Football and Agency Costs

<STD: Serious Topic Disclaimer>

In the interest of offering content both didactic and accessible, the following Serious Topic will not include proper source support. This means you shouldn’t believe any of it. If you find the discussion compelling, however, I encourage you to pursue a more demonstrable truth.

The football season is over, and I’m sure we’ll see plenty of personnel changes before next season rolls around. In light of potential head coach firings, some interesting parallels between corporate law and football have been rolling around in my mind recently.

Football is played at notoriously sub-optimal levels. Bill Barnwell over at Grantland is a great guy to read if you don’t believe me. At virtually every level of the game—from personnel decisions and player contracts to real-time game strategy—huge mistakes are regularly made. Part of the reason for this is that most football fans (and even football front offices) think about certain aspects of the game in the wrong way.

I’ll note that, as a general rule, front offices are usually extremely sophisticated actors with a lot of proprietary means of analyzing the best courses of action. So they often deserve the benefit of the doubt. But the fact remains that the NFL is fraught with poor decision making, and coaches are often faulted for all the wrong things. So I’m breaking out my slingshot and going after this Goliath.

The first thing worth mentioning is kickers. Take everything you know about kickers and throw it out the window. There is no such thing as a “good” kicker once you get to the professional level, and there are more than enough kickers to go around. Sure, it’s possible to have a good season or a bad season, but statistical analyses have revealed that successful kickers in any given season are actually slightly more likely not to do well the next season. Essentially, the moral here is: don’t pay kickers. There’s always a guy waiting in the wings who could do just as well. Anytime you see a team franchise tag a kicker, you know that front office is off its rocker. To use the baseball analogy: All kickers are replacement level.

Speaking of kickers, stop freezing them. There’s also no statistical evidence that freezing the kicker works (for those non-football fans, this means calling a time-out just before the kicker attempts a field goal, where the intention is to throw off his focus). There’s good statistical evidence that it uses a time-out, however! If I were to win the lottery tomorrow and buy a football club, the first directive I would give to my head coach would be not to try to freeze the kicker. Time-outs can almost always be put to a better purpose.

Of course, kickers should play a much smaller role in the game than they currently do. Coaches have at their disposal all sorts of grids and charts that map out the best course of action in a given situation. For example, at some point during a drive it becomes more valuable to go for it on fourth down than to try for a field goal. Coaches, in theory, know what that point is (based on their offensive potential, the defense they’re opposing, &c.). The same is true of two-point conversions. Every sideline general has the data on when and how often they should go for two (hint: it’s a lot more often than you think). But coaches often disregard this stuff. How often do you see a team go for it on fourth down? Well, it happens all the time when that team is down 14 with 4 minutes to go, but hardly ever in, say, the first quarter. But in theory a coach should take the course of action that best positions his team to score, regardless of the context, right? Well, sort of.

This is where I see the parallels between football and corporate law. Corporate law is a little more complicated than the one-phrase summary I’m about to give, but it boils down to resolving the inherent conflict of interest that arises out of the separation of corporate ownership and corporate management. In other words, agency costs. People who run a corporation are somewhat inclined to be risk-averse, but people who own the corporation probably don’t want them to be. Consider the following:

Imagine WidgetCorp is a corporation. They manufacture and sell widgets. One day, one of their engineers comes to the board and proposes a new widget. This new widget is very fancy and expensive to produce, but if it catches on it could be huge. So the board is faced with a question:

Option A: Produce FancyWidget. This offers a 25% chance of making $1 billion and a 90% chance of total failure, in which case the production cost of $100 million is lost. (.25)(1,000,000,000)-100,000,000 = $150 million. This is the corporation’s expected gain. If, in a million parallel universes, this same corporation were to pick Option A a million times, those millions WidgetCorps would average a gain of $150 million.

Option B: Continue producing RegularWidget. This offers a 100% chance of making $150 million, minus the production cost of $10 million. (1)(150,000,000)-10,000,000 = $140 million. This is the corporation’s expected gain under this plan. Again, if we had a large sample size, these numbers would bear out.

What should the board do? One answer might be that they should continue with Option B—the standard, safe option—because they are guaranteed to make money, whereas under Option A they will most likely lose $100 million. In that scenario, the corporation might fold, and the people who run the company would be out of jobs. Or the corporation might get taken over (and still, the managers would be out of jobs).

But stockholders are more likely to answer that the corporation should go for Option A. On average, it produces a better result, and besides, the shareholders probably have diversified their risk with other investments. The risk of losing value on their stock in WidgetCorp isn’t such a menacing specter, and they have a lot to gain.

In fact, the board will probably opt for Option B. They don’t want to lose their jobs. That means that the corporation has effectively lost $10 million as a result of its officers being risk averse (the next time you boggle over CEO salaries, consider that maybe that CEO is just being rewarded for low risk aversion). When corporate officers get paid to run the company, that’s an agency cost. But it’s also an agency cost when they make sub-optimal decisions at the shareholders’ expense.

