Profiling some successful rebuilding seasons

Submitted by taistreetsmyhero on November 26th, 2017 at 6:40 PM

This season had two possible trajectories: reload or rebuild. As Brian succinctly stated in his season preview, "This is a transition year between The Year and The Year, unless it isn't." And while many fans donned their maize-colored glasses in the preseason prediction threads, hoping for the lucky breaks to go our way en route to another 10-win season, the fancystats suggested that it was more likely to be a rebuilding year. Bill C. prognosticated before the season that "this program is probably a year away from ignition." 

"They'll have to get lucky on a couple freshmen and one right tackle, but teams have been luckier. Just not Michigan." -Brian Cook

For innumerable reasons (some within and many out of the team's control), alllllllll of which have been very clearly hashed out on this site, this was, in fact, a rebuilding season. The issue now becomes whether or not this was a "successful" rebuilding season. The MGoStaff posed themselves the question "are we on track?" The answers ranged from "mostly, yes" to "of course, yes." Ace compared the roster this year to the projected one for next season and concluded, "This year hasn’t been very fun. Next year will be."

While we can debate and analyze the quality of this season, in isolation, until we've beat the dead horse to a maize and blue pulp, the fact is that we won't know whether or not it did its job as a rebuilding one until after next season is over. What we can do, however, is partake in every Michigan fan's favorite pastime: make predictions about whether or not next year is finally "The Year", or, as Bill C. quipped, "Ignition."

But first, here is Seth describing his take on a baseline definition of The Year:

"The level I want to be at is Bo--i.e. go into every season with a real shot to win it all--and last year's team is a good example of that level. Football is too random to ask for more: Ohio State is the best team in the country this year by S&P+ and has two losses; Carr won a national championship with the Griese offense and lost to Illinois with senior Tom Brady. It's natural for your human brain to try to replace randomness with patterns that don't exist, and favor whichever pattern best serves your interests. If you haven't learned to keep those instincts in check this deep into adulthood, don't be surprised when your contributions to the marketplace of ideas are treated like they don't have any value."

The Year is a season in which the team makes a legitimate title run and is only a couple lucky of breaks away from ultimate glory.

A look back at rebuilding years that preceded "Ignition"

With all that in mind, I decided to profile some recent rebuilding seasons for other teams led by high-quality coaches that were followed by The Year and successfully "ignited" those programs. I chose the seasons based on a recent diary by MGoUser Eye of the Tiger that looked at win totals of elite coaches in the mythical Year 3 and Year 4 of their tenures.*

This analysis is descriptive, not predictive. It is designed to see how Michigan compares to 12 successful rebuilding teams across a small collection of metrics. I do discuss my interpretation of the results, but that is simply my fan perspective.

*Note: I did not include any season prior to 2007, when S&P rankings first started.

Quality of Wins vs. Losses

Higher is better on the left. Lower is better on the right.

While Michigan managed to more-or-less hit its expected win total of 8.5 games, one of the biggest knocks on the team is that they didn't beat any good opponents. This criticism holds true relative to this selection of successful rebuilding teams. The quality of Michigan's wins (as defined by the winning percentage of the teams Michigan beat this season, removing the game against Michigan to control for record) is only higher than those of the 2011 Florida State squad.

Looking closer at the 12 successful rebuilding teams, they averaged 0.33 wins against teams ranked between #20-25 in the final AP poll, and 0.58 wins against teams ranked between #10-20. Only 4/12 teams had no wins against Top 25 opponents.

In contrast, the quality of Michigan's losses is higher than any of the other 12 rebuilding teams' losses. All 4 teams that beat Michigan are ranked in the Top 25. The other 12 teams averaged 3.33 losses to teams in the Top 25 of the final AP poll, but also 2.00 losses to unranked opponents. Your take on this stat could be a glass half-full or half-empty one. It's always good to beat the teams you are supposed to beat, but it could be easier to fix an issue with losing to overmatched opponents than it is to get over the hump of losing to good teams.

