Fix football refereeing with legal theory (guest post)

We could be doing this so much smarter. Here’s how.

Hi everyone, Grace here, bringing you a terrific guest post from Joel Wertheimer. I asked Joel to write something for me a while ago because he’s great at thinking outside the box and challenging certain assumptions we hold. Here, he’s taken his knowledge of the American legal system and applied it to refereeing football matches, where he believes it holds up very poorly. I don’t suspect everyone will agree with what he’s written, but it certainly provides food for thought about how the sport should be dealing with how to make sure persistent fouls and bad challenges don’t spoil the game. I’ve even let him call it soccer.

Joel is a civil rights attorney at his firm Wertheimer LLC in New York City. Previously, he was Associate Staff Secretary at the White House during the Obama administration. But if you follow him on Twitter, you probably know him best for his frustrations watching his beloved Tottenham Hotspur. And with that, I’ll hand it over to Joel.

In 1974, famed economist Gary Becker argued that the best way to deter crime was to think of potential criminals as rational actors, who would think about the cost of committing a crime as a function of the likelihood of being caught on the harshness of the punishment. Becker theorized that you could create extremely harsh punishments to deter crime while reducing investment in enforcement to accomplish your goals. The model, in many ways, was adopted in the United States. It was an abject failure.

The problem, of course, is that potential criminals are not rational actors. Humans are not rational actors. Humans are very bad at two things that intersect here. First, people are very bad at distinguishing between low probabilities. The difference between one in one thousand, one in one hundred, and even one in ten, are difficult for us to think about. They are all small probabilities. The second is that humans discount hyperbolically, heavily discounting our future selves. One study put the discount rate on a year of incarceration at nearly 25%, namely that one year of punishment has a value of one year, the second just 0.75, and the third 0.5, and so on. By year four, the additional deterrent value of a year of punishment is actually next to nothing.

The Becker method essentially incentivized crime and then led to massive punishment for committing an offense, contributing to mass incarceration in the United States. Reformers seeking to reduce the amount of crime and create a fairer society concluded the opposite approach must be taken, and landed on Swift, Certain, and Fair. Swift meant there was little delay between misdeed and punishment. Certain meant that there was a high probability of detection. Fair meant that the punishment was not overly severe.

SCF has been a huge success when implemented, leading to reduced crime and reduced punishment. The deterrent value of a 50% chance of getting punished for 1 year versus a 10% chance of getting punished for 5 years is significantly higher, reducing the amount of crime committed, while being much less punitive towards those who are sanctioned.

So why in God’s name is soccer refereed like a Gary Becker fever dream? Referees avoid giving yellow cards and red cards because the sanctions are too harsh (as we learned from mandatory minimums) and then give them after some particularly violent act. Reckless tackles in the space between the boxes go uncalled until they receive a yellow or a red for the same conduct, with a red card a massive disadvantage to the penalized side. (For today, we will ignore my other bugaboo, the moral luck of a dangerous tackle that happens not to connect). 

Referees avoid giving penalties for obvious fouls in the box because the punishment is too severe to swing a match, encouraging players to foul, shove, pull shirts, in the box and hope it doesn’t get whistled.

Indeed, last season the Premier League specifically instructed referees not to call marginal fouls so as to not disrupt the flow of the game. This only led to more fouls, and more disruption to this observer.  Last year’s 1393 yellow cards were more than in any in recent memory.  Of course, had referees called every foul early, and told players they intended to do so, the flow of the game would have been improved, because players would stop committing so many fouls.Perhaps we would have seen fewer yellow cards by providing more certain and swift punishments upfront.

In order to improve soccer, we must accept referees as human and acknowledge players as responsive to incentives. That means referees should call the game tightly, specifically in the first fifteen minutes. Rather than wait to hand out yellow cards, referees should hand them out freely early on, reducing the number of fouls later.

It also means weakening the harshness of sanctions such that referees are willing to dole out punishments, rather than refusing to call fouls in fear of swinging the match. Ultimately, that means sin bin orange cards that allow referees to match the harshness of the conduct to the harshness of the foul. Such a rule will improve the flow of games, reduce the perceived unfairness, and protect players by reducing violent conduct.

Referees should also be given the option of awarding indirect free kicks or top of the D penalties for lesser infractions in the box, allowing them to call fouls for penalties that do not feel worthy of granting 0.8 xG to the recipient. Marginal handballs, shoving a player in a corner kick scrum, and a whole host of misconduct should be disincentivized without harsh penalty. This will increase the likelihood the referee blows the whistle conditional on the misconduct occurring.  Of course, the misconduct will go down once defenders are aware they will not get away with a shirt tug knowing that a referee “can’t give that in the box.”

Soccer should be free flowing, uninterrupted by constant cynical fouling. Calling more fouls is exactly how we accomplish that.