So back to football. When a team is on their opponent’s 10-yard line on 4^th and 1, what does it do? Well, it probably kicks a field goal. In the first three quarters, it almost definitely does. What are the correct equations here?

Choice A: Field goal. Let’s pretend a field goal this close has a 100% chance of success (a false premise). So, the expected gain is (1)(3) = 3 points.

Choice B: Go for it. Let’s pretend our NFL team can run gain at least 1 yard on a play 50% of the time. This is a pretty conservative figure; just run it up the middle with all you’ve got. It’s a very high-percentage play when the bar is this law. So, if we were on the 1-yard line, this is an easy equation: (.5)(7) = 3.5. The run is the better call. But, if we’re on the 10-yard line, things are a little more complicated. We have a 50% of 1) A first down somewhere between the 10 and the end zone, or B) a touchdown. Let’s assume turnovers are considered in the other 50% that represents failure. Essentially, if we don’t lose possession, do all the permutations open up a window to a higher expected point average than 3? The answer can only be ‘yes.’ The worst case scenario is we kick a field goal (ok, actually it’s a turnover, but those are unpredictable), but there’s always a chance that we’ll get the touchdown.

It’s also important to note that in the above scenario, even if our team fails to make the first down and surrenders possession, the opponent obtains possession with terrible field position. They’re susceptible to a safety and have a decreased chance of scoring points. On a kickoff, they’ll probably have better field position, and might even run it back.

Of course, there’s context to consider. Risk aversion is OK sometimes. I, as a poor law student, was once posed (by my corporations professor) the following hypothetical:

I offer you two choices. You can have $100,000, guaranteed, for free, or you can flip a coin and receive $0 for heads and $250,000 for tails. Which do you choose?

If you’ve actually read this far, you know the expected gain is higher for the second option. (.5)(250,000) = 125,000, which is greater than 100,000. But I’m risk averse because of my poverty. $100,000 would be so helpful to me that I couldn’t pass it up, even in the face of a potential gain of $250,000. Now, if my professor had offered to repeat this ritual with me every week, I would have opted for the second option every time (unless, perhaps, I had some immediate use for the money and could gain more from a guaranteed $100,000 than from a contingent $250,000).

Similar circumstances can arise during games. If my team is down by one with 0:01 on the clock, I want them to kick that field goal even if they have a higher expected gain via some other strategy. Football is notably different from corporate operations in that NFL games exist in discrete capsules—games. These artificial beginning and ending points can justify decision making that would otherwise seem irrational. But they can also dull our edges.

I often think what I would do if I were an NFL coach. I think I’d go for it on fourth down a lot. I might set up some rules that look something like this:

Go for it on fourth down when:

            1) You’re in enemy territory

            2) There are <2 yards to the first down

            3) It’s necessary to secure a win

Pretty simple, but I think it would be effective. Coaches should do this. They would, I really believe it, be more successful if they did. So why don’t they? Why does this problem persist? Why why why?

Well, I don’t know. I do have a theory, though. Let’s take another look at corporate law first.

Corporate agents do create agency costs (like any agent does), but there are a bunch of reasons why these are limited and why our corporate structure remains a good system. They’re called market forces. Let’s look at a few market forces that suppress agency costs:

1) Labor market: Being a good corporate manager makes you more desirable to other corporations. So even if the risk of loss associated with certain decisions is a deterrent, the potential gain associated with being recognized for sound decision making (and the expectation that good risk/cost evaluations will lead to better profits in the long run) should motivate good decision making.

2) Capital market: Running a corporation better makes it more valuable, and this motivates potential stockholders to buy into the corporation. This is a good cycle to get in, so corporate managers should be motivated to make good decisions to attract more investors and grow the corporation’s value.

3) Corporate control market: Poor decision making opens the door to other corporations to come in and take over. This also carries with it the threat of job loss.

4) Product market: Better decisions lead to better products, which lead to better sales and a healthier corporation. In the WidgetCorp example, maybe the FancyWidget has a place in the market that would make it very valuable, as well as increasing WidgetCorp’s market presence.

Essentially, if we presume the market is efficient, the corporate managers with the lowest agency costs are most valuable. In the NFL, however, the size of the market makes it pretty inefficient. This is my theory.

With only 32 teams, many of whom have institutional coaches who aren’t going anywhere until they fail miserably, the market does a surprisingly poor job of actually weeding out the bad from the good. Not to mention that the NFL is so full of illusory success and failure (how many statistics actually reach significant levels over the course of a season? We have no idea who’s good—coaches, teams, or players—in the NFL, really). The general level of sophistication is too low; too little information is available. I’m not sure these market factors—still totally applicable in the NFL—have a chance to make themselves known. In short, the market is just massively inefficient. We see problems like this all the time in professional sports: Even if we can calculate based on available statistics a player's value, usually only a handful of teams are even interested in buying. Supply and demand does wonky things in these scenarios. It's the proverbial water bottle-in-a-desert problem.

It’s unfortunate, but I think coaches might be right to make such bad decisions so often. It might be the best thing for their careers, even if it certainly isn’t the best thing for their teams.