Average Margin of Wins vs. Losses

Higher is better on the left. Lower is better on the right.

Another concern this season was that Michigan struggled to put away some clearly inferior squads early in the year. While the games against Cincinnati or Air Force never reached Hoke danger levels, they weren't the blowouts that they probably should have been given how the seasons of those two teams played out.

However, when we look at Michigan's average margin of victory compared to the other 12 rebuilding teams, it is greater than the median. On the other hand, Michigan's average margin in losses is worse than only that of 2011 Penn State (15.50).

The important confounding variable here is that Michigan beat worse opponents and lost to better ones than the other 12 rebuilding teams. This could potentially inflate the average margin of victory and deflate the average margin of loss in this comparson. Here is the chart of the 12 teams' average margin of victories against unranked teams and average of margin of losses against ranked teams:

Michigan still performs right at the median in terms of average margin of victory, but is still worse than 10/12 teams in terms of average margin of defeat. While the numbers are what they are, Michigan did look infinitely better against OSU than they did against PSU. Additionally, Michigan was in the game against Wisconsin until the injury to Peters completely deflated the team, which helped Wisconsin pull away. And Michigan was one possession away from tying beating OSU until the O'Korn armpunt, so an 11 point loss isn't qualitatively worse than 7 points. I don't know how the losses for the other 12 rebuilding teams compare, but I'm not too worried about this stat.

S&P Rankings*

*Michigan rankings are prior to OSU game

Not surprisingly, Michigan's defense compares favorably to the rebuilding teams, whereas the offense does not. I'll let the rankings speak for themselves, but the encouraging thing to me is that 4/12 rebuilding teams had worse passing attacks than ours. Given the youth at so many positions, the multitude of injuries, and the limited roster turnover, I think that Michigan stacks up very well to the other teams in the S&P rankings.

I have already posted that Michigan has historically followed up disappointing 8-4 seasons very well in the past. Here, I looked beyond the record to compare this season to those of 12 successful rebuilding teams. The advanced stats all suggest that this team can make the same leap the other ones achieved. The biggest obstacle, and one that many of the successful rebuilding teams had already overcome, is something that every fan already knows too well: win the big games.



November 26th, 2017 at 6:50 PM ^

Is it just me or does our season sit on the very large shoulders of one Mr. Grant Newsome?  If he returns to what he was I really like our chances.  If he cannot play and we have a new LT and return JBB/Ulizio, it looks much more grim to beat the NDs, OSUs, PSUs, and Wiscy's on the schedule.


November 26th, 2017 at 7:05 PM ^

But similar tough looking schedule in '85 (after 6-6 in '84) - ND, SC, Mary were all ranked when we played them. At MSU. At Iowa. OSU, etc. 

Same in '97 (after run of 4 loss seasons) - Colorado ranked, ND was supposed to be good before season, at MSU, at PSU, at Wisky, vs. OSU, vs. Iowa

Both seasons the key to our success was:

1. Great QB play

2. Good enough O line to run the ball in most cases

3. Stingy D

4. Turnover margin

I'm sure hoping Newsome can play again - even if just to know his hard work has paid off. This time, HE gets to wipe out a Wisky player!

I'm hoping the external pessimism is high (Michigan sucks, etc.) and Finebaum is crowing until the ND game. 


November 26th, 2017 at 6:56 PM ^

Actually they were a possession away from winning outright if O'Korn could have driven them to a TD and used enough of the clock. Anyway, good analysis and I think everything is setup for a big bounce back year in 2018. The schedule is not kind but good teams play well on the road. It's time Michigan starts winning those games.


November 26th, 2017 at 6:59 PM ^

7-6 and 4-4 in B1G to 12-2 and 8-0 in B1G regular season.

Some factors:

1. Schedule (home vs away and Non Conf)

2. QB play

3. Turnover margin

4. Key plays in close games

5. Refs

Iowa had all 5 work for them in 2015 until the MSU loss in Indy.

