Michigan Monday vs. MSU up at the Ozone
Michigan Monday vs. MSU is up at the Ozone. LINK: http://theozone.net/football/2012/PennState/michiganmonday.html
Tis a very good column. Really, go read the whole thing, especially commentary on Michigan vs. MSU (quoted from below.) Give Gerdeman credit: I haven't read a better description anywhere of the nature of this rivalry.
Big Brother finally rose from the ashes and bested Little Brother with a field goal with five seconds left in the game. Michigan's 12-10 win over the Spartans was ugly from the get go, and to me that's why it was just about perfect.
Michigan State and Michigan are supposed to brawl. Everything has to be next to impossible, or else it's just not right. Of my four favorite college football rivalries, Michigan is involved in three of them. [EDIT: what??? Michigan is 3 of your 4 fave rivalries, and you're an Ohio fan? what's up with that??] I love this rivalry because there is such a level of hate that respect can't permeate.
There is something to be said for the kind of hate that eats out your gut and invades your sleep. Mark Dantonio has it for Michigan. Dantonio lost, and there is no team that he would less want to lose to. It's that Little Brother Syndrome which has spiked this rivalry since Dantonio's arrival. He hates Michigan, and has done so long before he was ever the Spartans' head coach.
When it came down to it in this game, he had Michigan State playing as well as they could have expected to, and the Wolverines still came away with the win in a game where they never even reached the endzone. There's beating your rival, and then there's trolling them. On Saturday, Sparty got trolled.
Some bullets from the offense:
- Spartie did a great job vs. Robinson forcing him downfield . . . with the exception of Dileo.
- Speaking of Dileo, he was an excellent security blanket, and should be looked at more the rest of the season.
- Again, why isn't Gallon targeted more? The throwback screen is so effective, but was only called once.
- Dileo & Gallon, 2 little RR holdovers, are the best playmakers in passing game.
- Funchess reminds Gerdeman of Kellen Winslow. [EDIT: we should be so fortunate! May it be so.]
- Denard has issues with passes being batted down at the line, and also with linebackers (like Bullough) in the passing game.
- Other than his 38 yard run, Toussaint had 14 yds in 9 carries, along with 3 yds in 2 receptions. Not good.
Michigan on defense:
- Bell is one of the most tackleable players in the country. Elusive, Bell is not.
- MSU did a very good job of attacking downfield, but Maxwell just couldn't quite hit open receivers.
- Floyd is vulnerable, beaten deep twice, with 2 TD's if Maxwell was more accurate.
- Raymon is doing a very good job & getting better.
- Eventually, Michigan's secondary will get burned with the lack of a good pass rush.
- The DL only has 4 sacks this season, which won't cut it. Ryan can't do everything.
- If Nebraska, NW, & Ohio complete half their deep passes, Michigan's in trouble. If not, Michigan wins out.
Special teams:
- When your best offensive play is a fake, you're probably going to lose.
- Kicking game was stellar from field goals to punting.
- With Gallon returning the ball 26 yards, why doesn't he touch the ball more?
Final notes:
- Michigan found a way to win against a team with a pulse when things didn't look good.
- The point of Michigan's lack of TD's is this: the key to stopping Denard is to make him one-dimensional. No one remaining will attack defensively like MSU, but if Denard's big plays can be minimized, Michigan's scoring will go way down.
- Defensively, the game against Nebraska is key, because Michigan hasn't faced this good of an offense. If Michigan can rough up Nebraska, "then it's time to declare Greg Mattison a witch and have him tried."
- If Michigan beats Nebraska, put them in the Big 10 Championship game.
This was Gerdeman's best "Michigan Monday" of the season, imhe. Again, go to the link and read the whole thing if you want more analysis. My props to the Ozone for some of the best Michigan coverage out there.
October 23rd, 2012 at 10:12 AM ^
October 23rd, 2012 at 11:04 AM ^
I don't think Miller can hit that consistently, and since his legs are the greater threat, Floyd will see similar duty against the Buckeyes. Probably the same for Martinez, but Nebraska seems more prone to self-destruction, so bend-but-not-break may be more the rule.
On netting out the runs, first off he's not dropping the best and worst and taking the average. He's simply dropping the best. And you can't net out that run, because it happened, and it was the result of good blocking and good running. Is that one long run out of several attempts likely to recur in future games? Sure. In the absence of any indications otherwise, why would you assume no? To put it another way, if the game had gone on another 4 quarters, I would fully expect Fitz to have several below average runs but break another long one. His analysis assumes he'd stay below average ongoing.
