Has UM Just Been Lucky?

Submitted by Kilgore Trout on July 2nd, 2009 at 2:16 PM

Over the last few years, I have begun thinking more and more that UM football was getting kind of “lucky” and squeaking by and on the verge of things going south if a few breaks went the other way. I decided to look back at UM’s record since 1997 (as far back as I could easily get to on mgoblue.com and a reasonable statement of the “modern era”) and see what UM’s record was in close games. I defined close games as being decided by seven points or less. My reasoning for that distinction being that literally one play could have made the difference in the game. I realize that is somewhat simplistic, but hey, this is probably a somewhat simplistic exercise.

Here are the records. Overall and “close games” in parenthesis.

1997 – 12-0 (4-0)
1998 – 10-3 (3-0)
1999 – 10-2 (7-2)
2000 – 9-3 (3-3)
2001 – 8-4 (2-3)
2002 – 10-3 (5-2)
2003 – 10-3 (2-2)
2004 – 9-3 (3-1)
2005 – 7-5 (3-5)
2006 – 11-2 (1-1)
2007 – 9-4 (4-1)

Total - 105 - 32 (37 - 20)

Looking back, UM won 76.6% of games overall and 64.9% of games decided by a touchdown or less in the last 11 years of the Carr regime. The question then becomes, what is a reasonable expectation for a good and / or elite team for a record in close games? I did a little research and picked 5 “elite” teams and 5 blah / so so / meh teams and looked at how they did. Note: For the elite teams I took their record starting with the second season of their current coach. For the blah teams, I just took the last five years.

USC – 82-9 (11-9)
OSU – 76-14 (18-8)
Florida – 35-6 (7-4)
LSU – 31-9 (9-5)
Georgia – 74-18 (26-12)

Arizona State – 38-24 (10-5)
Oregon – 41-21 (12-7)
Oklahoma State – 34-28 (6-9)
Michigan State – 30-31 (5-16)
Pittsburgh – 33-27 (12-11)

What does this tell us? Umm… I’m not really sure, other than it took me way too long to look all of this up. Some things to note.

- USC hasn’t lost a game by more than a touchdown since Carroll’s first year. Wow.
- USC is pretty average if you get them in a tight game.
- Arizona State doesn’t play many close games. Must be kind of boring to watch.
- Not evident in these stats, but OSU was an amazing 11-1 in 12 games decided by a touchdown or less in ’02 and ’03 (big winning years). Krenzel was probably pretty underrated when it came to getting the job done.
- State was atrocious in close games over the last five years. If they break even in them, John L is probably still there.
- Michigan played more close games than anyone else in this data. A whopping 41.6% of UM’s games were decided by a touchdown or less. This probably doesn’t shock anyone that UM got into too many close games over Carr’s tenure. Only Georgia with 41.3% close games was in the same ball park. The funny thing is, over the last while, I would have picked Georgia out as a UM clone.
- As to my original hypothesis, with the exception of USC, teams with the highest winning percentages overall also had the highest winning percentage in close games. This tells me that Michigan “squeaking by” in those close games wasn’t really luck, it was because they were good. Or they were good because they were lucky. Or when it comes down to it, good and lucky are kind of the same thing. To me, that’s good to see.
- If UM would have won its customary 65% of close games last year, they would have been 5-7, which to me would have been easier to stomach. It wasn’t that far off, really. Beat Toledo and pull out one of Utah, NW, or Purdue and there they are.
- USC, Ohio State, Florida, and Georgia all made significant turnarounds in their respective coaches second years. Granted, none of their first years were as bad as Rodriguez’s, but it’s a good sign.

So there’s all of that, for what it’s worth.



July 2nd, 2009 at 2:28 PM ^

The thing that jumps out at me is the sheer number (57) of close games we had. We had far more than any of the other elite programs mentioned. I think that helps to explain our excellent record in those games. A lot of our close games came against teams that were less-talented than us, and we ended up beating them (as expected) but not before they put up a fight. A persistent criticism of Carr's teams was that they played down to the level of their competition too often. (1999 seems a prime case in point: did that team really face nine teams of comparable talent?) Other elite teams probably also would have beaten a lot of these teams, but perhaps by larger margins.

Kilgore Trout

July 2nd, 2009 at 2:32 PM ^

I went into that thinking I'd show that UM was unusually good in close games, but the biggest take home was that they play way too many of them. And very frequently it's against teams they shouldn't be that close to. App State and Toledo were bound to happen if you keep playing with fire (Ball State, SDSU for example).


