The English Premier League is adopting a new offside detection technology developed by Genius Sports, which will use iPhones and the iOS’ machine learning to improve offside violation accuracy. The new implementation is called Semi-Assisted Offside Technology (SAOT), and it’s supposed to address the issues with the VAR (Video Assistant Referee) system that have encountered delays and inaccuracies over the recent league seasons.
SAOT will employ 24 to 28 iPhone 15 Pros positioned around the field to capture high-resolution video at 100fps but tests have been recorded at 200fps as well. The high capture rate of the iPhones’ cameras and local processing capabilities contribute to the system’s ability to record and analyse detailed player movements, even under various lighting conditions.
This setup will provide between 7,000 and 10,000 data points, allowing the system to create a 3D virtual model of each player and accurately determine their positions relative to the offside line. Genius Sports’ “GeniusIQ” processing system will collect data to track individual body parts and predict positions when players are obscured from the referee’s view.
Even though performance comparisons with the existing VAR system have not been made, Genius Sports’ technology is expected to offer a more aggressive way of tracking players than traditional methods, which often use fewer data points and simpler representations of players on the pitch.
The Premier League will fully implement Genius Sports’ SAOT system by the end of this year, with ongoing usage throughout the 24/25 season. The effectiveness of the system’s performance will be evaluated closely by the league’s association.
(Source: Premier League, Genius Sports, The Verge, Wired)
