The Empty Baseline: Tennis and the Price of Handing Judgment to Algorithms

The Empty Baseline: Tennis and the Price of Handing Judgment to Algorithms

One afternoon on a Masters 1000 centre court, the ball fell close to the baseline on a serve at 30-40. The line judge, a man who had stood on that line for three full hours, raised his right hand and called it out with certainty. The crowd behind him fell silent for half a second, then erupted in protest. The receiving player walked up to the net, raised a thumb to signal a challenge. The big screen behind the court replayed the point in a three-dimensional rendering: a yellow parabola traced the ball's flight, and the point of contact appeared as a red dot, touching the line by roughly two millimetres. The chair umpire leaned into the microphone and read the result. In two seconds, a human judgment had been replaced by an algorithm.

That moment was not loud. There was no trumpet, no announcement. It happened quietly, exactly the way data usually takes power: without fanfare, only with precision. But behind that two-millimetre red dot lies an entire restructuring of authority that professional tennis has been carrying out, from the baseline to the prize-money ledger, from the umpire's chair to the data contracts no spectator ever sees.

Context: from human eyes to the red dot

Tennis is a sport built on a strange belief: that a human standing three metres from the point of contact, watching a ball travel at times faster than 200 km/h, can judge the contact point within less time than a single breath. For more than a century, we accepted that belief. Line judging became an organised profession, with ritual and an entire culture of argument attached: grumbling, head-shaking, rackets smashed into the court, fingers pointed at the line as if the line could speak.

From 2026, when ball-tracking systems entered the biggest tournaments to serve the player challenge mechanism, the boundary between human and machine began to blur. At first, technology played a supporting role: it appeared only when a player requested it, and only to confirm or deny a human call. But every time the rendering showed a line judge wrong, the credibility of human judgment thinned by one layer. After thousands of such moments, the question was no longer whether the technology was accurate, but why a human layer should still sit between truth and outcome.

By the 2026 season, the ATP announced that every tournament in its system would use live electronic line calling, meaning line judges would be removed entirely from the highest-tier events. Wimbledon also announced that from 2026 it would drop line judges, despite being the tournament most attached to tradition. The US Open had gone first in 2026, when the pandemic forced organisers to reduce on-court staffing and they discovered that electronic systems ran more smoothly than expected. An anomalous season, assumed to be a temporary exception, became the test run for a permanent reform.

What is notable is that this decision did not come from spectator pressure. Spectators still like line judges. They like the feeling of a human on court, the sharp calls, the fiery arguments they can retell over a beer. The decision came from tournaments and organisers, where the cost calculation and the demand for consistency won out. This is the point I always stress to readers: when sport changes, the motive is rarely in the crowd's emotion but in the balance sheet behind it.

Why tennis chose the algorithm

To understand why a tradition-rich sport would hand judging authority to machines, we must separate the problem into three layers: accuracy, cost, and consistency.

On accuracy, independent research over many years shows that the error rate of electronic calls on the contact point sits very low, typically reported below one percent under standard operating conditions. That figure matters more than it seems. Human line judges, under ideal conditions, also achieve high accuracy, but that rate drops markedly when the ball travels fast, when the angle is blocked, when a match drags on for hours, or when the crowd creates noise. In other words, human error is not evenly distributed: it clusters precisely in the most important moments, when stamina is exhausted and pressure peaks. Machines do not share that trait.

On cost, the figure few notice is the quiet driver. A major tournament must mobilise dozens, even hundreds, of line judges per round, along with travel, accommodation, training, and management costs. When an electronic system can handle all of that with an initial investment and fixed operating costs, the economics become clear. I take no position on whether this is good or bad; I only record that it is real, and that any analysis ignoring this layer is incomplete.

On consistency, this is perhaps the strongest point of the electronic system. The same contact point, whether on court one or court eighteen, whether in a final or a qualifier, will yield the same result. Humans cannot achieve that. A line judge can be right in round one and wrong in round four, not because ability changed, but because conditions changed. Electronic systems have no good days or bad days.

But stopping at these three layers misses the most important thing.

The price that is not on the invoice

When tennis removed line judges, it did not merely replace a technical function. It removed a layer of mediation between truth and spectator. And that layer, however flawed, had a role no algorithm can replace: it explained.

When a line judge calls, viewers immediately know who decided, who is responsible, and who will face questions if wrong. When an electronic system returns a result, viewers receive a fact but lose an explanation. Why was the contact point like that? What is the system's margin of error? If the system is wrong, who is responsible? No one answers, because no one has been given the responsibility to answer. This is exactly what worries me about the trend of using technology to deliver transparency in sports: it offers an answer without offering the origin of that answer.