We know the schedule next year is brutal, on paper at least.

But the next three are in our control

Look at our 3 close losses this year -

MSU game (Isaac fumble, McKeon fumbled cost us 3 pts, McDoom drop, etc.) and -5 in TO margin

Wisky - goal line fumble, 3rd and long conversions on long TD drive, 3rd down PI call, Peters knockout

OSU - Metellus drop, JOK missed passes, arm punt, etc. don't recover JT's fumble, etc., lose contain on same side of field and 2 different QBs burn us, etc.



November 26th, 2017 at 7:16 PM ^

And actually tweeted it to Brian and LSA last night... Michigan’s average season record since 1990 is 8.57 wins and 3.75 losses. While that includes 3-9 in 2008, and 5-7 in 2009 and 2014, it also includes 12-0 in 1997, as well as a couple of years without a *technical* loss, but at least a tie or three.

The bottom line is that Michigan, this year, simply regressed to its historic mean. We didn’t beat any good teams; and we lost the two most important rivalry games. But all wins and losses in all years count the same except in our own minds.

If a rebuilding year under Harbaugh simply means an average year in any other recent era, I’ll take it (for now). Sure, we have to start winning with some regularity against MSU and OSU, if for no other reason than than they have a higher emotional value. But to suggest that we are on the wrong track is a willful disregard of how the football program has performed since Bo stepped down.


November 26th, 2017 at 7:45 PM ^

Was the Year 1 A.B. (After Bo).

By comparison, it’s not for nothing that those of us who watched Bo’s teams play saw incredible execution on the field. An average Bo season had 9.2 wins and just 2.3 losses - and those were the years when most college teams played an 11-game season.

Personally, I think that history is both a help and a hinderance to the program. Bo hasn’t coached for almost 30 years; but he is STILL, arguably the soul of Michigan Football. We shouldn’t untether the program from what Bo built; but we shouldn’t be shackled to it either.

EDIT: I missed the other questions you asked in my original answer.

Simply put, what Rich Rod, Lloyd, Mo, Hoke, Bo, and - why not? - even Fritz, Fielding, and the rest did is mostly irrelevant to this year, next year, and every year thereafter. We *expect* to be competing for B1G titles at a minimum. But, sadly, for almost 3 decades now, Michigan Football has not lived up to the standard that Bo set, except as an outlier. That may take several seasons before it’s fully corrected.


November 26th, 2017 at 8:52 PM ^

that comprise over a quarter of your (still arbitrary) sample size.

My point is that the idea of "regression to the historical mean" is totally meaningless given that history has zero direct effect on future performance.

I wrote in this post what I think are reasonable expectations: achieving or building to seasons "in which [the] team makes a legitimate title run and is only a couple lucky of breaks away from ultimate glory."

We already managed that last year. There is no reason to look back at the Rich Rod and Hoke years as a basis to predict whether or not we will achieve the same in future seasons.


November 26th, 2017 at 9:31 PM ^

It’s not an arbitrary sample (which it would have been if I’d picked a random year); it’s the Bo Years against the Not Bo years since. Michigan has not performed to his benchmark, pure and simple.

There is no direct correlation between year-to-year records; that’s true. But it can’t be accidental, either, that some programs have better records, on average, than others. In effect, that’s why some programs are considered “elite” and others aren’t.

My personal view (your mileage evidently varies), backed up by the records from 1969, is that Michigan Football has regressed from a program that frequently competed for championships in the 70s and 80s; to one that sometimes competed for championships during the 90s; to one that rarely competed for them in the 2000s and beyond. A program that averages 4 losses per year over 28 seasons is not one that is going to be in contention very frequently. That’s been Michigan’s reality over that time.

I’m not suggesting that historical records are predictive; I’m suggesting that they are indicative, in that Michigan Football is not - and HAS NOT been - the program we believed it to be. If it were just about those few lucky breaks, as you put it, then some force of nature has put an increasing amount of negative influence over time on Michigan’s odds of them breaking our way.

