4 minute read
Computer Vision looks to the future
Words by LUKA PATAKY
LUKA PATAKY, SVP, Automated Content at Sportradar, explains why the company’s AI-driven Computer Vision technology is the key to the next phase of growth for sportsbook operators.
Sportradar has been collecting sports data for more than 20 years. When we first started this journey, we were limited in the amount of data we were able to capture. Some of the earliest soccer matches we covered were during the 2002 / 2003 season where we captured just five data points per game: team, sport, country, tournament and start time. This information underpinned a basic prematch betting offering of win, lose, draw markets.
Back then we understood that sports data had vast, untapped potential and that providing fast, accurate and reliable data to our clients would be a game changer for the industry. What’s more, we knew that we had to invest in technology to increase the amount of data at our disposal and leverage that data in a way that engages fans and opens new revenue streams for our clients.
Today, we collect an estimated 2,750 data points across most soccer matches which we use to power more than 185 pre-match betting markets and more than 196 in-play betting markets per match.
It’s clear that sports data has fuelled the growth of our industry. It underpins all our products and services and is the key to engaging bettors deeply and meaningfully. But the big question is, how to further increase the amount of data available to us and use it to open new revenue generating opportunities?
The focus must be on technologydeveloping a technology that will allow us to collect more data of a higher quality, while enabling us to deliver it to clients worldwide in the blink of an eye.
At the forefront of this is Artificial Intelligence (AI). It is going to be a game changer for the sports betting industry and the tech is increasingly used within our products and services. Right now, our emphasis is on a specific subset of AI, which is Computer Vision.
For the uninitiated, Computer Vision teaches computers to ‘see’ and understand visual inputs such as digital images or videos. Or, to put it in simple terms, the technology mimics human vision capabilities.
And, while we as humans are very good at understanding what’s happening in front of us, following a sporting event for example, we’re limited in the amount of data we’re able to process, much less manually record and upload.
By comparison, Computer Vision is able to capture well over a 100-fold multiple in the level of statistics and data it collects, doing so in milliseconds, versus what can be done on a human level.
It’s the depth of data collected by Computer Vision from every video frame that’s going to unlock new products and services for our operator clients, helping them to engage bettors more deeply and open new revenue generating opportunities.
Sportradar has been nurturing its Computer Vision technology for the past four years. It began life as an innovation project and we’re now at a stage where we’re incorporating the technology into our core processes and products.
We’ve deployed Computer Vision within table tennis and scaled the solution across our entire portfolio of more than 150,000 table tennis matches. It’s a significant strategic move to enhance the quality of one of our most popular betting sports by turnover globally.
The speed at which Computer Vision collects and processes data allows for the creation of new, faster bet types. We’ve developed sophisticated microbetting live odds algorithms to deliver three innovative quick betting markets.
We call these ‘micro markets’, quick bets related to in-game events, rather than the final match result. We offer micro markets for every rally that takes place during a game, with a new rally starting approximately every 20-40 seconds.
The current microbetting offering within table tennis allows bettors to bet on the number of ball bounces in an upcoming rally, as well as whether the ball will hit the net.
The introduction of micro markets across our table tennis portfolio provides our operator clients with an additional 360 betting opportunities per match. Over the course of a year, we’re creating millions more opportunities for operators to engage with their customers. It’s huge.
In addition to the creation of micro markets, Computer Vision is an enabler to enhance the viewing experience. The technology turns every video frame into a digital asset, whereby we can transform what people see on their screen.
The Augmented Stream we offer for table tennis includes not just betting data but also performance data. This information engages fans for large periods of time, keeping them coming back to your platform, resulting in increased customer loyalty.
It’s a powerful retention tool and in the future will be able to enhance personalisation by showing different camera angles, as well as facilitate digital advertising and sponsorship integrations which all create further monetization opportunities.
Where we are now with Computer Vision is just the tip of the iceberg.
We’re expanding the technology into tennis, basketball and other sports. Looking to the future and there’s tremendous opportunity to be unlocked. Consider the application of this technology within the Metaverse or Virtual Reality.
Away from sports betting, consider how a rights holder might be able to monetise this technology. We have the means to provide fans with an avatar which has ‘real-world’ parameters. That allows us to create immersive sports experiences that put fans at the heart of the action, for example, taking part in a real virtualised tennis match or other type of sports competition.
I truly believe that Computer Vision, as a result of the data it identifies, insight it facilitates and experience it creates, has the power to transform our industry and generate significant commercial benefit for the whole sports ecosystem.