We're the underdog in our road opener? Bet on Michigan

Submitted by Hannibal. on

It's a good thing that Michigan State is favored this weekend.  How come?  It is because of a strange statistical anomaly.  The underdog in Michigan's road openers for more than 20 years has an uncanny knack of beating the spread, and very often pulling the upset.  Here is a history of Michigan's road openers since 1986.  Notice how the favorite almost never covers the spread.  In most cases, it's obvious just by glancing at thes core whether the underdog beat the spread.  I have left a few games open, because I don't know how to find spreads for really old games, and some of these were close.  Edit -- I have updated this diary, thanks to the information in some of the comments below...

1986 – Michigan enters the game against Notre Dame as a heavy favorite and escapes with a 1-point victory.  Underdogs beat the spread (1-0) Edit:  we were actually a 7-point favorite

1987 – Michigan loses at Michigan State.  I can’t find any documentation, but I’ll bet we were favored in this one, so I’ll give this one to the underdog (2-0)

1988 – Michigan loses by two to Notre Dame.  I don’t know what the spread for this one was either, but two points sounds like it might have been right since both teams were ranked highly, so this one is an unknown.  Edit: we were a three-point dog. 

1989 – Michigan beats a crummy UCLA team by 1 point.  Once again, underdog beats (3-0)

1990 – Michigan loses to ND by four.  Both teams were top 5 ranked, but since the game was in South Bend, ND was probably a small favorite.  I'll say that they covered (3-1)  Edit: we were actually a five-point favorite.  Seems weird, but once again, the 'dog beats the spread. 

1991 – Michigan comes out flat against 4 TD underdog Boston College but wins by 22.  I’m pretty sure that this was an underdog victory vs spread (4-1)

1992 – Michigan is a slight underdog against Notre Dame and they beat the spread with a tie (5-1)

1993 – Michigan is favored to beat Michigan State and loses (6-1)

1994 – Michigan is the underdog to Notre Dame and wins (7-1)

1995 – Favored Michigan destroys Illinois.  Favorite actually covers.  (7-2)

1996 – Underdog Michigan upsets Colorado (8-2)

1997 – Michigan massacres Indiana by 37.  Favorite covers (7-3)

1998 – Underdog Notre Dame destroys Michigan in South Bend (8-3)

1999 – Michigan beats Syracuse by 5.  I don’t know what the spread for this one was, since we lost to them badly the year before and it was in the Carrier Dome.  This one is an unknown.  Edit:  We were favored by 6, so underdog beats spread.

2000 – Underdog UCLA upsets Michigan, starting a long, horrible road opener losing streak for Carr (9-3)

2001 – Michigan loses by 5, but was a touchdown ‘dog, if I recall (10-3)  Edit:  turns out we were only underdogs by 3, so Washington did, in fact, cover it.

2002 – Underdog Notre Dame beats Michigan (11-3)

2003 – Underdog Oregon beats Michigan (12-3)

2004 – Underdog Notre Dame beats Michigan (13-3)

2005 – Underdog Wisconsin beats Michigan (14-3)

2006 – Underdog Michigan beats Notre Dame. (15-3)

2007 – Heavily favored Michigan escapes Evanston with a 12-point victory over Nortwestern.  I don’t know exactly what the spread was, but the ‘Cats were coming off of a loss to Duke and then a 58-7 loss to Ohio State, so I’m sure the spread was more than 12.  (16-3)

2008 – 2-point underdog Notre Dame beats Michigan.  (17-3)

That’s 17 games where the underdog covers, and 3 games where the favorite covers.  That’s an 85% success rate for the underdog!  Since 1998, the underdog is undefeated against the spread (except for maybe ’99?)  I may have gotten one or two of the older games wrong and I have a couple of unknowns.  Even if a few games go to the favorites, it’s still an incredible anomaly.  Especially when you consider how refined betting spreads usually are and how tough they are to outpredict.  This phenomenon can’t just be summed up as “Michigan underachieves in first road game”, because we have covered the spread as the underdog at least every time since 1992, pulling upsets against Notre Dame twice and Colorado in 1996. 

Edit:  Actually the underdog is 20-3 over this time span.

Last I saw, we are a 3-point underdog to Sparty.  If you’re a betting man, bet on Michigan.

Michigan
betting
Michigan State
road opener
point spreads

Comments

tsabrak

October 2nd, 2009 at 10:00 AM ^

I think this is a neat find. Nice work. Also, I think that it's actually 18-3 based on your review. You shorted the underdog a win between the years 1995 and 1997 ('95 is 7-2, '96 is 8-2, '97 is back to 7-3)

Joined: 08/17/2009

MGoPoints: 130

Dan Man

October 2nd, 2009 at 3:21 PM ^

FAIL. Didn't we open as the favorite in this game? Can you account for line moves in your data? In any event, this odd statistic trend has no basis for predictive theory other than "this is the trend." For example, compare to "sandwich games" where a heavy favorite will frequently not cover if the following week they play a much more important game. This trend makes sense because the heavy favorite is overlooking their current opponent. Is there a theory why the underdog in Michigan's road opener likely wins, REGARDLESS of whether it's the home or road team? Of course not. Anyway, if you want to bet, go ahead, but I wouldn't be on the Mich/MSU game either way, and I CERTAINLY wouldn't bet on it based on your theory.

