Oh, Crap

Submitted by Brian on October 15th, 2008 at 12:35 PM

So a few months ago Chris Stassen, maintainer of the invaluable Stassen.com, emailed me to note that his name was not, in fact, "Harold." I had been calling him Harold for years.

Oops.

So I don't know if this is revenge or whatever, but yesterday this hit my inbox from Mr. Stassen. It's possibly the most frightening thing I've ever seen. The elderly, children, and pregnant women are advised to stop reading now.

I warned you…

Last year I stumbled across a research paper on the odds of a Vegas favorite winning a game (e.g., a 1-point favorite wins 53% of the time, a 7-point favorite wins 69% of the time).  The research was based on Vegas lines and NFL games, so one can argue against applying it to the college game.  I think college teams perform less consistently than pro teams (meaning that a big NCAA favorite is less guaranteed to win than an equivalently big NFL favorite)... but probably it gives a decent rough measure.

The other problem I've had is how to get "lines" for games which are purely only hypothetical (e.g., Texas Tech vs Kansas in the Big XII title game), or games which are weeks away and not big enough for Vegas to publish an early line (e.g., Michigan at Purdue).  I've finessed this issue by using Sagarin's PREDICTOR ratings (his best estimate of team strength) to compute spreads.  Again, this sort of abuses the research paper's results, in that his spreads aren't the same as Vegas lines (though they tend to be pretty similar).  Anyway, if we're only looking for ballpark figures, we don't have a lot of other options if we want to do this sort of computation.

There are lots of various ways to throw rocks at this analysis.  I did a similar one for Ohio State fans at the start of the season, and computed that Ohio State was twice as likely to lose 3+ games, as they were to go undefeated.  That was not too popular with the OSU fanbase, but I think I will be somewhat vindicated by the end of the season.

Anyway, even though there are weaknesses, I think the analysis' results are useful for looking at how a season "should" play out.  More important than the spreads and exact per-game probabilities going in, are the ways in which the probabilities of individual games combine into the probability for this or that overall record.

Without further ado, here's the result for the remainder of Michigan's season:

By Sagarin-predictor pointspreads, Michigan is an underdog of more than a touchdown in every remaining game:

+8 at Purdue,
+9 against Northwestern,
+15 against Michigan State,
+16 at Minnesota,
+21 at Ohio State, and
+34 at Penn State

Using these numbers, the composite odds for Michigan's final six games of the regular season are:

6 wins =  0.0% (8-4)
5 wins =  0.0% (7-5)
4 wins =  0.4% (6-6)
3 wins =  3.9% (5-7)
2 wins = 18.7% (4-8)
1 wins = 41.8% (3-9)
0 wins = 35.2% (2-10)

By that calculation, Michigan is more than 75% likely to finish either 3-9 (one win in their remaining games) or 2-10 (no wins), with 3-9 being a bit more likely than 2-10.

They're a 22-to-1 longshot to exceed 4-8, and about a 250-to-1 longshot to be marginally bowl eligible at 6-6.

Individual results [excised for space and horror considerations; it's just the numbers in detail]

The most likely single outcomes are:

(1) Lose all remaining games (35%)
(2) Beat only Purdue (14%)
(3) Beat only Northwestern (12%)
(4) Beat only Michigan State (6%)
(5) Beat only Minnesota (6%)
(6) Beat Purdue and Northwestern (5%)

-------------

Anyone still alive after all that? If so: computer ranking are pretty crappy measures of team strength even at the end of the year, and they can be wildly inaccurate with only six games of data. The thing that jumps out at me is the Penn State spread, which is a full ten points higher than the Vegas line. FWIW, the Vegas line is always a more accurate predictor than computer rankings. This is considerably more grim than the facts on the ground. Probably.

But… yeah. I'm building a bomb shelter.

Comments

Chrisgocomment

October 15th, 2008 at 12:44 PM ^

Yowzaa....can we really say it's surprising after what we've seen?  I'm hoping and praying that we beat Purdue, NW and Minnesota at the very least.....and THAT proposition seems unlikely without those crazy new-fangled computers.