The alternative is to believe, as I do, that Michigan Football was much further away from reliably (i.e. year in and year out) competing for championships than random events would dictate. It was simply not reasonable for anyone to expect that Jim Harbaugh was the sole missing ingredient needed to return our beloved team to the glory we had under Bo.

Is next year The Year? Maybe. I want to believe that’s true.


November 26th, 2017 at 9:49 PM ^

Michigan State and Clemson are two perfect examples of how much an elite coach can change a program.

As befuggled mentioned above, the college football landscape totally changed after the NCAA set scholarship limits. The parity in the Big Ten has never been higher. It is no longer reasonable to expect Michigan to win the Big Ten consistently. But there is no reason not to expect Michigan to perform just as well as MSU, which has had 2 seasons that were arguably as good as or better than any of Bo's best. Michigan can reasonably be expected to regularly field teams that are at least competitive in the division.


November 26th, 2017 at 10:06 PM ^

I completely agree with everything you just wrote. My point in noting the 8-4 average season, is that something was clearly decaying inside the program over more than 20 years.

Scholarships affected all programs equally. While that may have created more parity, some programs like Ohio State’s had very little change in season records over that time. The Buckeyes weathered the storm; why didn’t Michigan?

Your point about MSU is an interesting one, and it underscores my point. Dantonio changed that program systemically, giving it an entirely new identity in a very short period of time. We can (and do) make fun of the idiosyncrasies associated with that identity; but it’s been effective, almost entirely at Michigan’s expense.

MAYBE the ability to reliably beat John Cooper despite Michigan’s overall record during the 90s, and MSU through most of the 2000s, blinded us all to the atrophy we would otherwise have seen. “Average” seasons don’t mean much - unless they accurately reflect the deficiencies of a system in disrepair.


November 26th, 2017 at 10:34 PM ^

history and program standards. I don't think there were any "systemic" problems with The Michigan Football Program this year. The disappointing results were solely caused by youth (stemming from Hoke's recruiting failures), injuries, bad luck, and some poor coaching here and there.

I think your take is rooted in way too much feelingsball.


November 26th, 2017 at 9:51 PM ^

I guess I didn't care enough, or live early enough, to ever get fully caught up in the Bo mythos.  He always struck me as a very good regular-season coach who struggled to win bowl games.  He did well with the advantages he had, and he won a bunch of games and kept Michigan consistently at the top of the sport.  

This has been discussed elsewhere, but there were 2 Bo eras.  From 1969-1978, he averaged 9.6 wins and 1.5 losses; he never won a bowl game, but with so few opportunities that's not his fault.  But from 1979 until his retirement, he averaged a shade under 9 wins and and lost 3.  Michigan became what they've been for some time now; a good team with spurts of greatness.  That's sort of the steady state for all really good college programs.  But to say stuff like "Michigan has failed to live up to the standard Bo set 30 years ago" is whitewashing history.  Maybe he won a couple more conference titles, and the RR/Hoke eras were a step down, but this is what Michigan football has looked like for most of our lifetimes.  And that's fine to me because they can still reach the pinnacle with the right breaks.  But I'm done with the revisionist history.


November 26th, 2017 at 11:09 PM ^

Well the scholarships went from unlimited (?) to 105 after title IX was passed sometime in the early 70s. And then to 95 during the late 70s? Think 85 was in 1992. Anyway, the first two changes were certainly is a factor in the emerging parity in college football and then the higher # of losses later on once the Big 2/Little 8 era was over.


Also, I'd probably divide his tenure up differently. I think you can make a case that his last 5 years from 85-89 he had a team that was effectively as good as any team in the country for 4 of those 5 years (obviously not 87). The 88 and 89 teams went toe to toe with the national championships each year and arguably should have won both games. 85 and 86 were equally stout especially on defense. That was probably peak Michigan football in the modern era, with the usual sprinkling of facepalm moments to lose games that we should not have. 

