NFL Digital Athlete
Launch spot for the NFL's Digital Athlete program: game footage and CG player renders in one broadcast frame.
The player has a digital twin
Give the invisible work a body. Take the player you recognize, build a twin you can read, and make the jump between the two feel like one thought.
One team, eight disciplines.
Edit and VFX worked the shot together, so a new version of the live-to-twin moment could be tried and judged in minutes.
A player made of math
The NFL's Digital Athlete is a computer simulation of an NFL player, built with Amazon Web Services and announced in December 2019. It feeds on game video, player tracking, equipment, field surface and weather, then uses AI to make sense of how players move across a season. The league uses it in its player health and safety work.
LiveThe player you recognize
DataRead straight off the fieldIt is serious, useful work that almost nobody can see. We had thirty seconds to fix that.
No interface? Make one
There was no existing interface for the story, so we invented a visual language. Real players became CG counterparts, and angles, speeds and routes appeared around the action. Every sequence tied the football to the analysis.
The handoff is the whole trick
The digital players had to feel like the same athletes you had just been watching. Edit and VFX worked side by side to find the exact moment to step out of live action and into the twin, then hold it long enough to see what changed.
01Live action
02The footage dims
03The twin steps out
04Held long enough to readIf the movement felt wrong, nobody would buy the idea.
A complicated system, turned into a player, a movement and a new way to look at the play. Still football, with a little more going on underneath.
We gave the NFL's Digital Athlete program a visual language, connecting live football with CG players and animated analysis in a thirty-second film, from editorial through the final finish.
Background: Amazon press release · NFL.com