July 2nd, 2009 at 2:59 PM ^

Since you're looking at about a dozen seasons equaling almost 150 games, the sample size is sufficiently large that you can observe real trends. You're unlikely to see the effects of a random event like luck in such a large sample. Consequently, the results may demonstrate real trends caused by real factors. The trick is analyzing the data and explaining what it means.

Here are some of my ideas on teams that interest me.


Finding: Played a large number of close games and had a good record in them.

Conclusion: Michigan played many close games against inferior opponents. This explains the amount of games and record. One might assume this was because of Carr's tendency to turtle.

Ohio State:

The Buckeyes demonstrate regression to the mean. Over one stretch they were 11-1--caused both by lots of luck and philosophy. The rest of the time they were 7-7.


The Trojans are particularly interesting because they have so few close games but counter-intuitively are barely over .500 in those games. The explanation is probably, USC rarely plays close games against inferior opponents. They blow them out. When they do run into an equal foe they are about .500 which contrary to internet geniuses is about where you should end up against equals.


July 3rd, 2009 at 1:25 AM ^

With respect to the USC games, I must disagree with you. While I do agree that USC has probably played about .500 against equal opponents during the Pete Carroll era, I feel that USC has lost a significant number of games to clearly inferior opponents. Of it's 9 close losses, 6 came against clearly inferior competition (the exceptions being 2003 Cal, 2005 Texas, and 2007 Oregon). A number of these games, such as Oregon State in 2006 and 2008, were decided in the last minutes. I feel that a more likely explanation is that USC does not elevate their level of play at the end of a game as much as other periodically dominant teams (such as Miami before the rise of USC). Therefore, they are less likely to win close games as other elite teams, such as Georgia, LSU, and Michigan.

I feel that this is indicative of Carroll's game time coaching. In College Football, Carroll has been able to produce a dominant program at USC by consistently out-recruiting the rest of the country and developing his talent well. However, when Carroll was in the NFL, he was fired from both the Jets and Patriots after short stints as a head coach (1 and 3 seasons respectively). While I understand that there is a huge difference between college and pro football (ask Steve Spurrier) I think that his lack of success in the NFL can be attributed partially to his game time coaching, considering his track record with USC.

I do understand that my argument is based on evidence that can lead to multiple conclusions. However, I feel that it has enough merit to justify consideration.

Big Boutros

July 2nd, 2009 at 3:39 PM ^

I like this post and appreciate the number crunching that the OP did. I, personally, prefer to consider Lloyd's teams clutch rather than lucky. I think a good microcosm of Lloyd Carr's Michigan was the third game of 2004 against San Diego State.

Like so many other games in Lloyd's tenure, Michigan was actually losing this one at halftime, 21-17, and held the Aztecs scoreless in the second half to squeak by, 24-21. I would agree that Michigan was probably lucky to win that game in that the final score was not a reliable measurement of the football acumen of the winning team.

It was an especially fortunate win considering our two most important offensive skill positions were manned by true freshmen. That team was unique--it was Henne and Hart's freshman year--but the development of that game was fairly typical for Lloyd's regime.

However, I don't think Michigan's victory in that game--or any other close game--is entirely attributable to luck. Was it unlucky that Shawn Crable was called for roughing in the 2006 Ohio State game? Perhaps, but flipping a coin on whether the referee would throw a flag is a lot less reliable than simply not hitting Smith out of bounds (although, for the record, I don't believe Crable's hit was illegal). Were we lucky to win the 2004 triple overtime game against State? Maybe, but the team put itself in a position to win.

I'm not one of those jerks who doesn't believe in luck in any capacity, but I sincerely believe that Michigan's success in close games under Lloyd Carr is more directly attributable to second-half execution.


July 2nd, 2009 at 3:55 PM ^

Others have alluded to this, but I wonder how many of those games were the result of UM sitting on leads. I can think of many bonafide nail-biters and great comebacks, but I also can think of several games where Lloyd respected the other coach to the point of almost losing games Michigan had easily in hand (Ball St. comes to mind). Love the number crunching--great post.


July 2nd, 2009 at 4:19 PM ^

I am not sure that it is fair to:

"For the elite teams I took their record starting with the second season of their current coach"

...It might be fair if you were simply comparing coaches, but not programs I would say.