You Only Live Twice

November 27th, 2017 at 9:46 AM ^

and especially for a younger person it probably does come across as a certain amount of "mythos."  If you lived through the time, though, Bo was a personality and so much more than a coach.  The 1970s football attendance experience is difficult to describe in words for anyone who wasn't there... the magic of Bo, Canham and Ufer was unique.   I can't get my kids to fully understand it either. 

Bo was such a legend and so loved, his passing had an undeniable and lasting effect on the entire program.   Your numbers are on the money; the Bo mystique if you will was way beyond numbers.


November 27th, 2017 at 12:46 PM ^

Bo was all of the above. Plus, when he spoke it was magical ... both one-on-one or to a packed house he could ignite the situation. I was fortunate to have known him (and Don Canham) when I was a student at UM through my father's stint on the alumni board of control back in the 1970's. That was a unique time in history with the 10 year war, no playoffs, winning the B1G/Rose Bowl was the only prize that really mattered.

The current B1G environment in regards to recruiting and resources is just too equal - outcomes are more random than it was under Bo and Woody during that era of college football. I think with Harbaugh we have the best chance of competing at the highest level possible, but it will take time to rebuild the required team depth and pipeline so that there aren't any "rebuilding" years in the future. Urban walked into a fully loaded program when he took over from Tressel. Harbaugh did not have that same luxury.

The 2017 season was the first real step in our rebuild and it should only get better from here. 2018 will be better, but probably not fully there yet (unless we get that elusively lucky thing all on our side). The 2019 season should be when we find out if the "mission has been accomplished" ... as at that point it will be a 100% Harbaugh team - deep with his recruited talent at QB/OL and those players at year 3 in his offensive system. Defense will be ready and humming with great depth under Brown's leadership. The pipeline will have been built and ready to feed the beast for seasons to come.

Rome wasn't built in a day ... but it has been visited. Go Blue !


November 26th, 2017 at 9:21 PM ^

That's when the scholarship limitations went down to 85 (see here). When Bo started at Michigan, scholarships were (theoretically) unlimited, and during his tenure went to 105 and then down to 95, before going to the current 85 under Moeller. You'll notice Bo's record was better in the first half of his career than in the second half, and this is a big reason why.


November 26th, 2017 at 7:24 PM ^

in the room is...

1 - 5 against MSU & OSU... adding the other hated rival ND next year..

JH needs to turn this around starting next year or the drum for change will start to beat,

rightfully so...


November 26th, 2017 at 7:38 PM ^

But solid QB play, a running game that can move the ball in the 4th quarter, a D that forces FGs on short fields, etc.

I think MSU reverts back a bit next year without getting the favorable TO margin. 

ND isn't that great. I'm just hoping the hype is all in their favor next year and everyone is trashing JH all off-season long. Imagine his intensity and that of the staff and players. 

OSU - well we'll see - but 11-1 would be great, 10-2 with a Big 6 bowl (and a win this time) would be really nice too.

2019 is the year our schedule is tailor made for a NC. Just get back to 10+ wins next year. 


November 26th, 2017 at 10:13 PM ^

...that 1-5 record includes: a miracle finish by MSU in 2015; a royal screw-job in Columbus in 2016; and a game in which merely competent play by our third-string QB would very likely have won the ballgame against OSU in 2017. JH and the Wolverines have had some horrendous luck these last few years. Go Blue!


November 27th, 2017 at 1:35 AM ^

The record is what it is, but 1-5 with a ton of close, terribly unlucky losses makes me more optimistic about the future than if those 5 losses had all been blowouts. Sure, the mental hump of losing big games may be relevant, but what is more likely: Michigan keeps getting unlucky, or luck eventually evens out?


November 27th, 2017 at 11:52 PM ^

Same thing against Indiana and Minnesota in 2015.

