Select Page


Is a Future Without Human Referees Possible?

30 Oct 2023   ·   

Gerrit Maus from the Department of Psychology at the University of California discovered that the brain can create visual illusions to compensate for the speed of an object whose trajectory is followed by the eye. To determine its course, the brain might anticipate its position, even if it is incorrect. This is known as the flash-lag effect, which, in sports terms, is the scientific proof that humans cannot achieve perfect refereeing.


In football, a 2005 study found that the flash-lag effect played a role in offside decisions based on the angle line judges had in relation to the play and the proximity or distance of the attackers. Speed made players in motion appear ahead of their actual position. After analyzing all offside decisions from the 2002 World Cup, they found 26.1% of incorrect offsides where the flag was raised and 26.5% where it wasn’t.

Interestingly, a subsequent 2017 study in tennis showed a nearly identical error rate in refereeing decisions. By this time, the Hawk-Eye technology was already in use at the ATP Masters 1000. All shots that players challenged and which led referees to consult the technology were studied. The review showed that the human eye was incorrect 27% of the time, almost the same as in football.

In tennis, this optical effect has been considered for years, especially in game situations with the fastest ball speeds: serves. In 1980, to see which balls landed in, the Cyclops system, an infrared ray technology, was introduced. It underwent improvements because insects or wind-borne particles could trigger it. However, it remained active until 2005 when the multi-camera system Hawk-eye, a technology initially introduced in cricket in 2001, replaced it. While the study investigating tennis umpiring errors determined they occurred within a 33-millimeter margin, this technology boasted millimeter precision. A subsequent system, Foxtenn, started being implemented in 2023 and captures 2,500 images per second, compared to Hawk-Eye’s 340.

In baseball, the MLB is testing Trackman, an ABS (Automated Balls and Strikes) technology that also works with Hawk-Eye, to settle the sport’s most contentious calls: whether a pitch is a ball or a strike. In football, in addition to VAR, Goal-line technology for ghost goals and the SAOT system for offsides have been introduced. SAOT involves monitoring 29 points on each player using twelve cameras around the field, capturing 50 images per second, allowing for 3D gameplay recreation with minimal delay. It’s known as Automatic or Semi-automatic Offside.

The NBA is also in the final stages of implementing a Hawk-Eye system that tracks 29 points on each player and the ball in real-time. This is intended to replace the Second Spectrum, which used six cameras to monitor players’ torsos in 2D. Now, the NBA Replay Center will have algorithms to determine if the ball was ascending or descending when touched by a defending player or if it hit the backboard first. This technology isn’t just for refereeing; it’s also for general data collection to enhance broadcasts. It’ll provide insights like pass speed, sprints, shot angles, jump heights, distances between players, arm extension in a shot, and even the hand’s angle. Additionally, processing this data will aid biomechanical analysis, significantly impacting training, preparation, and overall performance potential.

This cascade of advantages shows that technological refereeing will inevitably become more common, and other potential applications may also bring artificial referees into play. In football, FIFA’s previous reluctance to introduce refereeing technology has been quickly forgotten. Years ago, Blatter wanted the game to be played the same everywhere since football is universal. It was also argued that pauses would disrupt the game flow and that fans enjoyed debating controversial refereeing decisions. With VAR’s introduction, those debates have vanished, and the system, though not without its controversies, is now well-established. Of the earlier reasons against technology, the most notable was the “slippery slope” theory. Platini once argued that each technological advancement would prompt the next. The then-UEFA president stated he wouldn’t allow advancements leading to a “less human” football.


The fact remains that advancements have consecutively occurred, and the debate about humanity’s role in sports is more heated than ever. Aldo Comi, director of Soccerment, a startup focused on data analysis and collection, has stated that cameras on the field will only increase to gather more and more information. This data will be used for more accurate refereeing and crucial insights to enhance tactical gameplay or performance. This technological surge is inevitably leading to the development of Artificial Intelligence to make refereeing decisions—a future without human referees. 

However, all indications suggest that, in the medium term, even with exponential data collection and the capability to train an AI for refereeing, technological solutions will complement rather than wholly replace humans. But that doesn’t mean the same number of humans will be employed in refereeing. In tennis, the ATP has already announced that by 2025 line judges will be replaced by a camera and AI system.

But we must remember that technology also fails. Recently, in Ireland, Hawk-Eye malfunctioned due to hardware issues during a Gaelic football All-Ireland semifinal between Galway and Derry. The ball clearly went between the posts, but the technology didn’t award the point, which had to be rectified after the game’s halftime. While this might be a simple technical error, easily fixable, it impacted the technology’s credibility. This leads us to debates or philosophies in robotics and automation: human trust in machines.

During a match between Andy Murray and Corentin Moutet at the US Open, the software for referees to review plays—a VAR equivalent—crashed, and they couldn’t review a mistaken decision. Murray commented that while technology is good if it stops working, it’s “a farce.” Due to such incidents, another tennis player, Alejandro Davidovich, protested at the Madrid Open, convinced that Foxtenn was malfunctioning and wrongly awarding points against him. In technology, the likelihood that a system will fail at some point is one. The issue with these errors, whether in software or system failures, inevitably leads to reactions like Davidovich’s, and recurrently so. The mistrust between users and robots has been a studied phenomenon for decades. Research has argued for the need to measure, quantify, and engineer trust in robot-human interactions.

In technology, the likelihood that a system will fail at some point is a certainty. The issue with these errors, whether in software or system failures as previously mentioned, inevitably leads to reactions like that of Davidovich, and recurrently so. The mistrust between users and robots has been a studied phenomenon for decades. Research has argued for the need to measure, quantify, and calibrate trust in artificial intelligence to address this issue. Recent analyses, in fact, have found that robots are not necessarily perceived as good referees in a sport such as football. The general public initially holds a subjective perception of technology. It only takes incidents like the aforementioned for that trust to be irreparably damaged.

In conclusion, the future of refereeing doesn’t paint such a bleak picture for humans. It is foreseeable that, in the long run, robotic referees will proliferate in areas where human resources are lacking, such as in developmental and amateur leagues. The shortage of referees is a global phenomenon that is already impacting grassroots sports. Here, technology can serve as a lifeline for these categories. However, in the pinnacle of competition, where every decision is crucial and watched by millions, what will truly be challenging is for human referees and robots to afford the luxury of doing without one another.


Building the future of the sports industry