Never Done Evolving is more than a tribute and more than a study in the form of avatars. It's the most accurate and detailed virtualization ever done with an athlete. An approach that will certainly help Nike find and train the next Greatest of All Times.

By analyzing all data and footage from official tournaments throughout Serena's career, machine learning modeled each era’s playing style. Then, to make this study's insights even more inspiring and tangible for everyone in the world, we transformed it into a live virtual match, in which Serena from 1999 faced Serena from 2017.

On August 9 2022, Serena Williams announced her retirement from tennis. As a brand that made part of most of her career, Nike decided to launch a tribute in the form of a study, showcasing what it takes to stay on top for over two decades.

Our Goal was to reveal the evolution needed to become the greatest of all times, in order to honor the former tennis player and inspire younger generations.

Two days after Serena Williams announced her retirement, Nike released a year-long study, in the form of avatars, showcasing her evolution to stay on top for over 2 decades. Analyzing all data and footage from official tournaments throughout Serena's career, machine learning was able to model each era's playing style. To make it accessible and inspiring to everyone in the world, we turned the study's insights into a live virtual match on YouTube. Starring 17 year old Serena, the age she won her first Grand Slam. Against 35 year old Serena, the age she won her 23rd Grand Slam. The game was the result of 130.000 games generated using Vid2Player technique, developed by Stanford University.

On August 9 2022, Serena Williams announced her retirement from tennis. 2 days later, on August 11, Nike released a study kept in secret for years, just waiting for this moment.

By analyzing all data and footage from official tournaments throughout Serena's career, machine learning was able to model each era’s playing style, Decision making, shot selection, reactivity, recovery and agility. Creating the most accurate virtualization ever done with an athlete.

Then we applied this data into the Vid2Player technique, developed by Stanford University, which generated over 130.000 games in which Serena from 1999 played against Serena from 2017.

The study's results were so mind blowing, we decided to make it accessible for everyone in the world, by turning it into a live virtual match on YouTube.

The live virtual match saw an increase of 1082% in organic views compared to other Nike content. It broke all Nike's organic views records on YouTube. And more than that, Nike is now applying this study's data not only to improve their products' technology, but also to improve their athletes' performance across sports.