better and this jersey is sharp.
I like simple. That might be a bit too simple even for me. It gives off too much of a generic vibe. Is that a Michigan jersey or a random high school from 1950? I do not know.
The above does avoid the clutter mentioned above. It even avoids the many, many iterations of maize trim that have never really come off:
I am so done with maize piping, and maize outlines on the numbers, and maize maize maize on a white road jersey. But the above suggestion needs something to distinguish it. The correct number of design elements isn't a jiggityzillion but it's not zero unless you're Penn State.
Maybe the stripes from the Sugar Bowl jersey:
That everyone liked those is indicative of how low our expectations are these days. I thought they were fine and they have the chest clutter—this partially self-inflicted with a superfluous block M—and weird thin numbers that kind of make it look like everyone is wearing a kids' size. But they weren't a collaboration between a six year old with a glitter gun and the first guy cut on every season of Project Runway so we liked 'em.
[After the JUMP: Bo Xs and Os, and moar Nike.]
thumbs up [Bryan Fuller]
Class size: fishy?
Since the 2016 Michigan recruiting class has already grown beyond the 14 scholarships that are known to be available, do you have a take or any insight as to how far Coach Harbaugh is willing to go with regards to oversigning? Is it simply a matter of players not being offered a 5th year or could we actually see Michigan take a step toward opening the Harbaugh wing of St. Saban's Memorial Hospital?
I have a hard time believing that we could get that draconian with recruiting, but is there a danger that we step too close to the line that coaches like Urban Meyer and Nick Saban crossed a long time ago?
A quick glance at the Depth Chart By Class shows a large number of redshirt juniors who are not currently contributors. At this point many aren't expected to be. Those guys can graduate and either take a fifth year elsewhere or head to the real world without anything about their departure being shady. There are between four and six candidates for the firm handshake on the roster.
Also, I've heard that there were a couple guys who were likely to take medical redshirts of the legit variety. I'm surprised we haven't heard any announcements about that yet—maybe there's enough room for the players in question to see if they can get back to where they need to be this year.
Add those two things together and you have 20 or so spots right now. It's reasonable to expect playing time and other attrition to get Michigan to the 25 they seem to be planning on—most teams in year one of a new coach see attrition like that.
I don't expect this to be a long term trend. Harbaugh's classes at Stanford ended up with 19, 17, 22, and 22 kids. That is a mere 80 in four years. (The transition class between Harbaugh and Shaw was 19, FWIW.) Harbaugh is clearly alarmed at the state of the roster and is trying to get in guys who he thinks are a good fit as quickly as reasonably possible.
It is possible it'll seem shady in February. Right now it looks like a reasonable approach.
Cat fight fix, more attrition stuff.
Ace's Stanford recruiting diagnosis reminded me of JH's public cat fight with Mike Hart and Jamie Morris. I did some digging yesterday on whether those fences were mended and found general statements from Morris about Jim being the right guy for the job before his hire in Dec/Jan, but nothing direct. This scuffle was a huge deal at the time and many wrote off Harbaugh for good. Do you know if this was all swept under the rug or if we've kissed an made up?
I'm not sure if Mike Hart buried the hatchet with Harbaugh. I didn't hear anything about it during the search, and it is possible that he was omitted from the "everyone call Jim Harbaugh" list, whether by accident or on purpose.
I have heard that Harbaugh and Carr had a conversation about a lot of things that did directly address those comments to the satisfaction of both men. Thus Carr's public advocacy of hiring Harbaugh even before that was accomplished. From my impressions of both men I'm guessing they're never going to be best buds; Carr was clearly practical enough to identify the best option for Michigan's coaching search.
Also in regards to Ace's post, it seems like this year will serve as a decent case study for how JH will handle 'crootin. We are taking guys at an astonishing pace this summer, and a class that is estimated to be near the 28 man limit is already filling up. Like most, I found that a bit unsettling and hope the additional public attention at UM curbs this activity (especially if Jimmy's going to cast stones at OSU 'crootin).
As I mentioned above, Harbaugh history in terms of attrition is very conservative. Some of the decommits Ace detailed aren't how I'd want Michigan's coach to go about things, but at least those guys were able to get the picture relatively early and find places.
[After THE JUMP: "his guys," speed, Harbaugh counterfactuals.]