In tennis, unlike football where VAR regularly becomes the centre of controversy, the replacement of line judges by machines happened relatively calmly. But calm does not mean problem-free. It only means the problem has not grown large enough to become a headline.

Look at how spectators react. When the rendering shows a contact point, viewers nod. They have no tool to verify it. They believe. But on what foundation is that belief built? A closed piece of software, a proprietary algorithm, a margin of error published in technical documentation almost no spectator reads. We have handed the final authority to a black box, and we call it transparency.

Fans look with their eyes, while I look with a probability distribution. And the probability distribution tells me this: when a system issues a binary judgment under conditions of error, there is always a grey zone, a small share of contact points whose result could be reversed under a different measurement. Tennis has never publicly disclosed that share in a form the public can access and debate.

Putting data after the moment

I want to recall a live-watching experience that changed how I write about this topic. In a fourth-round match at a major, there was a point in the deciding set that at first glance looked like an obvious line-judge error. The crowd booed, TV commentators complained, and social media exploded. Watching the replay once, I too would have written that the judge was wrong. But I decided to rewatch the entire set, noting every point, cross-checking against serve-speed data and the positions of both players. What I found was not in the controversial point.

The real problem was in the seven points before it. The line judge on that line had been repeatedly targeted by the server into the same corner, serve speed climbed through each game, and his viewing angle was blocked by the server's body for most of the points. By the controversial point, he was in a blind spot in terms of both stamina and position. His error was not an isolated event; it was the product of a systematically accumulated chain of pressure.

This is why I always place data after the seen moment, never before. If I open with the number, I turn the story into a soulless report and miss the most important thing: that every human error has a cause, and that cause usually has structure. Data does not lie, but it does not tell the story on its own either. There must be a person on court for the story to begin, and a person before the ledger for the story to be understood correctly.

Since that experience, I never write an analysis resting on a single metric. Every conclusion must come with a section I call the role variable: a detailed description of the operating system, how the human is placed inside it, and which conditions could change the outcome. With the officiating topic, the role variable is the standing position, the viewing angle, stamina after hours, and crowd noise.

The economics of replacing people with machines

There is a point commentary often skips: the savings from removing line judges do not disappear, they are reallocated. Where does the money no longer paid to on-court staff flow? Two directions are most plausible: investment in technology infrastructure and higher prize money for players.

On prize money, tennis has seen steady increases at the majors over the past decade. The total prize fund of a Grand Slam today typically sits in the tens of millions of US dollars, depending on the event and year, with figures around 75 million dollars at the US Open, around 65 million at the Australian Open, and comparable numbers at the French Open and Wimbledon, though each event announces in a different currency. These figures sound huge. But divided across all participating players and minus travel, accommodation, coaches, and medical teams, the take-home of players outside the top 100 is small enough that many still struggle to cover costs. This is the structural paradox of professional tennis: prize money rises, but its concentration rises too.

This links directly to the labour market in tennis. Tennis has no transfer market in the football sense, but it has a labour market: players hire and fire coaches, hire data-analysis teams, sign brand contracts. Whenever prize money and sponsorship shift, this labour market shifts with them. Every number in a sponsorship contract is a market confession about a player's expected value over the next few seasons. When analysing this market, I always remind readers that commercial value and competitive value are not the same. A player ranked twentieth can earn more than one ranked tenth, because the market pays for the ability to draw audiences, not only for ranking points.

And here is where the technology story meets the money story: when tournaments cut on-court staffing costs, they gain resources to invest in broadcast product, real-time data, and things they can sell to a global audience. A court without line judges but with vivid digital rendering is a court easier to sell to broadcasters. That is economically sound. But it also means the truth on court is gradually becoming a packaged product for sale, rather than a process for debate.

The contrarian angle: correlation is not causation

Here I must state clearly what many analyses of sports technology get wrong: they see the coincidence between technology adoption and reduced controversy, then conclude that technology reduces controversy. That is a causal inference built on correlation, and it does not hold.

The truth is that controversies fall not necessarily because technology is more accurate. They fall because controversy is redirected. With no human line judge to blame, spectators do not stop wondering; they only redirect their wondering toward the system. And because the system cannot be questioned directly on court, that wondering is compressed, with no outlet, then fades into silence. A problem that is no longer loud is not the same as a problem that is solved.

This is the tactical and operational blind spot I always try to point out: when we remove a mediating layer, we do not just remove its errors, we also remove its capacity for checking, contesting, and correcting. A line judge can be wrong. But precisely because he can be wrong, and precisely because he is seen to be wrong, the whole system must constantly self-correct. With machines, the self-correction mechanism still exists, but it happens inside the technical room, no longer outside the court, and therefore no outsider can influence it.