The problem is partially luck and partially that OSU is the Alabama of the North. They never have a down year. Meyer is averaging less than 1 loss in conference play per year. That means they win all the games where Michigan has a down year and maybe we split the remaining games if we're lucky.


November 26th, 2017 at 10:13 PM ^

...that 1-5 record includes: a miracle finish by MSU in 2015; a royal screw-job in Columbus in 2016; and a game in which merely competent play by our third-string QB would very likely have won the ballgame against OSU in 2017. JH and the Wolverines have had some horrendous luck these last few years. Go Blue!


November 26th, 2017 at 7:32 PM ^

We are NOT ready to be as good as OSU - need some more years to build what Urban has. 

The bounce back year is what I'm looking for - Bo never had 2 "off" years in a row. Lloyd never had 2 "off" years in a row - even the 4-loss '95 and '96 but had wins over OSU. 

After drinking beer post game yesterday, perhaps with dulled senses, I thought over the crap year we just lived thru - but we had a chance to win OSU last year and this year in the final minutes - different from Hoke, who always played OSU well, we are not losing to Rutgers, at home vs. Maryland, Utah, Minn - and needing last minute stops to beat Akron, UConn, a cfrappy NW team, etc. 

I'm old enough to have been around and grimaced at the end of '79, '84, '87, etc. seasons wondering if we had lost our edge - only to be blessed with '80, '85, '88 with wins over MSU, OSU, and 2 RB and 1 Fiesta Bowl win. 

JH deserves his chance to pull a Bo and have a "bad" UM season followed by one we can enjoy on YouTube game by game. 


November 26th, 2017 at 9:27 PM ^


And Michigan was one possession away from tying OSU until the O'Korn armpunt, so an 11 point loss isn't qualitatively worse than 7 points.


We were actually one possession away from beating OSU (it was 24-20).


November 26th, 2017 at 11:25 PM ^

I'll echo the rest of the comments in that this is really interesting stuff and gives me some optimism for next year. Outside of the rebuilding, I'd be curious to see a breakdown of how many starters each of those teams returned (that data might be difficult to track down, however). Michigan is going to bring back what 17 starters? And Ruiz might be an upgrade on Kugler.


Ecky Pting

November 27th, 2017 at 3:47 PM ^

I can only surmise you left out Saban's LSU teams since they predated the emergence of the S&P+ metrics. Of course, Eye of the Tiger's post you referenced mentions this important team w.r.t. the subject of re-building (not to mention, National Championships). I would also bet dollars to donuts that Eye of the Tiger has a dual-allegiance with the Cajun-Brady Hoke-lovin' Bayou Bengals, but I digress.

The 2002 LSU team was in its 3rd season under Saban, and went 8-5 (62%), after going 10-3 the year before. LSU was 8-4 in Saban's 1st season, a 5-win improvement over Gerry DiNardo's last team, who went 3-8. Of course, LSU went 13-1 (93%)  and won the National Championship in Saban's 4th season, which might be argued as a successful rebuild. Leading up to that, during the 2002 rebuilding year, LSU went 1-3 against ranked opponents, including 0-2 against top ten opponents.

Applying your Quality & Margin of Wins schema, the losing percentage of 2002 LSU's victorious opponents was 27% (much worse than UM Team 138). The winning percentage of its losing opponents was 44% (the same as UM Team 138). Also, LSU had an average margin of loss of 18 points (4 pts. worse than UM), and and an average margin of victory of 22 points (4 pts. better than UM).

Eye of the Tiger

November 27th, 2017 at 5:34 PM ^

Could you please explain? If it's "the year before a team really took off," then it doesn't mean "rebuilding." Oregon in 2009, for example, wasn't a rebuild. Kelly was the OC for the previous coach and retained much of Belloti's staff (and Belloti became the AD).

If you want to find controlled comparisons for Michigan in 2018, look at Florida 2008, Clemson 2011, etc.