AI In Sports Media: The Art of Storytelling
The proliferation of digital consumption over recent years has shifted the balance-of-power away from sports media companies and into the hands of the viewers, quite literally. Sports fans, with a multitude of options at their fingertips across multiple devices in an increasingly fragmented landscape, are more demanding than ever. Sports broadcasters and media platforms have responded by investing in data – primarily in an attempt to gain a better understanding of their audience and to enhance its coverage so viewers are more engaged in the action.
In this context, Artificial Intelligence (AI) is becoming indispensable in the fierce fight for attention. “The broadcast and media space has been three or four years behind professional sports teams in this space,” says Stats Perform’s VP Data, AI and Pro Products Ben Mackriell, who adds that a failure to understand the tangible benefits of AI is one reason for the sluggish take-up in some quarters. “Finding a competitive advantage in the pro space has driven analytics forward,’’ Mackriell continues. “In the media space, it goes back to making something consumable, and in some cases the metrics you will see in sports coverage now has been the same for the past 10 or 15 years. “But AI is not about how big, impressive or complex your modelling is. It is about smart people with solid processes creating simple and engaging content.”
Storytelling
Media companies are already aware of the importance of creating a data-driven narrative around their sports coverage. As a platform for AI and data-driven insights via its Opta Sports service, Stats Perform’s consumer-facing digital platform, The Analyst, has provided a touchpoint for fans and media platforms seeking a fresh perspective on a sporting event.
Duncan Alexander, The Analyst’s Co-Editor-in-Chief and the Co-Founder of the @OptaJoe Twitter account -- which has 1.2 million followers for its statistical musings on soccer -- says, quite simply, that the key to making AI consumable for broadcasts is that “it needs to fit into a narrative”.
He adds: “We have a golden rule that 80% of the time you want to be backing up what fans think, with the remaining 20% being a counterpoint. There should be a sense of fun so it can be part of a conversation.” The appetite for such AI-based insights is clear. Even some mainstream broadcasters have now adopted the ‘XG’ metric – coined by The Analyst when the platform launched four years ago – used for calculating the likelihood of a goal being scored in soccer from a certain spot on the field. “It shows how far the understanding and acceptance has come since 2017,” Alexander says.
The Analyst also offers a test bed for new AI and data-driven models, with the aim of presenting a new model to users as often as once per month on average. Ahead of the start of the 2021-22 European soccer season, Alexander and his colleagues worked with Stats Perform’s developers to whittle down half a dozen ideas into three models to be ushered out through The Analyst.
‘Meta-fandom’
If anything, the pandemic has brought fans closer to their digital sports touchpoints, Alexander adds, describing a status of “meta-fandom” where “depth and speed of data is very important”.
This, in turn, leads to greater pressure on the platforms presenting the content. “Over the years we have developed a series of tools that broadcasters can use,” Alexander says. “We recently relaunched PressBox Live, which serves up information constantly and instantaneously during games, by using AI and automated insights in various ways.”
These statistical findings can create a story around the action on the screen. Mid-game analysis via different models are being increasingly used by media platforms and commentators to provide a sense of anticipation or a novel reflection on an incident.
Capturing attention
“There is more of a demand for getting into the coach’s head and capturing someone’s attention within a couple of seconds, but the content has to be consumable and quick,” Alexander adds. “It’s an exciting time. The last 18 months have shown how much potential there is.”
AI’s growing role at the heart of the art of storytelling in sports coverage appears to be inevitable. “Mobile devices and cord-cutting have changed the dynamic of how sport is watched,” says Dr. Patrick Lucey, Chief Scientist at sports data and analytics provider Stats Perform, pointing to the increasing number of digital enhancements that have become part of a standard sports viewing experience. “We still need to collect bigger and deeper data as more data will produce more insights, simulations and predictions.”
Such data can be used to provide the fan with greater control of their viewing experience – an approach that is increasingly desirable to viewers who are consuming coverage on their terms via over-the-top (OTT) platforms. “With OTT coverage, the viewer can control the experience, and AI can provide additional layers that will engage them,” Alexander adds. “This is something US platforms have done very well over the past couple of years.
“With AI, new experiences can be generated through content, visualizations and even camera angles, changing the way viewers interact with the coverage. If used correctly, AI can give you a lot of power.”