Dytarious Johnson is mean
The question we no longer have to answer about basketball
Does it concern you yet that Harbaugh and staff are going after so many 3-star or less recruits (and even unranked ones) rather than shooting for more 4- and 5-star types? Might JH be underestimating his own standing and instead still be in "I'm at Stanford" mentality (i.e., "I need to find the hidden gems because the 5-stars are going to USC, Alabama, and such")?
Thanks for the blog, and give Ace a raise.
Hail to the bloggers,
This is so overblown. Michigan has ten commits. Five of them are composite four-stars (Swenson, Onwenu, Peters, Falcon, and Evans). Of the five who aren't, one committed to Brady Hoke (Harding), one is (probably) a fullback (Reese), and one picked up Nebraska, LSU, and Florida offers after his commitment (Davis). The two other guys are Kiante Enis and Dytarious Johnson. Enis ran for three thousand(!) yards last year and Johnson looks like a BAMF on his Hudl film.
That is not a high flier rate thus far. The two guys who truly qualify are both gentlemen an expert talent evaluator has seen in person.
Meanwhile, here is a list of high four star recruits who Michigan is thought to lead for: NJ WR Ahmir Mitchell, NJ WR Brad Hawkins, PA TE Nasseir Upshur, MD OL Terrance Davis, WI OL Ben Bredeson, MI DE Khalid Kareem, and NJ DE Ron Johnson. They are at or near the top for five star NJ DT Rashan Gary and CA LB Caleb Kelly.
They won't get all those guys; they'll get a healthy chunk, and they'll get involved with more guys down the road. It's not going to be an Alabama class but it should be comfortably top ten.
And that's only half the reason recruiting concern is overblown. The other half:
That class was Andrew Luck and three stars. It followed a class that was all three stars, and ranked ninth in the then Pac-10. Stanford was slightly better than that when those classes bore fruit. Recruiting is important; coaching is more important.
[After the JUMP: Countess impact, concerns that Michigan's skill position players are no better than Iowa's, outrageous afro.]
you asked for it
"Soon he will start appearing in historically significant photos and no one will remember that he was not, in fact, present."
Harbaugh put his Jim Harbaugh on the Declaration of Independence, and war was avoided. The British decided to do anything else at all; Harbaugh was forced to invent the game of baseball so he could play it with himself.
Shot clock effect on upsets.
Given that lowering the shot clock from 35 to 30 seconds had little to no effect in the NIT, and that we can expect the same for a full season, I wonder if a side effect of the change might be fewer upsets. While efficiency might not change, the number of possessions will. I would think that with more possessions the better team is likely to win, because more possessions mean less randomness and greater reversion to the mean overall.
Give EMU 50 possessions against Michigan vs. 100 possessions against Michigan, and I would think that they would have a better chance to win with 50 possessions than 100. Could the 30 second shot clock actually make March Madness less maddening by reducing upsets? Thoughts?
-A slightly amused reader who still hopes for upsets
I think that's correct. I still remember that game back in the Amaker era when Illinois was at their apex and Michigan was rolling out Dion Harris and walk-ons named Dani. Michigan's strategy was to run the clock down without running offense and have Harris take a contested shot—the most Amaker strategy ever—and it worked for a while.
Anything that increases the number of trials without making those trials significantly less reliable indicators of talent should reduce upsets. It should be a real effect, but it might be so small as to not be reliably measurable. Maybe Kenpom will address it once he's got a big ol' bag of data.
I have gotten a lot of questions/assertions about the 30 second shot clock—far more than I think the change warrants. The differences are going to be minor. The median NCAA team saw only 10.7% of its shots go up in the period of time just erased. Some of that time can be reclaimed by being more urgent about getting the ball up the floor. (For example, the NBA's back court violation is an eight second call, not a ten second call.) The net impact is likely to be less wasted time and approximately equal efficiency. That's a good change for the game.
More on shot clock
I don’t believe this will affect the quality of shots as much as it will affect substitutions…
On a number of occasions I watched several teams, Wisconsin and Michigan included, essentially ‘waste’ at least 5 seconds tossing the ball back and forth outside the 3 point arc without any other movement. Case could be made this was simply being used to offer the players a short rest on offense, meaning that the top players likely play longer before substitution.
This may mean that teams with deep and talented benches gain an advantage…so the question may become whether it is the team with the best starters or the team with the best top 9 that wins.
