Case Study: Statsyuk delivers pro-grade hockey player tracking with Roboflow

Statsyuk turns single-camera hockey footage into player tracking with Roboflow in hours

Statsyuk, a sports analytics company, aimed to make professional-grade hockey player tracking accessible to all teams, a feat that previously required expensive multi-camera rigs. Their challenge was to use computer vision to turn footage from a single moving camera into structured data, accurately tracking players even when they were heavily occluded or bunched together during fast-paced gameplay. To accomplish this, they turned to Roboflow and used its RF-DETR object detection model.

By implementing Roboflow's RF-DETR, Statsyuk built a highly accurate and efficient data-processing pipeline. The solution provided reliable player detection through scrums and netting, which was crucial for downstream tracking and analytics. Using Roboflow also helped them structure clean datasets and fine-tune models with limited data. The result is that Statsyuk can now provide detailed, shift-by-shift data to coaches, players, and scouts, bringing pro-level analytics to junior, collegiate, and minor league teams that previously had to rely on manual tracking or basic box scores.

Statsyuk