Also, how many big OOC games does LSU and UGA schedule? How many outside of the south? We use to be decent at going out west or inviting in a respectable 2nd OOC opponent.


July 2nd, 2009 at 4:25 PM ^

really points out the "killer instinct that a coach/program has. I agree 100% that Lloyd was very concerned about not embarassing the opposition almost to a fault. It made games that should have been easy pretty tough, and when you take momentum into account it made finishing even harder. Don't get me wrong, I was very happy to have Lloyd as a coach. He was a great one, and I respect that he stuck to his priciples no matter what. That being said, I can't wait to see RR run up the score on some teams this year. When this machine starts chugging along, it will be effective and deadly. I think he will make a point in the first four games to send a message to the rest of the college football world. They won't really be believers until we go to play MSU. That game will be a big one, and it will tell us about what kind of team we have this year. Go Blue!!!


July 2nd, 2009 at 5:04 PM ^

I appreciate all the work you put in...One thing I'm hoping will happen is that Rodriguez will help us all have healthier hearts than we did under Carr. One of the things I like about Rodriguez's tenure at WVU is that if WVU could find some weakness in your defense they would exploit it again and again until you were dead, dead, dead...


July 2nd, 2009 at 6:21 PM ^

You can probably thank the Mike Deboard mentality for both why there were so many close and why we won so many of them. We were in a lot of close games with teams that really weren't in our league. But with a good defense and running game, as well as a usually more talented team, our close games may not have been as close as they appeared.


July 2nd, 2009 at 8:38 PM ^

I appreciate the amount of data and math involved in this post, but no way can you just look at a bunch of scores and decide "it was a close game."

The combined memory of this board could probably (with help) remember many details of the last 12 Michigan seasons, certainly the major momentum story arcs etc. But I don't think you can/did.

Also, what about all these other teams you have listed? Tell me all about how that 2003 ASU vs. bum-fuck U game went down.

Finally, as has been stated by others (including me in another post about defenses) Lloyd-ball allowed many teams to get within 7 points within the final minute or two.

I really do admire the work, and there is a lot of interesting info to ponder; I just don't think final scores tell a very accurate story about close games, therefore the data is compromised.


July 3rd, 2009 at 2:14 AM ^

This Article from Rivals/Yahoo spells it out pretty concisely....

It gives you the worst Michigan Defenses in terms of Yds and Pts in school history... other than last year... you find that most of them have come since 1997:

Ten worst scoring defenses
Year Ppg Record
2008 28.9 3-9
1962 23.8 2-7
1958 23.4 2-6-1
2004 23.2 9-3
1994 22.3 8-4
2007 21.4 9-4
1999 20.6 10-2
2002 20.4 10-3
2005 20.3 7-5
2000 19.1 9-3

Ten worst total defenses
Year Ypg Record
2000 389.9 10-3
2008 366.9 3-9
1982 366.8 8-4
1994 362.0 8-4
2002 347.7 10-3
2005 345.2 7-5
1984 343.2 6-6
2004 336.2 9-3
2007 335.7 9-4
1999 333.0 10-2

Part of that may have something to do with the proliferation of spread offenses in to the Big Ten... but IMO


July 3rd, 2009 at 10:36 AM ^

From the title to the amount of work you put in, this was a very fun read.

Thanks for the entertainment and thanks for the time you put in to writing it.


July 3rd, 2009 at 10:52 AM ^

I think Carr's grossly predictable playcalling is what kept so many games close. When the other team knows what's coming, it has a tendency to negate a portion of your talent advantage. Also, as a previous poster stated, the conservative nature of his playcalling did tend to keep games closer, too.

Another factor was Carr's use of the "prevent defense." This often backfired and allowed the opposing team to get back into the game. I have never understood why a coach would abandon what got his team the lead in the first place during the last part of the game and expect it to be an improvement. I would rather see the team play straight up and continue to be aggressive for the duration of the game.

Whatever the case, I definitely agree with the poster who can't wait to see RR hang some serious numbers on a few teams.

In the meantime, how 'bout that 1981 UM/Illinois game:



July 3rd, 2009 at 5:02 PM ^

It's a topic some pals and I have discussed ad nauseum. I believe that the 7-5 year was something that could have easily happened five times over the last ten Carr years, I really do. For all the lamenting about the bad breaks our team gets -- and they have been legion -- we forget Wisconsin muffing punts at Camp Randall, or Braylon going apeshit against State, or Navarre going all blind masseur swordsman against Minnesota in the 4th.