-Howard [ed: a basketball referee]
There's another effect: if teams do decide to make those five seconds up by being quicker that's going to result in more pressure to get up and down the floor and more tired legs late in games. That'll be something to watch next year: does the percentage of bench minutes go up as a result?
Again: probably marginal impact but one that I would argue is unambiguously good.
[After the JUMP: another theory of baseball competitiveness, sea cucumbers.]
skill position recruiting under Brady Hoke
Where is the talent?
Hoke's recruiting classes were consistently highly ranked, bringing in 4 and 5 star athletes coveted by other reputable football programs, yet I keep reading that we lag many Big Ten teams in talent. I find it hard to believe Hoke is this bad at talent evaluation and missed on this many athletes. What do you think of the innate talent on the roster? Is this a pure development issue, or was Hoke that unlucky or bad at identifying talent?
Peter from Horsham, PA
It could be there, but concealed by the Hoke chaos.
Last year's D was 14th in yards per play and returns eight or nine starters, depending on how you define the term. The only personnel losses were Raymon Taylor, Frank Clark, Jake Ryan, and Brennen Beyer; Michigan gets Desmond Morgan off an injury redshirt and adds Wayne Lyons.
It didn't feel like that good of a defense, though. Part of that was the quality of competition. Michigan is 41st in FEI despite having shiny regular stats. (MSU is one spot ahead of them in 40th.) Part of that was the offense putting the D out there so often in bad situations that they were going to crack eventually. Many, many games over the past two years have had a similar pattern. The defense scratches and claws to keep Michigan in it and then they collapse in the fourth quarter because they're out of juice.
The other part of that was Michigan going to a man press defense their team simply could not hack. Blake Countess was exposed for half the season before they went back to their old ways. Is that "talent"? In some ways. But Hoke took an All Big Ten sophomore and destroyed him because he couldn't see that he was being put in a terrible situation for his skill set.
Meanwhile on offense… I got nothing. With the exception of projected TE Devin Funchess, Michigan recruited horribly at QB and WR under Hoke. The Chesson/Darboh class was all right, but they appear to have missed on all three guys they took the next year and the jury is very much out on last year's class. QB… I want to throw a rock in Hoke's general direction every time I think about it.
Recruiting very badly at QB is a way to have your recruiting rankings far outstrip your ability to play football. It's only one spot. Michigan's piles of OL are five. Michigan has nothing to show from the Hoke era other than a three star they took early with little competition and Shane Morris, who it is now clear was wildly overrated by the services. Morris is the kind of boom-or-bust guy you can take happily in a Harbaugh context where the roster has five or six or seven QBs on it; he is a disaster waiting to happen if the two guys in front of him are Russell Bellomy and air and the guy behind him is a middling three star. Harbaugh set about fixing that, at least.
Meanwhile the places Hoke did best—the lines—are the ones where you have to wait the longest to get a return on your investment. OL are particularly unpredictable.
And, yes, I do think Hoke was pretty bad at IDing talent. Look at the Dukes/Jones/York WR class, or the performance of the tailbacks he brought in, or the Wangler/Ferns/Furbush/Winovich LB class. Do not get me started on the tailbacks. Look at the decision to move to press man.
Defensive line is an obvious exception, where Michigan's turned two three-star Glenville projects into real players, but that just re-confirms that Hoke was a pretty good DL coach who ended up in way over his head.
Uppity so-and-sos with their thesauri
Excuse me while I correct your nerdom with my (apparently) superior nerdom. Ferret Tecmo Bowl Bo Jackson would not be the correct name. What you are looking for is Ferret Super Tecmo Bowl Bo Jackson. In the first Tecmo Bowl, Bo Jackson was just a man. It is in Super Tecmo Bowl that he became an unstoppable run-circles-around-the-cpu-god. Surely you did not mean to name the ferret after the lesser of the two Bo Jacksons.
Bro. I said I would brook no criticism about my asymptotically perfect ferret names, and yet here you are, brooking your ass off. Well, BEN, clearly "Ferret Super Tecmo Bowl Bo Jackson" is clunky and unwieldy and stupid, as any six-year-old child could see. It is obvious to any Tecmo Bowl connoisseur that we are talking about the destroyer of worlds Bo Jackson, "Super" or not. Humor, prose, and lovemaking are all about omitting unnecessary flourishes. So too are ferret names.
[After the JUMP: Ben gets commuppanced hrrrrrd]