# Fear and Paranoia in Ann Arbor

Submitted by Ryano on October 6th, 2010 at 2:41 PM

I feel like there hasn’t been enough analysis leading into the Michigan/Michigan State game this weekend, so I have really wanted to contribute. It occurred to me that we have a metric that has not been tapped yet (to my knowledge): mgoblog’s Fear/Paranoia and Desperate need to win levels. These numbers, part of every game preview, offer a window into the psyche of a man who spends his week living and breathing Michigan football. Certainly something must be gained from this data.

I started by averaging each level based on whether Michigan won or lost the game. I used every game Brian has predicted since the beginning of the 2006 season. The averages are as follows:

F/P: Fear/Paranoia level

D: Desperate need to win level

 F/P D Win 4.3 8.5 Lose 6.7 7.0

So this makes sense.  When Michigan is rolling, Brian tends to be less afraid, and more desperate to keep the trend going.  When Michigan is not doing so well, opposing teams tend to freak him out more, and the F/P level goes up. At the same time, Henri the Otter of Ennui starts making appearances, and Brian becomes less enthusiastic about Michigan’s chances of doing anything right ever; the D level goes down.

This is interesting, but we also have another variable in each game’s preview: Brian’s prediction. What happens when we combine these levels with Brian’s prediction for the outcome of the game?  We get the following:

Bad Upset = Win predicted, but Michigan loses

Predicted Loss = Loss predicted, Michigan loses

Predicted Win = Win predicted, Michigan wins

Good Upset = Loss predicted, but Michigan wins

 F/P D Bad Upset 5.9 7.3 Predicted Loss 7.6 6.7 Predicted Win 3.8 9.1 Good Upset 9.3 3.7

When the prediction matches the outcome, these numbers line up pretty well with the Win/Lose averages up above. But it’s interesting to see what happens when the prediction is incorrect. When Brian’s F/P starts to creep up over the 'Win' average, and the D level starts dropping, watch out for the upset against Michigan. Also, look at the ‘Good Upset’ numbers … granted, the data set is small because rarely does Brian predict a loss and get it wrong, however this is somewhat amusing. When the opposition looks frighteningly good (or Michigan just looks terrible), and Brian loses all hope and turns up the Morrissey, look for the Michigan win. The key is the D level has to drop; if both F/P and D both go up, it will likely be a loss.

In conclusion, the Fear/Paranoia and Desperate need to win levels are a force to be reckoned with. This is not a semi-humorous tally of + and - at the end of each preview, these are actually a reliable indicator as to the result of the game. These should be carefully reviewed this Friday to get an idea of what Michigan has in store for Michigan State. Actually, in all seriousness, I was surprised to see these numbers made any sense at all when totaled like this!

You found a trend with four seasons worth of data points? I'm shocked!

This isn't as much as it sounds.  During this time, only 3 times was a loss predicted, and Michigan won ... not exactly a reliable data set.

Are they really reliable indicators?  Can you post some 95% confidence intervals?

Wonder how this would change if the tomato cans were removed, Mathlete style, as these seem to be the games with the lowest fear / highest desperation levels.

Also, is a large spread between the F and D metrics (i.e. |F-D|) good for M?  If the team is expected to lose (high F, low D), maybe they play relaxed and pull out the upset.  If they're heavily favored (low F because the chance of upset is low, high D because a loss ruins the season), they would also of course tend to win.

Finally if you took the product of F and D for each game, you'd get something you might call "importance."  Does a team perform better / focus more for games that are important to it's fanbase?

The big thing that I am worried about is how State shut down the run last year.

By the way, anybody hear Valenti today?  He believes that Jones is going to Spy on Denard and nobody in the Big Ten is better at taking angles than Greg Jones.

As per ESPN:

[Michigan State] has outrushed the Wolverines 364-112 in the last two meetings

That's a very scary stat, until you realize DENARD!!! I doubt they shut us down this year like they have the past 2 meetings. As a matter of fact, I bet Denard beats that 112 yards by the half easily.

As for Jones, Adam Rittenberg also sees this as a way MSU might try to handle Denard, and Jones is no joke. He hasn't been one of the top tacklers in the Big Ten the last 2 years for no reason. The boy can play, so it'll be interesting to see what happens when that Unstoppable Force meets that Imoveable Object.

