Harbaugh vs. Romer


Patrick Barron / MGoBlog

I’ve always subscribed to the theory that success in college football requires four ingredients, all of which are necessary but not sufficient for success — in other words, to win the type of games Michigan has the intention of winning under Jim Harbaugh, we’ll need excellence in all four: recruiting, player development, scheme, and in-game coaching.

Michigan has recruited at a high-level under Harbaugh and will continue to do so. The program’s foremost recruiting analysts attribute this to the combination of the staff’s ability to sell the university’s greatest assets (academics, tradition, Ann Arbor, life after football, etc.) while at the same time constantly putting the Block M in the national spotlight. Be it with the Jordan brand contract, Signing of the Stars, or having Derek Jeter, Michael Jordan, and Charles Woodson all on the Big House field during the team’s home opener. When you combine Michigan’s resources with Harbaugh’s incredible work ethic, it isn’t hard to imagine more prospects with a stature similar to Rashan Gary’s pledging to the Wolverines.

Jake Rudock is a walking testament to Harbaugh’s player development acumen. The in-season transformation Harbaugh made with his signal caller in 2015 was nothing short of stunning and further reaffirms he is truly second-to-none at developing the game’s most important position. But it isn’t just the quarterback. We’ve seen upperclassmen like AJ Williams and Ben Braden take great strides in the short time this new regime has been in Ann Arbor. Especially considering what Harbaugh was able to accomplish at Stanford with incredibly limited talent, we can sleep soundly about this vital piece of the puzzle.

I won’t pretend to be an expert on the schematic elements of football — while I played a bit in high school, my knowledge of the game on a technical level is well beneath those who break down film for a living and can truly analyze what’s happening on a player-by-player level (hey, knowing what you don’t know is just as important as what you know). Though it is not just coaches who fit that description — MGoBlog’s Brian Cook puts every play of every Michigan game beneath the magnifying glass in his weekly Upon Further Review columns.

Believe it or not, one of the most encouraging of those columns I’ve ever read was Cook’s analysis of last year’s offensive performance against Ohio State. While Michigan lost handily and only scored 13 points, the team had a very well-reasoned and creative offensive gameplan and likely would have been quite competitive against the extremely-talented Buckeyes if not for attrition on the defensive line and a puzzling (and perhaps distracted) approach from ex-defensive coordinator DJ Durkin. From the piece:

We’ll talk about the OL numbers in a bit. Now I want to focus on Michigan’s gameplan and that RPS number. I loved this gameplan, with a couple of exceptions we’ll talk about. When Michigan got stalled it was because a play that should work got blown up because either 1) OSU’s talent level greatly exceeded Michigan’s or 2) a Michigan dude just blew his assignment. Aside from a couple of run blitzes that Michigan wasn’t prepared for OSU didn’t have much to respond with.

They didn’t actually need a tactical response since they had Joey Bosa going up against Erik Magnuson and Mason Cole, but I thought the Michigan coaching staff got the maximum out of what they had. That was just 13 points because they got zero help from the defense and OSU was able to out-talent them enough to bend but not break. Michigan drives:

  • 10 plays, 43 yards, indefensible punt from the 36 on fourth and 5 / 7 plays, 28 yards, punt / 14 plays, 72 yards, chip shot FG / 11 plays, 92 yards, TD / 6 plays, 15 yards, punt / 7 plays, 57 yards, chip shot FG / 20 yards on first play, Rudock knocked out on next, GG

That was against a D that had three top 20 picks and three more guys go before the third round was over; Michigan had 307 yards on just six drives. If they get the 9 or 10 possessions with Rudock that you’d expect before the game they easily crest 400 and approach 500.


This is a ton of good stuff! Michigan got piles of mismatches in this game, whether by alignment or personnel. There were correspondingly few plays that I thought were bad ideas from the start. (One of them was asking Butt to block Bosa on a power play.) They took advantage of many RPS+ plays and left big chunks of yards on the field on others. If this was year four I’d be upset about some of those breakdowns; one year removed from a Brady Hoke offense I chalk most of that up to the lack of player development characteristic of his regime.

The other thing that’s impressive about the RPS: the rest of that chart. It is hard as hell to scheme your way to yards when you are being mostly dominated across the front. Michigan managed it despite a lot of doors being emphatically closed.

If you’re on this site, there’s a good chance that you have already read that piece and know that Harbaugh has the ability to not only to get his teams to realize their full potential in a hurry (Michigan’s 10-3 record last season exceeded even the most optimistic expectations) but also to exploit his opponent’s weaknesses while mitigating his own team’s as much as possible.

The final piece is in-game decision making. This subject will be one of the primary focuses of these very pages; my interest in it began in earnest after a conversation with Craig Ross.