Joined: 06/30/2008

MGoPoints: 1097

ikestoys

October 2nd, 2009 at 10:46 PM ^

You get to the right conclusion but with the wrong argument. This is a classic example of correlation shown with no causation. Since I doubt that anyone will be able to come up with some solid rationale for some sort of causation, all this data presents is an example of variance.

Joined: 09/26/2009

MGoPoints: 1723

Beegs

October 2nd, 2009 at 10:12 AM ^

Great digging. This is one of those interesting trends similar to the 37-4 record that the winner of the UM-MSU game has when it outrushes the opponent over the last 41 years. However...I suspect that the RR era will see us sort of throwing out all that stuff from a "valid trends" perspective. Things just feel soooo different now with the new schemes and different type of coaching styles and different types of players. I think these historical trends - while interesting - will be relatively meaningless going forward.

Joined: 08/17/2009

MGoPoints: 103

Slinginsam

October 2nd, 2009 at 11:14 AM ^

I bleed Maize and Blue, but the spread opened at - 1 1/2. It has moved to -3. Most of the time, the direction that the spread moves is right. As an example, the ND-Michigan spread opened at ND -5. It closed at ND -2 1/2. We know what happened there. Let's hope the money is wrong. GO BLUE!

Joined: 06/17/2009

MGoPoints: 16

jamiemac

October 2nd, 2009 at 11:43 AM ^

Where do you base the assertion that most of the time, the direction the spread moves is right? I always fade the public in those times, and usually win. Also: at no point did I see them ND/UM line at ND -5. I tracked this rather religiously over the summer and into September. It was -3 all summer long, but did dip to -2 before creeping slowly, but surely all the way back up to -3.5, where it stayed most of game week. It closed at -3. Not sure there was a lot of one way action on this game at all.

Joined: 06/30/2008

MGoPoints: 8446

jamiemac

October 2nd, 2009 at 11:24 AM ^

Official lines for some of the above games where they O.P. said he didnt have them. Does not change much, but figured y'all would like to know. Lines from covers.com 1986: UM -7 at ND. 24-23, UM 1987: UM -6.5 at MSU. 17-11 MSU 1989: UM +3 at ND. 19-17 ND 1990: UM -5 at ND. 28-24 ND....so this year, the dog did cover in the road opener. ND replaces a new QB. UM virtually intact from a year before. Thats the only way I can explain why UM was favored here. 1999: UM -6 at S'Cuse. UM 18-13 2001: UM +3 at UW. UW 23-18.....UM missed a 2-pt attempt in the final minute, or else this is at least a push. So, the Dog gets an extra cover than the O.P. credited it for due to the 1990 ND game, but losses a cover it has been credited for due to the 2001 UW game. Does not change the record the O.P. passed on, however. Remember: UM is 9-21 ATS in their first road game of the season, dating back to 1979. However, they are 6-2 ATS as a dog in thier first road game. Think about that: Generally speaking, UM is in a role its been successful in only 30 percent of the time the last three decades. But, in the fine tuned, specific role they're in, it's a 75-percent success clip. Interesting numbers indeed. Bookies will be rich tomorrow after this game if UM wins.

Joined: 06/30/2008

MGoPoints: 8446

Hannibal.

October 2nd, 2009 at 1:18 PM ^

Thanks for the updates and corrections. By my final count, that makes the tally 20-3 in favor of the underdogs over the 23 year period. The only years in which the favorite covered the spread were 1995, 1997, and 2001. In other words, the underdog covers 87% of the time.

Joined: 09/09/2008

MGoPoints: 9051

AMazinBlue

October 2nd, 2009 at 12:40 PM ^

any spread over 12 points, take underdog. Michigan rarely covered big spreads under Carr. His get a lead and protect philosophy made betting on UM difficult. I don't see that being the issue with RichRod. But tomorrow will be tight due to weather. I hope we drop 45 on them, but being on the road, 1st road game, bad weather and the QBs first road game, it will be tight. The anticipation is killing me. Last year destroyed the "we should win this one" mentality. So every game is exciting because of the "who's really better" factor. This season has a lot more electricity than any I can remember.

Joined: 10/12/2008

MGoPoints: 7197

Slinginsam

October 2nd, 2009 at 1:02 PM ^

http://www.vegasinsider.com/college-football/odds/las-vegas/ The left side column shows where the spread opened. They begin the Sunday night before Saturday's games. The ND game opened here at 5. They showed it closing at 2 1/2 at game time. Like I said, if you watch the spread movements in both CFB and NFL, MOST of the time, the direction in which the spread moves is right. Hope I'm wrong here, but if I were to wager, I'd take MSU. I hate to say it, but I think this will be the week where the dam bursts on our D. Got this sick feeling in my stomach. Believe the bettors think so, too.

