Next Gen Stats is the NFL’s player-tracking and analytics program, run with three partners: Zebra Technologies for the location hardware, Wilson for the instrumented ball, and Amazon Web Services as the cloud and machine-learning provider. Zebra’s player-tracking partnership began in 2014, and AWS has been the official technology provider since 2017.
The tracking layer is radio, not cameras. Each stadium carries 20 to 30 ultra-wide-band receivers, and every player wears two or three RFID tags in their shoulder pads; tags also sit in the ball, on officials, pylons, sticks, and chains. The tags report position about 10 times per second, charting movement to within inches and producing more than 200 new data points on every play. On top of that raw stream, AWS scientists build machine-learning models - pass completion probability, expected yards after catch, and defender movement prediction among them - using tools such as Amazon SageMaker, and store nearly 300 million data points per season.
Next Gen Stats turned American football, long described in coaches’ anecdotes, into a dense spatial dataset. It is the data foundation for broadcast graphics, the league’s annual Big Data Bowl analytics competition, and a growing layer of predictive models on top of the tracking feed.