ATX Wolverine

October 15th, 2008 at 12:47 PM ^

Out of curiosity, how is "Chris" confused with "Harold"?  That's just embarrassing.  I'm pretty sure if I'd been randomly calling some dude "Harold" for the last 4 years and they finally got fed up enough to tell me their name is actually "Chris", I would want to avoid Harold/Chris for at least 6 months to a year afterwards.

UofM Snowboarder

October 15th, 2008 at 12:48 PM ^

I always find that statistics are hard to swallow and impossible to digest. The only one I can ever remember is that if all the people who go to sleep in church were laid end to end, they would be a lot more comfortable. - Mrs. Robert Taft. 

jamiemac

October 15th, 2008 at 12:49 PM ^

.....for at least 2 of those games are out:

MSU -3.5 at UM.......first time I saw this line it was UM -1, then last week MSU -1....now Sparty is laying a FG and a hook.

OSU -20.5 at home vs UM......line has steadily grown from 14.5 since the start of the season.....not sure if OSU can even score that many points. Of course, UM could get shutout.

As for the other games.....I would have no trouble taking UM and the points if the actual lines are as high as those computer numbers point.

We are in games, folks, as long as we can avoid turnovers. I realize thats a lot to ask of this team, but eventually the madness has to stop.

Right!?!?!

Yinka Double Dare

October 15th, 2008 at 1:16 PM ^

Exactly.  We've only been truly blown out once -- by Illinois, mainly in the second half as we were right there at the half.  Other than the million unforced turnovers we played better than Notre Dame, and the other games were all close one way or the other. 

Penn State might be able to reproduce what Illinois did to our defense, but I don't think anyone else on the schedule can, so yeah, if we avoid fumbling a million times they should be in it in the other games.  And if we're in it, we'll have a chance to pull some of these games out.  KC Lopata absolutely must get his confidence back though if we are to have a chance -- we just can't afford to give away points even more so than usual given the wild inconsistency of the offense. 

Anonymous Coward (not verified)

October 15th, 2008 at 2:55 PM ^

Direct quote from the ND blog Dr. Saturday posted to:

"I don't think you can say, 'Take away the turnovers and Notre Dame would've won.' This is still an inexperienced offense and turnovers come with an inexperienced offense."

Direct quote from the link you posted to:

"The Blue-Gray Sky is right that, under any circumstances, a -5 turnover margin is terrible performance, and you can"t "take away" the Irish miscues."

While ND fans wrote Dr. Saturday for inclusion on his weekly "Life on the Margins" feature, LotM demonstrates how costly major mistakes are to teams.  It doesn't show that a "better" team lost, it shows that a team made a bunch of mistakes that ended up providing the margin of defeat. 

You have to go a step further to argue like UM fans in the wake the the ND loss and claim the bad plays are random, bad luck or bad calls.  (I'll cut Brian some slack in that he had stated his belief that turnovers are largely random many times prior to that game.)  Again from the ND blog: "I've never subscribed to the theory that turnovers are largely random, the downward turn of Fortuna's Wheel. Turnovers generally result from bad plays by the offense, good plays by the defense, or offensive miscues that create a window of opportunity for a defensive player to force a turnover."

Of course ND would have won if you take away 5 turnovers (not so clear Michigan would have beaten ND if you took away the net -3, although it would have been a pretty even game).  But we made them, UNC did not, and we lost.  Unlike UM, which has still never lost to a better team, just beaten itself.

 Sorry for the long reply, I'm just really bored at work.

-irishdevil

Anonymous Coward (not verified)

October 15th, 2008 at 12:52 PM ^

Well, before we all collectively jump off a cliff, keep in mind that Vegas/Sagarin lines will change over the course of a season.  If Michigan should somehow beat Penn St. (or at least make it a very close game), then the ensuing lines should tighten up.

In other words, based on how our season has gone to date, this is a good reflection of how it will go in the future if nothing else of significance changes from here on out.  The odds of that occuring are probably lower than the odds of Michigan finishing 2-10.