There was nothing odd about the run that would require it be netted out. If it was a broken play, or a fumble that kicked around an picked up 25 yards downfield, or a fake punt, it might make sense. But it was nothing more or less than a very good running play.
October 23rd, 2012 at 11:34 AM ^
I don't really care all that much about the dropping Fitz's long run issue, because anyone who had watched any of Michigan's games this season can tell you that Fitz has had trouble running the ball and that's the whole point the guy is making. BUT the logic is simply that his long run is an outlier. It's non-representative of his ability to consistently run the ball. That's all (and don't kid yourself, when he has one run for 38 yards and 9 runs for a combined 14 yards, that one run is a massive outlier). Keeping the long run in there skews the YPC in a misleadingly positive direction. When dropping one data point drops the yards-per-touch from 4.2 YPT to 1.3 YPT, that one data point is clearly muddying the waters.
October 23rd, 2012 at 12:04 PM ^
Removing outliers is non-sensical in football. Big plays are part of the game. Ignoring Fitz's earlier play this year (I'm not addressing that here), let's use this one game and extrapolate: if Fitz had 10 runs every game this year, 9 of which went for 14 yards and one of which went for 38, what would your expected result be in the next game? It would be foolish to assume 1.6 YPC.
You'd expect a lot of poor runs, and one big play. That one big play helps the offense tremendously. As a defensive coordinator you'd have to assume that event would recur.
October 23rd, 2012 at 1:09 PM ^
If your level of analysis is "I expect a lot of poor runs and one big play" then YPC or any calculated per-play average is an inappropriate metric. You'd be interested in frequency of big plays to poor plays, not average yards per play metrics. Leaving the 38 yarder in there and then proclaiming that Ftiz earned 4.2 YPT and labeling any further delving into the data as abritrary is ridiculous. Ignoring the total lack of a normal distribution in his yardage numbers is grossly inappropriate from any statistical standpoint.
As everything involving statistics and empirical analysis is concerned, it all boils down to the question you're asking. If you're asking how Fitz is doing on a per-play basis, that 38 yarder is really distorting his numbers. If you're interested in the probability of Fitz getting a big run per 10 plays, that's an entirely different question and one that isn't answered by looking at YPC.
October 23rd, 2012 at 1:33 PM ^
Removing Fitz's best run and proclaiming him bad in this game is what I have an issue with. I don't really care about statistical exactness. I know that if you were a coordinator and said, "throw out that big run early in the 2nd quarter, and he only averages 1.3 YPC, he's not a threat" (again, based on this game alone), you'd be making a huge mistake. And that's what the OP is doing. No I don't think you assume 1.3 YPC vs. 5.2 YPC.
How effective was Fitz in this game... What other question matters? That run was the primary driver to a field goal (and gave us multiple attempts for a touchdown from inside the 10 yard line), which was 25% of our points on the day. It was 12% of our total yardage on the day. It was a simple lead play with no special circumstances requiring it be netted out. Since it was relatively early in the game, it may have forced the defense to key more on Fitz, leaving other options open (that may or may not have been efficiently leveraged).
In footbal terms, foolish to throw it out.
October 23rd, 2012 at 2:06 PM ^
I don't care about statistical exactness
So you don't care about the statistical validity of the statistic that you're arguing is more valid?
To use your example of the DC looking at Fitz's stats (yes, I know no DC would prepare by looking at stats rather than game tape), he could just as easily say "on more than 90% of his touches (12 plays out of 13 touches = 92%), he'll struggle to get a yard and a half past the line of scrimmage (YPT after removing the 38 yarder = 1.3 yards)." Even if you leave that play in, the standard deviation is more than 11 yards, meaning that every time Fitz touches the ball, he'll go for somewhere between a 7 yard loss and a 15 yard gain 68% of the time, which suggests something is very off about his per-play stats. (The standard deviation after dropping the big play is 5.1.)
Leaving the outlier in, from a statistical perspective, distorts the per-play averages even if the play itself wasn't invalid. But I think it's pretty clear that we're just looking at this from totally different perspectives (empirical v. intuitive). Agree to disagree, I guess.
October 23rd, 2012 at 5:53 PM ^
I'm arguing for practical extrapolation of what happens on the field. Assuming his best run is removed eliminates a valid reflection of his (and his OL's) ability and potential threat.
