The Road To A Terrible Bowl Has One Less Bump

Submitted by LSAClassOf2000 on

THE ROAD TO A TERRIBLE BOWL HAS ONE LESS BUMP

Decided to do one more update along these lines...

With that win over Indiana, you will be pleased to know that bowl eligibility has become slightly more likely, at least by the revised Massey numbers that I threw into the matrix this morning.

How much more? Well, consider that – as of right now – our chances to win at least two games now stand at an estimate 26.65%, which is merely the sum of the probabilities of winning all three or any two games. At this same point last week, we were talking about a number which sat around 12% or so, but of course having no result for the Indiana game at that point would make that result naturally lower.

The matrix for the remaining games now looks like this:

 photo 11-4-2014PROBMATRIX_zps148cbe2c.png

Like last week, the blue boxes would be hypothetical wins and in each box, you’ll find the most recent available projection from Massey Ratings. Over in the “PROB” column, the green boxes are the scenarios where we end up bowl eligible and the yellow boxes are where we get sent home with a year’s supply of Rice-A-Roni, a ton of Turtle Wax and the board game version of the show that we’re now on.

Here’s the rough distribution for any number of remaining wins:

 photo 11-4-2014REMAININGWINS_zps4503062d.png

It basically says exactly what you would have thought. That is to say, it says we now stand a realizable chance at being a 5-7 team after having at least put ourselves in a position to be a potential 4-8 team.  It’s definitely progress and the picture grows more clear if not more rosy as we slide into this game in Evanston on Saturday.

If we win on Saturday and don’t change any numbers in the matrix for the moment, then bowl eligibility is essentially a 50/50 proposition (well, 52.48% technically), so the numbers rather make the next two games key, but particularly this next one. Why? If we lose to Northwestern and change no other numbers for the sake of providing an estimate, bowl eligibility is a 5.52% chance away then. 

Comments

Tater

November 4th, 2014 at 11:38 AM ^

The next two weeks, we will see if David Brandon's "resignation" has, indeed, freed the players and staff to perform on a higher level than they have this year.  It certainly looked like it against Indiana, but Indiana has almost "attained" tomato can status because of their QB situation.

I think the chances are in Michigan's favor more than the statistics would indicate, due to possible "addition by subtraction" concerning Brandon.  The coaching staff has gone from being under pressure to having nothing to lose.  Brandon won't be telling the coaches what plays to call and which QB to play.  

There is a possibility that the current Michigan team is better than it was last Friday.  We will find out the next two weeks.

michelin

November 4th, 2014 at 1:41 PM ^

then I would expect our rating to rise.  So, unless MD doubtfully beats MSU the same weekend (and especially if they lose or get banged up), I would expect our chances of beating both NW and MD to rise.   My guess is that our overall chances of bowling with a NW win would be about 55-60%.--a bit higher than the 52.5%.

treetown

November 4th, 2014 at 5:09 PM ^

If the Wolverines go on the road and beat Northwestern (a task they have had a tough time doing) their chances against Maryland at home must rise - especially with the suspension of one of Maryland's key WRs.

I like stats and complex stats as much as anyone but this type of prediction works better for teams which are consistent - and the Wolverines have not been a predictably consistent team in a long time.

michelin

November 4th, 2014 at 10:50 PM ^

....on Sunday, Michigan opened as a 2.5-point road underdog visiting Northwestern, but is a 2-point favorite as of late Monday afternoon.  Although betting sentiment does not always reflect win probabilities, my guess is that the chances of beating NW have risen after the terrble loss at Iowa (now a three game losing streak).  Even if the win percentages posted above already considered the weekend's games, I wonder if the models give more weight to recent games.  (whether they should consider streaks is a topic for debate).

uminks

November 5th, 2014 at 2:41 AM ^

had slipped to a MAC level team. If Michigan can beat NU on the road, I will be more impressed and would say a 6-6 record is very likely.  If OSU knocks off MSU, I think we will lose by 20+ points. If MSU defeats OSU, then OSU may not play as hard and it could be a close game in Columbus? The question becomes will Hoke get fired before or after the bowl game if he manages to finish 6-6?

4-8 15%

5-7 50%

6-6 30%

7-5 5%

wolverineesq

November 6th, 2014 at 4:54 PM ^

are the probabilities really independent? Shouldn't there be some sort of correlation? I.e. a loss to Northwestern should make the model less confident we'll beat Maryland or ohio.