A tennis line-call review flashes across the broadcast graphic: a sub-millimeter gap, decided in an instant. Many viewers still squint at the replay and cannot tell fair from foul with their own eyes. That gap between what we see and what the system declares is where the whole debate over machine officiating begins.
The Disputed Call Replays
Cameras and sensors were supposed to end the arguments.
Photo by Instead, they just moved them somewhere else.
A close line-call review shows a razor-thin margin on the broadcast graphic, and many fans still can’t perceive it as decisive. A slow-motion offside frame gets played on loop while two commentators point at the same image and reach opposite conclusions. The evidence sits right there on screen. The agreement doesn’t follow.
For most fans, this means an automated verdict rarely ends a dispute. It just relocates the argument from the field to the living room.
How the Machine Actually Decides
The systems doing this work behave less like an all-seeing eye and more like a very fast estimator.
Photo by Semi-automated offside technology, a system that uses limb and ball tracking to model player positions at a single instant, debuted with ten cameras mounted under each stadium roof at Euro 2024 [Engineering].
Ball-tracking cameras don’t photograph every millisecond directly. They interpolate a trajectory from a handful of angles, then report the statistically likeliest position. Engineers, not referees, set the calibrated thresholds that turn that estimate into a simple in or out call.
The machine never actually witnesses the play. It calculates the most probable version of it. That built-in estimate, not any camera malfunction, is what quietly feeds the lingering unease among fans.
Trust Cracks Beneath the Precision
The precision gains are real and measurable.
Photo by In the English Premier League, video review reportedly raised correct refereeing decisions from 82% to 96% after its introduction [Tencent]. One study found referee accuracy climbing from 92.1% to 98.3% once similar review systems stepped in [Em360tech].
Players still ask for explanations anyway. Coaches still treat the ruling as something worth disputing. Post-match interviews now include questions aimed at the technology itself, as if it were a referee who could still be talked into changing its mind.
Academic reviews of AI in sport keep circling the same worry: “Ensuring transparency and fairness in AI algorithms matters, and continuous monitoring and evaluation are necessary” [NIH]. Better tools didn’t erase the instinct to question the umpire.
Echoes of Past Officiating Fights
Photo by This argument predates any algorithm. Fans still cite decades-old blown calls as proof that officiating carries some baseline fallibility. Those missed calls became defining scandals long before any camera array could second-guess a referee.
Leagues introduced review technology specifically to end fights like those. Major League Baseball plans to give each team two challenges per game on ball-or-strike calls under a new automated system starting in the 2026 season [Tencent]. Soccer has now run semi-automated offside through a men’s World Cup and a Women’s World Cup.
The referee changed from human to machine. The grievance stayed exactly the same. Research reviews also warn that AI can inherit human biases, pointing to documented home-team bias in soccer officiating that a poorly built model could easily reproduce [NIH].
The Same Call, Seen Differently
The same ruling reads differently depending on who’s watching and what they stand to lose. In fan surveys, supporters of the losing side consistently rate a call as less accurate than supporters of the winning side, even when both groups watch identical footage.
Broadcasters describe the automated call as proof of progress. Fan forums describe that same call as reason for doubt. Each side receives the moment differently:
-
The league sees consistency and speed, advantages a tired human referee can’t match.
-
The losing fan sees a threshold that happened to fall the wrong way.
-
The engineer sees a model behaving exactly as it was designed to.
The fan squinting at that replay isn’t really arguing with a machine’s eyes. They’re arguing with a choice made long before the whistle blew: where an engineer set the line between in and out, how many cameras got bolted under the roof, which margin counted as decisive. AI didn’t remove human judgment from the game. It moved that judgment upstream, into design rooms most fans never see. Next time the graphic loads and the verdict appears, watch for the small pause before it decides. Someone already made that call, long before the ball ever crossed the line.