Another contrarian angle concerns the human element. We often assume that removing humans means removing emotion and bias. That is true at the technical level. But at the experiential level, humans are not the source of the problem; humans are part of the product. A great tennis match is not great only for its strokes, but also for its people who can err, can be persuaded, can react. When every human layer is replaced by an algorithm, the match becomes more accurate and also flatter. An empty court does not make the result wrong, it only strips away our illusion that sport is a math problem with a single answer.

The people left out of the reform

Every reform has winners and losers. Here, the clearest losers are the line judges. This is a professional group, largely part-time, tied to the sport for decades, who suddenly see their profession shrink within a few seasons. Commentary often frames this as a technological inevitability, an unstoppable evolution. But as someone who has observed the industry for years, I must say clearly: this is not a natural process, it is a decision. And every decision has a decision-maker, interests, and distributive consequences.

The winners, beyond tournament organisers and technology suppliers, also include spectators in one sense. More accurate contact points, potentially shorter matches, fewer drawn-out disputes. For TV viewers, that is a better experience. But I want to return to the financial base of the sport. If hundreds of line-judging positions are cut, how much is saved, and where does it flow? Does it return to players, especially those in qualifying rounds with unstable income? Or does it stay with organisers and sponsors? These are questions tennis, to this point, has not clearly answered.

I do not have enough data to assert anything about the specific reallocation rate. That is a data limitation I must admit. But I can say this with relatively high confidence: when an industry cuts labour costs, the benefit of that cut rarely flows automatically to the remaining workers. Historical tendency shows it usually flows toward the parties with the greatest bargaining power.

Data limits and what I cannot know

At the end of every analysis, I always include a section I call data limits, and I consider it the most important part, especially when discussing technology.

First, the accuracy of electronic line-calling systems is published by the suppliers and tournaments themselves. That does not mean the figures are wrong, but it means they have not been independently verified at a level the public can access. This is a methodological problem: a figure published by a party with an interest in that figure looking good needs cross-verification by a third party without a corresponding interest.

Second, system error is not evenly distributed across all conditions. Fast balls, sharp angles, changing light, and court humidity can all affect the accuracy of the rendering. Technical documentation mentions this, but the degree of impact under real match conditions remains a grey zone.

Third, and most important, I cannot know whether the electronic system is affected by any factor beyond the physics of the ball. This is not a hypothesis about fraud. It is an epistemic limit: when a system is closed, we cannot test hypotheses about it merely by looking at its outputs.

The truth lies deep beneath the ledger, where headlines never reach. And in this case, reaching it requires something tennis is providing less and less: access to how the number is produced, not only to the final number.

The human, once again, placed wrongly

I want to return to a point mentioned above but needing depth: spectators. In football, when VAR appeared, the fans in the stadium became the most forgotten group. They sit in the stands, are not shown the replay, do not hear the exchange between referee and VAR room, and only learn the outcome after a vague wait. Tennis has been luckier here: stadium screens display the rendering result for everyone at once, so at least in-stadium spectators are not treated as outsiders.

But luckier does not mean better. The core issue remains: spectators do not participate in the formation of truth, they only receive the result. In football, I always defend the view that an in-stadium explanation mechanism is a necessary condition for genuine transparency, not just a slogan. In tennis, because there is no such mechanism, transparency is limited to the level of display, not the level of explanation. Displaying a result is not explaining how the result was produced. These two are different, and that difference is the whole problem.

Once again, I do not write about tennis; I only take notes from data. And what data tells me is this: whenever a sport automates judging authority, it must simultaneously automate the duty to explain. If it does only half, it is not more transparent, it is only quieter.

Signals for the next cycle

So what happens next? I always try to close with a forward signal rather than a summary, because a summary is for the past while a signal is for the future.

The first signal is the emergence of independent audit demands for line-calling systems. Once technology has become essential infrastructure rather than an assistant, pressure for public certification standards will rise. I believe that within a few seasons, the probability of this becoming a formal discussion topic is moderate, higher than I thought a few years ago.

The second signal is the rise of data-transparency clauses in contracts between players and brands. As a player's commercial value depends increasingly on performance data collected by systems the player does not own, the question of personal data ownership will become a negotiating point. This is a trend I follow with high interest.

The third signal is the possible emergence of a generation of players trained from the start to optimise for an environment where every call carries probability and every point carries data. This generation will play differently, tactically differently, and perhaps will have a different relationship with officiating: not confrontation, but coexistence with a system they understand better than the old judges did.

Finally, I want to leave a question without an answer, one I believe will remain valid for years. If every on-court decision can be made by an algorithm, where does the remaining human part of tennis lie? The answer may not be on the court, but in the stands, where spectators must still believe in a truth they no longer have any way to verify. That is the question I carry with me whenever I write about this sport, and I have no intention of answering it soon.