I met Craig for the first time after listening to WTKA AM 1050’s MGoBlog Roundtable, where he’s a regular panelist along with Cook and a few other experts. He’s an author of two books on Michigan Athletics, one of which is now out of print. He mentioned that he had some copies lying around and if anyone were interested he’d happily give one away, so I took him up on it. When I went over to his house, we ended up talking for an hour and a half about where the program’s been and where it’s going.

The host of that show, recruiting authority Sam Webb, playfully refers to him as “the elder statesman,” but he isn’t kidding. A veritable encyclopedia on Michigan Football, Craig recalled to me vivid details of games that happened long before I was born and talked about the relationships he’s developed with people within the program, none more notable than National Championship coach Lloyd Carr, who he remains close with.

Here’s a passage from his second book, The Search for the Unified Field Theory, that outlines a major theme of his work:

So here’s the deal. Einstein’s follies aside, I believe there is a unified field theory in football, a fundamental set of ideas or equations that can help to explain “what happened” in a football game, why the game was won or lost. More than this, I think this theory allows a coach to make choices and structure his team and program in a way that squeezes out a few more wins than might otherwise be the case.

And therein lies our aim: finding it.

Craig and others have good reason to believe that not punting very much, or at least substantially less than coaches tend to do currently, is a part of it. The natural starting point of this conversation is the work of David Romer, a Cal Berkeley economist, who wrote the seminal paper, “Do Firms Maximize? Evidence From Professional Football.” An excerpt from a New York Times profile of Romer explains the motivation behind it:

Romer, a lifelong sports fan who is a professor at the University of California, Berkeley, came up with the idea to rigorously examine fourth-down plays after listening to a radio broadcast of an Oakland Raiders game in his car about a decade ago. Although the Raiders had the ball in striking distance of the end zone, one of the commentators remarked that they would be smarter to kick a near-certain field goal rather risk going for a touchdown.

“I am pretty analytic,” Romer recalled telling himself. “That is a pretty shallow way of thinking about it.”

So, after stewing over the idea for a couple of years, he set out to tackle the great fourth-down debate. Armed with high-octane statistical tools, Romer and a group of Berkeley students analyzed the results of thousands of plays from more than 700 N.F.L. games. The research team encountered data issues that most economists never face.

Romer’s work has been largely met with skepticism in coaching circles because conventional wisdom in sports is an incredibly difficult hurdle to overcome. Coaches at big-time universities or professional franches with nine-figure payrolls often stick to the proverbial book — and why not? If you fail underneath the brightest lights it’s easier to defend oneself when the decisions one has made echo the way things have always been done. Wins are at an extreme premium and there is likely no career in the world with less job security or more turnover than professional or division-one college coaching. So you do everything you can to hold onto that gig, because they don’t come around much.

But doing great things, in the private sector or on the football field, involves bucking convention. Romer’s proven that this means punting less. Way less.

You can read this article for a quick tutorial on Romer’s work and its application. While the NYT model doesn’t explicitly reference the professor, their “4th Down Bot” comes to a very similar conclusion, just with more data (games) to back it up. It boils down to this:


Especially near the middle of the field, coaches need to be more aggressive. So let’s look at where Jim Harbaugh’s decision making fell on the spectrum through the 2015 season. A few points first.

  1. While Romer and the New York Times used NFL data, Romer believes the ultimate findings would apply to the college game all the same (Craig recounts an email conversation that him and the professor had in his book where Romer explains why).
  2. Bill Connelly’s S&P has designated garbage time as being up by 25 points or more in the second quarter, 22 or more in the third, or 17 or more in the fourth. I filtered this out arbitrarily, in a more liberal manner than Bill does, based on my recollections of the individual games. It shouldn’t affect our findings.
  3. Most importantly: this is not an indictment of Harbaugh’s decision making on fourth down, merely an analysis of how closely his decisions echo Romer’s advice. There are specific team, opponent, or situation-variant times where it would be best to not follow the chart above.
  4. This data does not reflect the final outcome of the decision made but only the decision to go for a fourth-down conversion or punt itself.
  5. We’re going to examine field goal vs. conversion attempt decision making at a later time. This piece is only about punting vs. going for it.

Green designates a decision that follows what the NYT table recommends. Yellow means it did not, but was close to doing so and defensible given the score, situation, or general trend of the game (more on that later). Red designates a time when Harbaugh probably should have gone for it on fourth down opposed to punting. Below the first chart are Michigan’s non-garbage time (my judgement) fourth down conversion attempts.




About those yellows: they were marked that way for a variety of reasons. (Maybe I was a bit too friendly to our coach, I’m a fan and proud alum, after all — leave a comment if you disagree with any of my designations and we’ll continue the conversation).

For instance, Michigan punts on fourth and seven from the MSU 38 with 1:12 to go in the first half of a 10-7 game. That’s not technically the correct decision but consider that a.) Michigan State is out of timeouts, b.) Michigan had allowed exactly seven points over the prior 17 quarters, and c.) had one of the best punters in college football in 2015. MSU runs three quick plays and we go into halftime quietly.