When Michigan is rolling, Brian tends to be less afraid, and more desperate to keep the trend going.  When Michigan is not doing so well, opposing teams tend to freak him out more, and the F/P level goes up.

It's more likely that we desperately need to beat crappy teams, but are less afraid of them while we're less likely to need to beat the best teams but are more afraid of them.

I am worried because State's offense is basically UMASS with better athletes. I also tend to agree with the worries of the first couple posters. I think State's defense may stifle our running game a little bit. I understand that Denard is a whole different monster and that our offensive line is much improved this year, but I still am not totally convinced about our running game yet. I think this game is going to reveal to us how good our offense really is. Not that Michigan State has a great defense, but I think their defense will be the best we have faced up to this point. I am hoping that this game both legitimizes and falsifies the data that has accumulated about our offense over the course of the first five games. Mainly that our rushing offense truly is lethal and dominant.

But this is the very reason why I am nervous--I am not totally convinced or certain that our offense is as good as it has appeared to be thus far. This game is huge because it will reveal to us whether we are good enough to play with the big boys in the Big Ten again.

I share your worries, but I do derive some comfort knowing the both Michigan and MSU played ND and we didn't need a trick play in overtime to win.  We marched down the field and stuck it to them.

The other thing that keeps me from losing sleep is that you generally can't set up your D to shut down the run without giving up something else, namely, the pass.  Don't know about you, but I have been very encouraged with our passing game. We have some pretty good receivers.  If Sparty structures their D for the run, let Denard pass all day, or at least until they adjust.  Then, the run should be there.

As for fear and paranoia, somehow the Pink Floyd song "Run Like Hell" keeps jumping into my head and I can't get past that.    Here's the first couple of lines and chorus just in case you never heard it.

Run, Run, Run, Run, Run, Run, Run, Run,
Run, Run, Run, Run, Run, Run, Run, Run.

I think there's an entire chapter of the playbook that we have yet to see.  Run plays that is.

That and this offense has gelled like nothing I've seen since the 97 Defense.  Even to the point of Denard kneeling in the end zone while the rest of the offensive team swarms him.

Expect the run game to be slowed down from previous levels.  And that is when the offense will unleash hell.

Gotta love an analysis analyzing the analyzer.

This kind of breakdown of Brian's predictions is something Brian himself must remain ignorant of, lest it influence his prediction-making.  It's like psychohistory -- it only works if the subjects remain unaware.

I don't think MSU will be dominating our O-Line on Saturday.  I think the offense will have a good showing. If they do stick more people in the box then DROB can burn them wth the pass, and if he does the run will open up. I think we will score 5 or more TD's on MSU.

Now the defense still worries me.  If they cannot stop the run it will be a long afternoon and the team wth the ball last may win. However, if we do stop the run there will be some 4 and outs by Sparty but there could be some big passing plays.  Final analysis, we cannot let MSU run on us and make long time of possession scoring drives. MSU's best defense will be long time consuming drives that keep DROB and company off the field.

The D for this game should be pretty high, I bet it's at least 9.

• +1 for RR hot-seat
• +1 for 3 in a row
• +1 for  bowl eligibility
• +1 for "see? it won't work in the B10."

Greg Jones as a spy, take comfort in that actually being the worst way that State could defend Denard. As a football coach spying a man up has this mirculous little effect on the psyche of the other players on defense. They begin to relax a bit thinking,"I don't have to make tht play, Jones has him." I've seen it time and again. Which is fine, as long as Greg Jones gets there. However, if a lineman happens to start blocking Jones, or even just gets  a piece of him, then you have no one responsible for DENARD! States best option would actully be to bltz the holy bejeesus out of Michigan, fill all the gaps with a man and get Denard in the backfield. It's the only way to stop him.....from running.

play a different guy because the stud is up against a mean defense?  I've made that mistake and discovered that the studs usually still perform in the big game.  Denard still gets 150+ rushing, and State's three TO's seal their fate.  Our D will play just well enough to win.

IIRC we should have Shaw back, and perhaps Carvin J.

You had me at "I feel like there hasn’t been enough analysis leading into the Michigan/Michigan State game this weekend".

this post made me both paranoid and desperate.