Blue Durham

October 15th, 2008 at 12:53 PM ^

is based on NFL spreads.  Isn't it a rare event for any NFL team to be favored by 21 points?  I don't think I have ever heard of a point spread over 30 points, so this is based on a very small sample this far out.  It may be extrapolated out from the smaller spreads, though, but still.

Also, it has been said that since they are "pros" in the NFL, on any given Sunday...

But until only recently, I used to think that on any given Saturday, Alabama would beat the living crap out of EMU.  And I don't think that is the case anymore.

I don't quite have as pessimistic outlook as these numbers indicate.  But close. 

papabear16

October 15th, 2008 at 1:12 PM ^

Caveman's right.  That'w what made me laugh.  The rest of the post was obscured by tears.

The application of pro numbers to the college game is a poor fit, mathematically.  That said, I can't argue that the ultimate results seem that unrealistic.  I think we get a few breaks and win one or two more, but sheesh.

MMB 82

October 15th, 2008 at 1:15 PM ^

so that any wins will be a pleasant surprise. And I am 100% behind RR, so F you naysayers.

 

That "chemical burn" scene from Fight Club was strangely calming..... 

imafreak1

October 15th, 2008 at 1:33 PM ^

The third quarter. Every game, you can blame the third quarter on RR.

Maybe it is easier to start with what is not RR's fault right now. The offensive line, fumbles, unexceptional QBing, hideous safety play, inconsistency from young players.

Things that will bother me are lack of effort, personal fouls from frustration, offenses using the max protect and throw to wide open guy to repeatedly score TDs, and only winning one more game.

 Right now, I see a Michigan team that plays hard and tries to over come adversity. They've been in every game (even Illinois was close into the 4th) and that is important. I also like how RR has found and trusted the good young talent. He takes chances when appropriate and does not appear to lose faith. When Threet hits open receivers the offense hums. All good things.

WolvinLA

October 15th, 2008 at 1:59 PM ^

A good amount will be blamed on him - he's the coach.  Just like a lot of people blamed Charlie the Enormous for ND last year, and although it's not completely analogous, he had some of the same problems to deal with.  If we had a 10 win season he would be a lot of the credit, so he will likely take a lot of the blame. 

That said, I don't think anyone else would have done much better.  I don't think that makes him a bad coach, or that he should, by any means, be fired.  This is one season that will have to be sacrificed so that we can have WVU type success, but in the Big Ten.  Totally worth it, in my book.

imafreak1

October 15th, 2008 at 1:20 PM ^

Since most numbers bases analysis is interesting, this is interesting. However, there is huge volatility in this particular analysis especially as applied to Michigan.

1. Timing. Doing this analysis now, after M's worst loss, predicts the basement. Lines change wildly over the course of a season and are still volatile week to week. Just look at Minnesota. They get a huge bump from beating Illinois but is anyone comfortable that game is an accurate gage of the relative strength of those 2 teams?

2. If jamiemac is correct, the Vegas line for the MSU game (at least) is crazy different from the computer line.

3. This Michigan team is considerably more volatile than your average team. They also have more room (and expectation to believe they will) to improve than your average team.

Having said that, I wouldn't be surprised by 4-8 but think losing all remaining games is very unlikely. Unless Threet gets hurt 3-9 is pretty tough and will make me sad. I also think that the games Michigan wins and loses will continue to be an odd jumble of inconsistency that eludes prediction. This team, at its worst, is not much better than ND from last season but they have a much higher ceiling. These stats treat them the same.

Anonymous Coward (not verified)

October 15th, 2008 at 1:24 PM ^

This is very interesting, but I'm not following the methodolgy. Can you explain in a little more detail?

Also, everyone should keep in mind that the Vegas line is not designed to be an indicator of score or outcome.

TrueBlueMan

October 15th, 2008 at 1:34 PM ^

This is very interesting, but I'm not following the methodolgy. I use PREDICTOR all the time but what are then other numbers you are using? Can you explain in a little more detail?

Also, everyone should keep in mind that the Vegas line is not designed to be an indicator of score or outcome.

