Possible Outcomes For The Remainder
OUR (ESTIMATED) PROSPECTS IN A HANDY GRAPHIC FORMAT
Just a brief diary this week about what the statistical crystal ball might hold for the remainder of the season.
For those of you who regularly check Massey and Sagarin and other sites which calculate ratings and/or projected win probabilities, you may have already figured this out, but for those of you that aren’t in the habit or have not yet dared to dream (a terrible dream, of course) at looking through these stats, here’s the current estimate.
In all seriousness, of course, we’re not quite doomed yet, although it would be fair to say that our chances at redemption grow fainter by the game, and by redemption, I mean at least gaining a berth in a certain bowl which, if you live in Metro Detroit, is within rather convenient driving distance. Actually, that’s probably not a conventional definition of the term “redemption”, so we’ll just keep it at “bowl eligible” and leave it for now.
Here’s the matrix – blue indicates a hypothetical win, with each number being the Massey estimate of this occurring (numbers current as of yesterday mid-morning). The estimated probability of each of the remaining possible combinations of wins and losses is in the final column:
So, yeah, the most likely single combinations of events is beating Indiana and then cruising to a cool 4-8 mark on the year, followed by losing out and going 3-9. After that, the next most likely combination would put us at 5-7 with wins against Indiana and Maryland. You can see where this is going, of course. Our collective chances of winning X number of games are below:
Bowl eligibility – 3 or 4 wins – currently sits at 12.18%. Hypothetically, if we beat Indiana, the chances at becoming bowl eligible rise to 20.06% if no other numbers change. If we lose on Saturday, the estimate on bowl eligibility would fall to a very manageable (if you want to call it that) 1.73% - if we don’t change any of the other probabilities in the matrix.
October 30th, 2014 at 8:43 AM ^
This place we are in. It is a sad place.
October 30th, 2014 at 9:18 AM ^
I think a lot of people thought of 7-5 or so as the floor for this team. It's sad to think, that it would have taken a few breaks here and there to even have the hope of reaching 7-5 as the ceiling.
October 30th, 2014 at 9:53 AM ^
...was a pretty good team! 5th year senior quarterback, a first rounder at WR, promising young RBs, a very deep defensive roster that returned nearly everyone, and so was ready to make the leap from "good" to "great," etc.
They defintely looked like an 8-4/9-3 team.
I miss Theoretical Michigan 2014.
October 30th, 2014 at 1:55 PM ^
I thought Theoretical Michigan was an 11 or 12-win team. I still think they were, personnel-wise.
October 30th, 2014 at 3:00 PM ^
Yeah, personnel wise I agree with you. But man, the last few years have been a crash course for UM fans in the things we took for granted for almost 40 years since Bo took over. Things like finding competent coaching, recruiting the right guys, proper roster management and retention, not changing with the times, support of the program alums, etc. They've all struck at various times to make the last 7 years pretty bad, and they've all been things no UM fan has really ever even thought about in decades. Hopefully the program just comes out the other side in better place instead of cannibalizing itself into oblivion.
November 1st, 2014 at 9:59 AM ^
Not that he is all that much worse than others, but you know how bad things are when your 5th-year-senior QB is a major part of what's killing your offense.
October 30th, 2014 at 2:29 PM ^
I remember when I laughed at Lee Corso when he said Michigan was only going to win 3 games this year on College Gameday...
October 30th, 2014 at 9:29 AM ^
I could see this team finishing the regular season 6-6, but I can also see them not winning another game. We've seen enough games now to see that the offense is putrid and any defense with a pulse can shut them down. Given the fact that every team seems to play good and with more emotion against Michigan, I think every game is a toss up, except OSU.
October 30th, 2014 at 9:48 AM ^
The second-most probable series of events is actually us losing out for the rest of the year and going 3-9 (14.53%).
October 30th, 2014 at 10:11 AM ^
Whoops. That's what happens when your paper notes / edits do not make it to the typed copy. Fixed it, forgot I did and left it out, and fixed it again just now. Thanks for the catch.
October 30th, 2014 at 10:05 AM ^
Footbaw!
