Sportradar DeepKinetic Foundation Model

Sportradar

Vote

The 7th Annual Awards

Nominee ✨

• Best Use of Technology

Share

ABOUT THIS ENTRY

Sportradar advanced its AI-driven sports intelligence capabilities with the launch of DeepKinetic, a foundation model for basketball and a first-of-its-kind application of AI in sports. Built using player tracking data from thousands of NBA games, DeepKinetic delivers a deeper real-time understanding of player movement, decision-making and game action, unlocking new possibilities for AI-powered broadcast products, and next-generation coaching and scouting insights. 

How does this represent "Excellence in Engagement"?

Sportradar’s DeepKinetic sets a new benchmark for fan engagement by transforming how basketball is understood and experienced in real time. By applying a first-of-its-kind generative AI foundation model to billions of 3D body-pose data points from NBA player tracking, DeepKinetic can model player movement and decision-making to generate live, probabilistic insights about how plays develop and what may happen next. 

For fans, this unlocks a deeper and more intuitive understanding of the game. Instead of simply showing what happened, DeepKinetic enables broadcasters and media partners to explain why a play unfolded the way it did and how the next moment might develop, turning complex data into compelling, easy-to-understand storytelling during live games. This moves fan engagement beyond traditional statistics and post-play analysis toward predictive, context-aware insights that enhance the viewing experience in real time. 

By making advanced AI insights accessible and meaningful to fans, DeepKinetic establishes a new standard for interactive, insight-driven sports experiences. The model also creates a scalable foundation for future innovations across media, performance analysis, and interactive products, demonstrating how artificial intelligence can deepen fan connection to the game while setting a new industry benchmark for data-driven engagement in professional sports. 

Objective

The objective behind the development of Sportradar DeepKinetic was to create an AI foundation model capable of learning how players move on the court and modeling basketball action with unprecedented precision, unlocking predictive insights that deepen understanding of the game. 

DeepKinetic was designed to deliver tangible impact across the basketball ecosystem, from coaching and scouting to media and broadcasting to betting, while representing a breakthrough in how artificial intelligence understands, models, and predicts human movement and decision-making in real time. Building on this foundation, Sportradar plans to expand the technology to additional sports, including soccer and tennis. 

Strategy & Execution

Powered by official NBA player tracking data, Sportradar’s DeepKinetic was designed to translate complex on-court movement into insights that resonate across the modern basketball ecosystem, from coaches, scouts and broadcasters to fans seeking a deeper understanding of the game. DeepKinetic will bring fans closer to the action, enabling them to gain a deeper understanding of what’s unfolding on the court in real time like never before. 

At the league level, DeepKinetic enhances coaching, scouting, and performance analysis by modeling how plays develop and identifying patterns in player movement and decision-making. For broadcast and media partners, the technology provides real-time probabilities and advanced visualizations that help explain why plays unfold as they do and what might happen next, turning complex data into clear, intuitive storytelling for fans. 

Over time, DeepKinetic will unlock always-on predictive insights, more immersive fan experiences, and new commercial opportunities. The technology will power Sportradar’s next generation of products, including advanced coaching and scouting analytics, simulated betting experiences, and enhanced broadcast visualizations, bridging the competitive and entertainment sides of the game while helping fans engage more deeply with basketball. 

Organizations

• Sportradar

Links

•   Website

Credits

Carsten Koerl
Founder & CEO
Sportradar

Behshad Behzadi
Chief Product, Technology and AI Officer

Nicolo D'Ercole
EVP for Technology and AI at Sportradar

Celebrate in NYC

Join us for the 7th Annual Hashtag Sports Awards™ on the evening of June 24, 2025 as we celebrate finalists and reveal the winners of this year's awards live during the show.