It’s fourth and nine from the Northwestern 34 with 4:19 to go in the second. It’s 21-0 Michigan. The chart says to try a field goal but that’s too deep for then-freshman kicker Kenny Allen, who had missed a 44-yard attempt earlier in the season. Hell, this is on the outer edge of where you would trust even a seasoned NFL kicker. So Michigan punts then Jourdan Lewis houses an interception on the ensuing possession that erases any doubt about the outcome.

This isn’t to say I agree with all of the times Harbaugh deviated from the research. I was upset by the decision to punt from Ohio State’s 36 on fourth and five less than five minutes into that game. Given the talent discrepancy between us and the Buckeyes, who littered the first round of the NFL draft (literally, probably), points would be at a premium and we passed up an opportunity to get some early.

Ultimately, we have a 77.8% “success rate” here, if you will — out of Michigan’s 45 non-garbage-time punts and their nine non-garbage-time fourth-down conversion attempts, Harbaugh’s decision to punt or go was Romer-approved or at least defensible given the complexion of the game (in my opinion) 77.8% of the time. If you want to go by the book (meaning the punts I deemed as defensible but not Romer-approved counting against our success rate instead of for it), our score is 59.3%.

Three questions arise from this:

  1. How does this compare to other coaches?
  2. Is Harbaugh always going to make decisions like he did in 2015 or will he become more aggressive or conservative based on the team he has or the trends of individual games?
  3. Should Michigan emphasize this more?

I don’t have an answer to the first question (yet). He’s somewhere between David Shaw of Stanford and Kevin Wilson of Indiana on the spectrum. There are few situations where I have felt like pulling my hair out when the punt team runs off the sideline since Harbaugh’s homecoming, at least (which was not the case under the previous regime). This is something we’ll look into in the future.

Two. Harbaugh’s malleability on fourth-down philosophy appears rather opponent-variant. Against Penn State (a 49-10 victory) earlier this year, Michigan went for it on…

  • 4th and goal from the one with 9:51 remaining in the first (no score).
  • 4th and two from the PSU 39 with 8:33 to go in the second (21-0 M).
  • 4th and seven(!) from the PSU 34 with 1:59 to go in the second (21-0 M).
  • 4th and four from the PSU 28 with 7:57 to go in the third (28-0 M).

So, yeah, that’s pretty aggressive, and all of those decisions of course agree with the research. The aggressiveness may be attributed to the struggles of kicker Kenny Allen, but it’s an interesting data point nonetheless. The next week in a 14-7 slugfest between Michigan and Wisconsin (both of which have some of the nation’s premier defenses), Harbaugh was markedly more conservative. Punting on…

  • 4th and three from the Michigan 43 with 7:06 left in the first (no score).
  • 4th and six from the Wisconsin 41 with 2:59 left in the fourth (14-7 M).

Michigan outgained Wisconsin nearly 2.2:1 and could have won this 23-0 if not for three missed chip-shot field goals and a long return off of a flukey interception that set up Wisconsin with a short field, so the behavioral deviation makes sense. Sometimes, “punting is winning.”

And finally, does this even matter? I’d like to think so, since I just wrote 3,000 words on it, but the answer isn’t so simple.

It’s easy to think that it doesn’t. That Harbaugh being more aggressive or conservative on fourth down won’t affect Michigan much at all considering the twin-barrelled killing machine that he is in the process of building and the conference appearing to be trending back towards those halcyon “Big Two, Little Eight” days of yesteryear. When you’re more talented than the vast majority of your opponents, the difference of a few yards here and there shouldn’t be make-or-break. But it will once the tomato cans have been knocked down and we take 23 south to meet the bastards.

Michigan recruits very well, as I said at the beginning of the page. By “very well,” I’m saying that we are one of the few programs that recruiting won’t be a limiting factor of the program’s national championship aspirations. But in comparison to Michigan’s blueblood peers, we are at an inherent disadvantage in terms of talent acquisition that will have to be made up for in other ways. Because Michigan’s competitors are willing to offer “incentives” that we are not, are willing to perhaps take ethical shortcuts to obtain a player that we are not, don’t have the hurdles of promoting an academically rigorous environment (which top-level prospects aren’t always interested in), and typically have a more-fertile in-state recruiting base without an in-state competitor, it’s going to be an uphill battle for Michigan to bring in the type of athleticism and talent that Alabama or Ohio State does.

The way to compensate for this is clear: be better than them at developing players (have coaching erase the deficit in raw natural talent), be better than them at creating gameplans that exploit the opponent’s weaknesses, and by making sound in-game decisions — both in playcalling and tactically, like on fourth down.

These choices might not matter when lining up against teams like Rutgers and Illinois, but think about Romer when “toe meets leather” in Columbus tomorrow.


2 thoughts on “Harbaugh vs. Romer

  1. Pingback: Harbaugh vs. Romer II | Big House Analytics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s