Introduction

AI in
Sports Broadcast

Authored by

Sportradar is the playmaker of the sports industry, with a vision that brings together all elements of the sports economy. And the brain behind that vision is artificial intelligence (AI).

Teams, tech platforms, broadcasters, brands, and betting operators all rely on Sportradar’s AI-backed products and services, leveraging the unique breadth and depth of its sports-focused data.

By doing so, they can draw actionable insights in real time, both into the action and into the fans who watch it, to improve decision-making and consumer engagement.

Sportradar’s data, drawn from more than 20 years as the world’s leading sports technology firm, provide the raw inputs into the most advanced AI techniques in the sector.

This data advantage is what attracted Behshad Behzadi to join Sportradar this year from Google, where he had led its AI- focused Google Assistant, Google Lens and Google Smart Display projects. Here, he explains what AI will mean for the sports sector.

sportradar.com

Scroll down to continue

Sportradar’s 4Sight, the company’s AI-enriched streaming technology

Hitting, kicking, throwing or running with balls, pucks and shuttlecocks – little has changed since team sports first formed leagues more than 100 years ago.

But unlike the games themselves, how sports matches are consumed and engaged with has been transformed over the decades. Where once it was just the spectators on the sidelines, then radio and network, cable or satellite TV, now streaming is beginning to take over.

This new medium opens a new realm of AI opportunities, which can in turn spark a revolution for fan engagement. It will broaden and deepen the connection between sports and their fans, while opening access to ever more sports and events.

US viewers recently saw how AI-produced audiovisual content is already changing the game. There was a glimpse of it in the coverage of the Paris 2024 Olympic Games from the subscription broadcaster, Peacock.

Viewers could produce their own 10-minute Olympic highlights packages, choosing whichever sports they wanted, all knitted seamlessly together by AI editing.

Even the commentary was delivered by generative AI, using a digital rendition of the famous American broadcaster, Al Michaels’, voice. “The fake Al Michaels is surprisingly good in Olympics highlights,” read a headline in the Washington Post.

If that is the future of broadcasting, then AI will soon bring more consumer choice to more sports and more levels of competition. This will be transformational for how fans consume sports.

Much focus on AI’s use within sport remains on the increased depth and number of data points that can be captured through the deployment of this technology, like optical skeletal tracking data, radio frequency identification data (RFID) and even biometric data. And while the integration of deeper and real-time data is important, the real game changer is AI’s ability to process these raw coordinates to feed more possibilities for fan engagement, monetisation and drive better understanding of sports.

Sportradar’s NBA Advanced Visualisations

Sportradar has been breaking ground like this, investing in technologies that can contextualise and productise the data to create novel and meaningful experiences. One such example is our application of Computer Vision (CV) within niche table-tennis tournaments.

By leveraging CV technology, we’re opening the door to new advertising, betting and revenue possibilities. With the power of AI, CV can process 120 frames per second to accurately score the action even faster than the chair umpire and convey the live action to markets in real time.

With simulated commentary like Peacock’s from Paris added in, AI can curate and monetize the action even for the least visible sports. This will expand the sports economy into what has previously been uncharted territory for the general consumer, also enabling rightsholders to realise the potential of their sports data.

But fans, too, can similarly benefit from improved insights, served in a way that is appropriate to their level of expertise in any sport. Viewers of the same sport will have different levels of understanding of the action. AI acts as your expert sports buddy, conceptually watching the games with you and embedding contextual explanations of the rules or deep tactical analysis into individual streams.

It can confer the encyclopaedic knowledge that comes from having watched all the games ever recorded in that sport, and read all the books, blogs and social-media posts about it. With this deeper understanding of the action, the viewer’s engagement with and attachment to the sport can grow, accelerated by generative AI.

Young audiences are especially receptive, with 18 to 29 year-old fans likelier to use mobiles and tablets to consume sports through OTT streaming than traditional broadcast media. Modern fans crave real-time and bitesize content, and the industry is evolving to keep pace with these new consumption patterns.

These new methods permit more interaction with the viewer. And in a world where the battle for consumers’ attention grows hotter seemingly by the day, the ability to understand them and speak to them individually is a potent weapon.

This is now possible by using AI-backed sports technology. Leagues and teams, retailers, ticketing and betting companies, even sports broadcasters, can better understand the fans with whom they are engaging.

Behshad Behzadi, CTO and CAIO, Sportradar

This means receiving deep, data-driven insights on every single connected fan, for whom content can be hyper-personalized – optimised for the individual across media, betting, ticketing, marketing and a host of other mediums.

Our FanID marketing service powers such personalisation for rightsholders and brands, providing the first sports-industry-specific data clean room to achieve a deeper understanding of sports audiences. It creates individual customer profiles, targeting fans with customised adverts and their own, tailored feed of actionable narratives.

Generative AI is also transforming the content itself. Already, data visualizations incorporate real-time, animated overlays into the live sports stream. Some streams are incorporating more gamified elements such as leaderboards, challenges and rewards that make the viewing experience more interactive and immersive. These are supported by AI chatbots and virtual assistants, bringing data insights and user-relevant storytelling and interactions.

At Sportradar, we’re already applying elements of this technology through emBET, a streaming solution that integrates sports betting functionality and live data information. Insights, polls, voting and quizzes, are directly fed into any OTT streaming platform.

But there is much scope for deepening the influence of AI too. As the incipient market matures for augmented and virtual reality (AR and VR) hardware such as Apple’s Vision Pro and the Meta Quest, it will open the way to more creative and immersive sports-focused content.

This might mean watching a game from a specific seat or angle, or even from an individual player’s perspective. It could mean playing simulations against an avatar of a favourite player or athlete.

Social-media integrations, based on AI analysis of trends and fan sentiment, will allow sports organizations to connect with their audience still further. And with real-time translations, captions and subtitles or audio descriptions, accessibility can be improved for a global audience and for fans with disabilities.

None of this is possible unless the AI is trained on diverse and reliable datasets. Now, thanks to those datasets, how we consume sports will be unrecognisable to what we know from the past 100 years – even though the games themselves will remain much the same.

Introduction

AI in Sports Broadcast

AI in
Sports Broadcast