Anonymous Coward (not verified)

October 15th, 2008 at 1:35 PM ^

And you will never be disappointed I guess...but I'm more of a glass is 1/20th full kinda guy

joeyb

October 15th, 2008 at 1:37 PM ^

I would like to see our odds after the PSU and MSU games. If we keep it close vs PSU, those lines will drop quickly, especially if our offense scores a few times. Plus, a lot of people have been talking about us beating MSU because all they have is Ringer and if we stop him like we stopped PJ Hill, their offense stops functioning. If those two games turn out this way, the lines shift and we have a much better odds of beating Purdue, Minnesota, and NW.

Imagine if they had done this analysis on USC after the OSU defeat. They would have had a near 100% chance of being 12-0 or possibly 13-0. We all know how that turned out. This is college football and it is largely based on emotions. Statistics cannot account for that. I still think there is a good chance we hit 6-6.

Anonymous Coward (not verified)

October 16th, 2008 at 2:54 PM ^

"A lot of people" also said Michigan would beat Toledo comfortably.  Michigan was a 16-point favorite.

There is not a "good" chance of Michigan making it to 6-6.  There is a small but nonzero chance.

Anonymous Coward (not verified)

October 15th, 2008 at 1:37 PM ^

wins this week, all those numbers go flying out the window. Gone. Pfft. Seeya.

Which is why, back to your original premise, Brian, extrapolating results in college football even one week forward is foolhardy.

Anonymous Coward (not verified)

October 15th, 2008 at 1:39 PM ^

sounds like one or both of these games could have below 100K in attendance...... unfortunately another tradition may fall.

UMFootballCrazy

October 15th, 2008 at 1:48 PM ^

Even if these numbers are flawed, they paint a picture of the reality of things as they are.  I have no hopes left.  As I have watched the season unfold, the grim reality has settled in that our beloved program does not reside in such lofty heights so as to be immune from the problems that lesser programs have in making a major coaching/system change.  We too can go through a rebuilding period.  Its been a good run and good things are coming, but for now all we have to do grin and bear it and take comfort in the fact that Coach Rodrigues has done this before.  He needs more than talent, he needs his kind of talent.  His system takes time to learn and perfect.  I expect that by the time my eldest is applying for pre-med at UM in two years things should be sorted out and we will be the terror of the Big Ten.  Till then...its all pain and suffering...think expatiation...purgatory...a cleansing for the glory to come...

Maybe on the bright side some of these haters will give up their death grip on their season's tickets and it will be possible to buy a family block to enjoy and pass on to my kids and grand kids...

BlueNote

October 15th, 2008 at 1:48 PM ^

downst - gotta love the obscure Ace Ventura Pet Detective reference

I like your optimism.  You must have hope or else you'll go crazy.

"Sanity may be madness but the maddest of all is to see life as it is and not as it should be."

WolvinLA

October 15th, 2008 at 2:18 PM ^

No, Jim Carey's character asked what the chances are that they would be together.  He asked, "one in a thousand?" and she said "More like one in a million." 

He then said "So you're saying there's a chance!"

Later in the movie, when he found out she was married he said, "What was all that one in a million talk?"

Anonymous Coward (not verified)

October 15th, 2008 at 1:49 PM ^

i see tickets are available from the box office.  guess it will not be a sell out.  no surprise.

NW game may not even break 100K.  that would be another tradition gone. 

 

Ace

October 15th, 2008 at 1:54 PM ^

...my midterms now seem infinitely less painful. However, my reasoning behind dropping $60 on a Penn State ticket seems even more flawed than it was five minutes ago.

chitownblue (not verified)

October 15th, 2008 at 2:10 PM ^

Remember - the intention of Vegas Odds isn't to be predictive - it's to establish a spread on which people will gamble. So, if the perception held by teh vast majority of gamblers, who likely haven't watched that many games, is that Michigan is completely horrendous, they will need more points to bet on UM. The spread is a number based on perception - not an analysis of the teams.

helloheisman.com

October 15th, 2008 at 2:55 PM ^

But Vegas spreads are kinda like the stock market...there are a lot of people who watch closely and do thorough research along with those that don't.  If you believe in efficient markets, you should believe that Vegas lines are a good indicator of relative chances as well.