Netting out an outlier is all about context, at least in a practical application. Fitz's long run came against the first string defense of a good D1 opponent, in the second quarter of a 0-0 game, on a base running play. Those are all valid conditions, so the carry should be included. For the same reason, I wouldn't net out a 7 yard loss under the same conditions. Doing so would remove an important datapoint re. the RB and OL's ability and potential threat.
The only time carries (or passes, sacks, etc) should be netted out are under non-normal conditions, at least in football. If you gained those yards against the 3rd string defense of Deleware State, in the 4th quarter of a 63-0 blowout, on a double-reverse throwback lateral, then it should be considered for removal.
October 23rd, 2012 at 11:42 AM ^
Michigan thought the same thing last year, and Miller/Posey torched the secondary. I think Posey is better than the guys have now, but Miller's deep ball is pretty good.
October 23rd, 2012 at 12:11 PM ^
He missed a lot of those throws as well. But that entire day was a break in tendency from what Ohio showed all year. It was a big gamble on their part, given they likely hadn't been practicing throwing the ball all over the place like they did in that game. And if you don't practice it and don't play it in games, it's unlikely you can execute it in the next game. But what did they have to lose?
I've seen Miller be dead-on and miss horribly this year. I think we probably take that chance against their receivers. But given I also think we'll score on them, maybe we'll be more prone to bend-but-not-break.
October 23rd, 2012 at 3:26 PM ^
Well Braxton Miller is probably a better running back than Bell. And don't sleep on Hyde, he has been a sledgehammer all season. Bell is probably better, but I wouldn't suggest it's a huge gulf.
October 23rd, 2012 at 9:33 AM ^
I was nodding my head in agreement right up until the end when he said Nebraska has the best offense Michigan has faced.
Does anyone really believe that Nebraskas offense is better than Alabamas?
October 23rd, 2012 at 9:41 AM ^
Nebraska has the #12 offense in the nation.
Alabama is #39.
October 23rd, 2012 at 10:17 AM ^
At the end of the day, I would rather have Michigan trying to stop Nebraska's offense, as opposed to having to deal with Alabama's again.
October 23rd, 2012 at 10:47 AM ^
Cool, Mich is the 3rd ranked pass defense in THE NATION, Minn is 6th... Do you think these are the 3rd and 6th best pass defenses in the NCAA?
October 23rd, 2012 at 11:40 AM ^
It just depends on how you define "best." Alabama might seem like the better offense because of their talent, but they don't put up the best numbers.
October 23rd, 2012 at 11:12 AM ^
Of course, Alabama's offensive numbers are depressed by the fact that its first string rarely has to play in the 4th quarter.
October 23rd, 2012 at 9:36 AM ^
not tall. When he spiked it and came stomping across the field toward Hoke like some kid with his Dad's helmet on, my wife collapsed in appreciative laughter. I am completely down with him as QB. Perfection doesn't happen in college football; Denard is better than perfection.
October 23rd, 2012 at 9:47 AM ^
Can't you pretty much say that about any team?
October 23rd, 2012 at 9:59 AM ^
If Nebraska can outscore Michigan over the course of four quarters, they will more than likely win.
October 23rd, 2012 at 10:40 AM ^
Drew Dileo reminds me of Crash Jensen, who was drafted by the Dolphins as a QB the year before they got Dan Marino. Jensen lasted 12 years in the league as a WR/H-back/slot receiver. His specialty was third down, and he had an incredible knack for getting open. He wasn't big or fast, and didn't have a lot of gaudy stats, but he carved out a very nice career for himself.
October 23rd, 2012 at 12:01 PM ^
He reminds me of Wayne Chrebet. Size and speed are wonderful attributes, but the basic ability of get open and make the catch is where the rubber meets the road.
October 23rd, 2012 at 12:21 PM ^
Wayne Chrebet is a hero for slow, white people everywhere.
Never forget!
October 23rd, 2012 at 10:41 AM ^
I'm not as worried about Nebraska's offense anymore. There gaudy offensive output is in large part a product of them beating up on crappy teams like southern miss, arkansas state, idaho state. Our defense will still have to show up and play well, but we are the best defense they have faced all year.
October 23rd, 2012 at 12:15 PM ^
Using Big Ten numbers only:
They're averaging 32.3ppg in conference, behind Ohio, Penn State, Indiana, Michigan and Wisconsin (although they do lead the conference in YPG by nearly 25 yards).
OTOH they've played the number 3, 6, and 8 defenses. Michigan has played the 4, T9, T9 defenses.
October 23rd, 2012 at 10:47 AM ^
October 23rd, 2012 at 11:54 AM ^