Why we’re right to be afraid of robot referees - and why it isn’t for the reason you’d think

By Mike Warren

When horse racing photographers started experimenting with photo finish cameras in the 1880s, they had no idea what was to come.

For two thousand years, determining the outcome of a horse race had been entrusted solely to human senses. But the early days of the technological revolution sparked belief in a better tool.

The photo finish camera was billed as a game-changer. No longer would close calls be blighted by controversy as ageing judges squinted to decide a winner. Sport was about to get fairer.

For the next 80 years, technology struggled to match that ambition. Image distortion plagued early efforts, often making trailing competitors appear in front. By the 1964 Tokyo Olympics, human ‘observers’ were still seen judging track events... twenty-two of them.

Things have changed since then, with today’s ultra-HD cameras subdividing seconds into ten-thousandths and beyond. But our obsession with applying technology in the quest to make sports officiating fairer is as relevant as ever.

Now it’s AI’s turn. In the last year, it has put Wimbledon line judges out of work, reduced the NFL’s chain gang largely to spectators and been rolled out in more football leagues worldwide.

Even the most traditional sports like boxing and gymnastics launched trials into their own AI scoring systems, while the Spanish FA announced plans to use theirs to evaluate human referees.

Naturally, this has brought questions.

Like the early photo finish cameras, AI’s accuracy and reliability have been debated, with every high-profile mishap or error demanding attention.

Technology actually has a compelling argument here. The evidence is clear. Yes, mistakes happen, but AI has already proved itself better than the human eye. It makes far fewer errors, and the data shows it’s not even close.

There are, however, valid ethical concerns.

On the one hand, technology can shield human officials from hate and abuse. But, like the protestors who stood outside SW19 dressed as Wimbledon Umpires holding “Don’t let the bots call the shots” placards, do we really want to just submit to the machine?

There’s also a question of clarity. Like many supporters, I spend my weekends sitting in stadiums waiting for decisions that are vague at best. Do fans and players actually understand how AI is

making these calls? Doesn’t that matter?

And then there’s history. Referees and umpires are woven into our sporting fabric. Surely replacing them risks losing something no machine can ever quantify?

All legitimate questions.

But should we even be trying to make sport fairer in the first place?

Like auto-tune in music, AI referees might be technically correct. But is ‘technically correct’ what we really want?

John McEnroe’s tirades against Wimbledon officials, the ‘Hand of God’, the three seconds in Munich, Thierry Henry’s handball, and the 2021 Abu Dhabi Grand Prix.

Arguably, five of the biggest moments in the history of sport, remembered and debated up to this day, not for delivering fair results, but because they gave us unfair ones.

This is a key difference between humans and AI. Computers certainly aren't immune from making mistakes on the field. And they probably never will be. But a technical glitch isn’t the same as a human referee making a contentious decision on the pitch. Fans can complain about the computer error all they like, but they can’t debate it.

This isn’t an argument for throwing out the rulebook. Officiating must be clear, trustworthy and credible. But the risk is that by chasing fairness, we might be jeopardising why we watch sport in the first place.

European football, some argue, is in competitive decline. Wealth disparity between clubs may be the root of the issue, but VAR threatens to accelerate it. In the words of economics professors from Molde University, it’s reducing the uncertainty of outcome.

Data shows VAR is leading to more penalties being awarded, rewarding more attacking teams (i.e. the better teams) in the process. It also restricts the opportunities for ‘dark arts’ weaker teams have long employed to level the playing field. Where once shirt-pulling and cynical fouls slipped under referees' radars, they’re now being flagged more often. Essentially, the best teams are getting better, and the rest are feeling the impact.

It’s early days, but the concern is that by handing over the whistle to AI, we might inadvertently be making sport what we all should fear the most. Predictable.

The idea that there could be such a thing as ‘too much’ fairness in sport may seem ridiculous.

But as sport and spectacle continue to entwine, perhaps we should at least be considering it.

After all, there’s a reason why Baller League and other new leagues like it have introduced rules like wildcards that tip the odds in one team’s favour. It’s the same reason why riders in Modern

Pentathlon were randomly assigned horses 15 minutes before the competition started. It’s also why ski competitions, test cricket and The Open Championship are inherently intriguing.

Unfairness makes sport uncertain. And uncertainty makes it interesting.

The challenge for those in charge of sports over the coming years will be deciding where the

balance lies.

Rules are an essential part of sporting competition. They’re why sports matter. Taking them away shouldn’t be the answer, whichever sport you are.

But while fairness should be an ambition for everyone, doing so without any consideration of the consequences to entertainment might in fact be the riskiest move of all. Because it doesn’t matter how fair sport is if nobody is interested in watching it.

As we navigate the years ahead and whatever newfangled technology it brings, we’d do well to

remember it.

Mike Warren is a Strategist at Dark Horses