Semi-Automated Offside and the Ball-Tracking Era: What It Means for Every Sport

The Premier League's semi-automated offside system went live across all matches from the 2024-25 season. Average review time dropped from around 70 seconds to under 30. That is the headline. The more interesting question is what the underlying architecture tells us about where officiating technology is heading across every major sport.

What Semi-Automated Offside Actually Does

The traditional VAR offside review is slow because it is manual. An operator freezes the video feed, identifies the moment of the pass, locates the relevant body parts of the attacker, and draws a line. The precision of that line is limited by the frame rate of the broadcast cameras and the skill of the operator working under pressure.

Semi-automated offside replaces that process with an optical tracking system. Ten cameras positioned around the stadium track up to 29 data points on each player's body, 50 times per second. A separate sensor system monitors the ball. The moment of the kick is identified automatically from ball acceleration data. At that exact frame, the system draws the line from tracked skeleton data, not pixel estimates drawn by a human.

The result is faster and more precise. Hawk-Eye reports accuracy within 2.6cm at the joint level. The human-drawn line approach has a margin of error that is hard to quantify precisely, because it varies by operator and camera angle, but "not 2.6cm" is a safe characterization.

Why Cricket and Tennis Got There First

Ball-tracking for LBW decisions has existed in cricket since 2001. Hawk-Eye line-calling arrived in tennis in 2006. Both sports have a structural advantage: the trajectory of the ball follows predictable physics within a fixed, known geometry.

Football is harder. The ball moves in three dimensions at higher speeds across a larger playing area, and the relevant question for offside is not where the ball ends up but where specific body parts are at a precise instant. Cricket-derived tracking systems could not be directly ported over.

The engineering foundation still transferred. Camera calibration methods, sensor fusion approaches, and computer vision models for fast-moving objects in outdoor lighting all carried over. Football officiating required additional investment in skeleton tracking, which a cricket LBW review does not need. That investment arrived when the problem became commercially important enough to justify it.

The Gap Between Elite and Grassroots

Below the professional level, none of this infrastructure exists. The 4.87 million cricket matches logged on CricHeroes each season are umpired by humans with no tracking support. The same is true for the overwhelming majority of football and tennis matches globally.

Cost is the primary barrier. A full Hawk-Eye deployment at a professional venue runs into six figures per season. Calibration and maintenance require dedicated technical staff. This is rational for a Champions League final. It is not rational for a district cricket league or a state-level under-18 football tournament.

The officiating quality gap between professional and grassroots sport is consequential. It shapes which decisions get disputed, which games get decided by errors, and whether athletes trust the process. It is also not inevitable.

Broadcast-grade precision is not what grassroots officiating requires. What is required is a system that can detect contact events, track basic ball trajectory, and generate a reviewable output when a decision is disputed. The hardware to do that at significantly lower cost is available. The software models are being built.

Where the Architecture Goes Next

The current commercial bracket of ball-tracking systems is designed around broadcast infrastructure. Smaller systems, built on fewer cameras and edge compute, are beginning to close the gap.

The near-term engineering challenges are monocular trajectory reconstruction (inferring 3D path from a single camera), contact detection in outdoor conditions with variable lighting, and on-device inference that removes network dependency for venues with poor connectivity.

None of these are fully solved. The error bounds at grassroots level will be wider than what Hawk-Eye achieves in a controlled stadium. But "wider than Hawk-Eye" can still be substantially better than a human official with a difficult viewing angle and no review mechanism.

The Premier League rollout demonstrated that tracking-based officiating works at scale and increases trust in decisions when the methodology is transparent. That proof point matters for adoption. The harder and more important challenge is building the version of it that a cricket board, a club, or a school sports program can actually run.