There are times of the year when America’s major professional sports receive national attention. America’s pastime, baseball, has a “fall classic,” Thanksgiving is dominated by the NFL, and the country dedicates an entire weekend to the Super Bowl. Christmas Day is a time when the NBA shines with activity from noon to midnight (though the NFL is also looking to get in on the action). When Americans watch LeBron James and Kevin Durant battle it out on the court, they usually don’t think about referees, but it’s impossible to play without referees.
Despite its importance in determining the outcome of a match or championship, the creation and enforcement of rules in sports is often ignored, at least until a referee makes a wrong call.
During the 2010 World Cup finals match, English midfielder Frank Lampard fired a missile that hit the crossbar and beat German goalkeeper Manuel Neuer. Almost everyone in the stadium believed the game was tied at 2-2. Everyone except the referee. Real-time footage showed that Lampard’s shot had indeed crossed the goal line, but the goal still did not count. Germany won 4-1, sparking a conversation among fans around the world about the officiating of this beautiful game, and shortly afterward the introduction of goal-line technology.
Almost a decade later, the English Premier League introduced Video Assistant Referees (VAR) with a similar goal of reducing human error in refereeing and increasing the fairness of matches. But in the six seasons since its introduction, VAR has caused more controversy than it solves. How could that be? Don’t fans want more correct decisions?
Probably not, says Daisy Christodoulou, author of the book “I Can’t Stop Thinking about VAR” and guest on the February 2025 EconTalk episode “Coase, the Rules of the Game, and the Costs of Perfection.” Christodoulou and host Russ Roberts apply economic theory to understand why the pursuit of perfection often leads to unsatisfying results, why continuums are often more useful than categories, and how comparative judgment can improve the creation of consistent rules that align with individual preferences.
As the title suggests, the Coase Theorem is the core theme of this episode. The Coase theorem states that in some cases, individuals can resolve disputes involving externalities more efficiently than governing bodies. This is partly because attempts to create perfect rules that are supposed to “solve” externalities leave little room for the complexity of individual situations. But Russ Roberts argues that rule-making technologies like VAR can cloud refereeing and unnecessarily complicate disputes that could be resolved with common sense. In his words: “Everyone knows what a goal is. Everyone knows what a handball is…yet somehow it becomes difficult when you get to the details of ensuring.” Christodoulou agreed, adding that although VAR was introduced to make referees’ decisions more clear, its role so far has mainly been to cause confusion and frustration. She found that with the introduction of VAR, English football disrupted 11 bottom-up processes as well as top-down rule-making structures, resulting in inconsistencies between accuracy and other fan preferences.
There must be a trade-off between referee accuracy and match excitement. An agonizingly long video review session may produce the right results, but it often reduces goal-scoring power.
Christodoulou believes this represents a trade-off between consistency and common sense. Using the judicial system as an example, Christodoulou argues that people often value a certain level of discretion in enforcing and interpreting laws. However, discretion inevitably leads to inconsistency, accusations of bias, and potential unfairness. In other words, it is impossible to overcome the trade-off, and attempting to do so often results in the worst of both worlds.
This is the result of VAR: overly complex rules are applied and inconsistent.
Christodoulou said that since the introduction of VAR, handball rules have increased from 11 words to 121 words, but what qualifies as handball can vary widely from minute to minute, even within the same match. Christodoulou believes that handball rules misidentify continuous variables as categorical variables. Categorical variables describe mutually exclusive concepts, whereas continuous variables exist on a spectrum. Many of the decisions referees make during games are categorical as to whether an incident is a foul or not a foul. The problem is when the line between the two is blurred. For example, it is very difficult to explain in plain language what handball is and what it is not. In fact, many aspects of everyday life can hardly be explained by language itself. However, this does not mean that it is impossible to determine what is and what is not handball. Christodoulou argues that this can be achieved through comparative judgment and tacit knowledge.
Christodoulou explains using the example of an evaluation form. Please point this out. Determining which of two works is better than the other is much easier than determining how much better the two works are alone. The determination of quality in a vacuum varies widely both between individuals and, crucially, within the same individual. On the other hand, comparisons can be combined to create high-quality distributions. According to Christodoulou, comparative judgment grading provides more agreement and consistency than rubric-based grading.
“So there’s this strange paradox that what feels like an incredibly subjective rating method is actually shown in the data to be very objective. And the flip side is also true. There’s this very objective rating scale, seemingly very objective rating scale, that has all these checklists that say, ‘Is this distinctive? Does this piece of writing feature that?’ Is this distinctive?’ But when you crunch the numbers, people don’t agree at all… It’s really very subjective. ”
To apply this insight to refereeing, technology can first create collections of potential handballs, for example. You can then use crowdsourcing to encourage fans, players, and referees to decide which of the two clips in this dataset is more like handball. By repeating this process, a distribution for determining handballs is created. Here, Christodoulou suggests that AI can be trained to recognize patterns from examples of what is considered a handball that can be relayed to referees during handball disputes. The referee will then decide whether the instance in question is above or below the handball set line.
As always, there are no solutions, only tradeoffs. There are no perfect rules. Attempting to impose rules from the top down results in unclear rules that are applied inconsistently, leaving fans (and players) frustrated. Although it may seem counterintuitive, the key to promoting goodwill toward referees is more likely to come from attempts to overturn further managerial and referee decisions than from recognizing the power of common sense or tacit knowledge.
Kevin Lavery is a graduate student in the master’s degree program in economics at Georgetown University. He holds dual Bachelor of Science degrees in Economic Analysis and Political Science from Western Carolina University.