Joined: 06/17/2009

MGoPoints: 16

jamiemac

October 2nd, 2009 at 2:07 PM ^

Nice link.....but: 1.) It shows me this week's lines, not Week 2. 2.) And, if you are refering to the LVSC (Las Vegas Sports Consultants) number in the far left column, that needs to be clarified. That is not an actual point spread. LVSC releases what they think the line should be. Its just a recommendation. You cant place a wager at LVSC, for example. So, that column does not represent the actual opening line. But, it does represent what oddsmakers were intially recommended by LVSC in terms of what the line should be. Sometimes they match up, but other times the oddsmakers will tweak it to fit their own analysis as well. For example: PSU at Illini. The LVSC column said PSU -4. This was not the opening line. The line did not open at 4 and then move to -7 due to overwhelming public money on PSU. Within the hour of LVSC posting their recs, off shores like Caribsports, Pinnacle Sports released their official opening lines as PSU -6, -6.5. I looked these up at both sites Sunday afternoon and that's where they opened the betting. A lot of place opened this line up higher than the LVSC rec. They basically looked at the rec of -4 and said, WTF, are you kidding me, have you seen the Illini, that's NOT GOOD ENOUGH!! So, I dont think the ND line opened up at -5. It was probably the LVSC you are refrring to. I track summer lines pretty religously. It went from -3.5 to -2.5 during preseason betting. The first actual line to bet on that I saw for this one during game week was -3.5. Whatever, Michigan won outright. Fuck the Irish. As for the cover always going to the way the money goes....hey, I can just go with my personal experience of fading the public bandwagon. But, I dont fade all public bandwagons and those dont enter my conscious, as a result, so we both could be right there. In Week 3 alone, I took the Vols, Noles and MSU on simple public fades and covered each easily. But, of course, I am ignoring public bandwagons that won because I had no action on it. My research is very self centered there. Anyway, good stuff to talk about. We can only beat the Book by talking amongst ourselves.

Joined: 06/30/2008

MGoPoints: 8446

Firstbase

October 2nd, 2009 at 7:48 PM ^

I've been a Michigan fan since I listened to Bob Ufer while picking apples on our Casnovia orchard as far back as 1968 with my dad. (I can still picture the little brown transistor radio we used sitting on the corner of a crate.) Back then, I recall stacked Michigan teams beating opponents into submission by scores like 72 - 14 etc... It was fun! My point with this meandering post is just to say that MSU wasn't too much of a factor back then, and really isn't now. They can have a successful season here and there (and kudos to them when they do), but they'll never match the Michigan Mystique. MSU would have to pull off double-digit wins over Michigan for several years running before even thinking about claiming "bragging rights." And Dantonio? His little outbursts, pouting and diatribes are completely sophomoric. What a bitter little man he is! He is obviously "green" with Michigan penis envy. It would be wonderful to see a 2-touchdown victory tomorrow. It would take me back to those thrilling days of yesteryear. I can almost hear Ufer's "aooga" horn as I type!

Joined: 09/30/2009

MGoPoints: 9369

ikestoys

October 2nd, 2009 at 10:41 PM ^

"It's a good thing that Michigan State is favored this weekend. How come? It is because of a strange statistical anomaly. The underdog in Michigan's road openers for more than 20 years has an uncanny knack of beating the spread, and very often pulling the upset." This isn't a predictor at all of future performance. If you want to make an argument that the line is off because of x, y or z reasons that actually concern this game then great, but those prior games are independent results from the game tomorrow. The results you've tabulated are just variance. The betting market wouldn't purposely allow such a glaring hole to last for this long because the market doesn't want to go broke.

Joined: 09/26/2009

MGoPoints: 1723

jamiemac

October 3rd, 2009 at 10:11 AM ^

Truthfully, yhe most relevant thing here is how Michigan does ATS in their first road game, not how the favorite or underdog in the game does. As a program, they have struggled in their first road to meet Oddsmakers "expectations" since 1979. That makes this a Michigan tradition as old as stud WRs wearing the #1 jersey. I dont know about your causation or correlation, but as a life long observer of this program and as a guy who also follows the point spread game, I can tell you this a legit flaw in the program's armor. We'll see if Rodriguez turns the tables on it. Another ID of the program, however, is how well it does against the points when its an underdog. So you have dueling IDs going here. Typically Michigan is laying chalk, sometimes big time, in thier first road game. Given conflicting IDs, a second year coach and a different role than usual in their first road game, its hard to put a ton of stock in these numbers in this thread. But, I find them fascinating nevertheless an thanks to Hannibal for pulling them out.

Joined: 06/30/2008

MGoPoints: 8446