October 30th, 2014 at 10:30 AM ^
and it appears that Hoke all along had Tied His Cart to The Nuss Horse. Hoke had done largely the same thing with Borges, which I wont go into.
I sincerely hope that The NussO can WIN these Next 3-- to get to 6-5, and then win the Bowl game. ANYTHING Less-- I dont think I can handle.
Hoke REALLY needs to Intervene with Nuss and say-- look I want more RunPlays to LT-Cole. FIND THOSE Plays and Block schemes. And then FIGURE-OUT the Pass-Play Ocall sequence.
Otherwise-- Hoke continues-- NEITHER of us will be here for 2015 . . .
October 30th, 2014 at 11:01 AM ^
are my own personal observations from Digging Around In The Weeds trying to find something within the NussO. And no one likes to have to Dig Around In The Weeds.
I mean-- there is Virtually NOTHING Else Out There, of any kind-of Sustained Ocall Consistency. At lease outside of DG RunPlay-HeroBall anyway.
But I TELL you-- there IS A FORMULA out there, for RunPlays to LT-Cole. And then PassPlays to a Butt-Funchess Combo. Then maybe RB-Draw, then back to LT-Cole . . .
October 30th, 2014 at 12:38 PM ^
We'll just be down here hanging out with Indiana, Purdue and Illinois.
October 30th, 2014 at 1:11 PM ^
Hmmm....
Mullins or Bill Cowher?
Mullins or Jon Gruden?
Mullins or John/Jim
Mullins or <gasp> Mike Singletary?
October 30th, 2014 at 1:45 PM ^
I predicted 11-2 this year. Wow. I was not fully aware of the possibilities.
October 30th, 2014 at 1:49 PM ^
data and stats. College football is a game of emotion and this team will rally hard to win out and save Hoke's job....or beat Indiana and lose the rest.
October 30th, 2014 at 1:49 PM ^
I expect an IU win due to their QB and their defense.
Maryland and Northwestern and Michigan if they played in a round robin format on a neutral field in a series of 10 games would all fall out 5-5. They are almost identical mediocre teams. So I say we split those. And get hammered at OSU. So I have 5-7.
October 30th, 2014 at 3:02 PM ^
Isn't IU down to their like 4th string QB or something at this point?
October 30th, 2014 at 8:26 PM ^
Yes they have a true freshman string bean who was like a 2 star who they pulled the redshirt off 2 weeks ago. So if we lose to him, fuck it all.
October 30th, 2014 at 2:44 PM ^
I got to stay out of Michigan to keep my head attached?
October 30th, 2014 at 3:09 PM ^
Thank you for this analysis and the mathematics certainly pass the "smell test" and seems reasonable.
The one question is whether these games are truly independent events as if they were throws of a die. After watching Penn State nearly upset OSU, one can see that morale and motivation are big parts of college football. If the Wolverines beat the Hoosiers couldn't that in some way help their morale and raise the chances of beating Northwestern and Maryland?
Curious if some type of Bayesian or conditional probability analysis could be done?
Thanks again!
October 30th, 2014 at 5:36 PM ^
To answer both you and the poster below, there actually is a Bayesian correction built into the ratings and predictions which the model calculates on his site, as I recall. Although he states that the basis of the measure is past performance, which it would have to be anyway really, using the numbers to estimate outcomes usually turns out pretty reasonable results.
October 31st, 2014 at 10:25 AM ^
One thing Penn state had going for them is home field advantage and the "White Out" Big incentive for them to play harder under those conditions.
October 31st, 2014 at 10:25 AM ^
One thing Penn state had going for them is home field advantage and the "White Out" Big incentive for them to play harder under those conditions.
October 30th, 2014 at 3:21 PM ^
Doesn't this model assume independent events? Or are the W-L probabilities already adjusted for non-independence? I ask this for clarification, not as a criticism.
October 31st, 2014 at 1:18 AM ^
Thought for the most part the schedule looked fairly easy if you took out the 3 losses to the better teams on the road. Then I thought they would lose one they should have won for 4 losses. Though, I was worried we have a greater chance of winning only 5 or 6 games then we would have winning 10 games.
October 31st, 2014 at 10:25 PM ^
Comments