Sportradar Chief Product Officer – Computer Vision, Luka Pataky, examines the ways in which the firm’s Computer Vision solution can add multiple new dimensions to enrich the user experience.
Sports betting is constantly evolving. To keep up with the demands of increasingly tech-savvy and data-hungry customers, betting operators need to present more opportunities to engage with sports, offering exciting new products powered by deeper, faster and more accurate data, enabling them to attract new and retain existing customers.
A key part of this evolution is to also create a host of new commercial opportunities for both betting operators and rights holders. What if a new technology were available to help form new micro-betting markets, devise opportunities for more stats, facts, and insights to appear on screen to inform viewers while also offering more digital advertising space for commercial partnerships?
That’s where Computer Vision, a state-of-the-art AI technology that understands images and extracts information from them, comes in. As developed by Sportradar, Computer Vision is able to capture and process the deepest level of live sports data on the market, faster than ever before.
Computer Vision is capable of capturing a 100-fold and more increase in the level of statistics and data previously collected and the new technology can analyse up to 120 images per second, in real-time. Not only is this analysis executed at very low latency, but it is far more dependable and consistent in its accuracy, as it is free from human error.
This Computer Vision technology is initially being applied to Sportradar’s table tennis offering, as a first real live sport. Because of the controlled environment table tennis tournaments are held in, gameplay can be captured and analysed with one, fixed camera, so the sport is an ideal entry point. Table tennis is Sportradar’s fifth largest betting sport by turnover, and we’re planning to apply Computer Vision technology to the matches we cover later this year.
As a sport with extremely short rallies – an average of just 3.4 seconds, where the average stroke occurs every 1.7 seconds – table tennis presents particular challenges when looking to grow faster betting markets. With such quick movements dictating points, Sportradar’s Computer Vision technology is well placed to provide accurate, error-free data to operators in a time frame that creates more betting opportunities within a game.
Our technology team has worked tirelessly to train the AI to understand the rules of table tennis as well as implement more data touch points than ever before. Once we position the camera at the live table tennis game, the AI can detect the player’s movement, and identify a serve, a hit or a ball bounce to inform the betting data. It’s this near-immediate access to fast and accurate data that is benefitting our operators.
We are introducing micro-betting markets, a form of live betting that offers a super-quick turnaround on low-value bets and ‘gamifies’ the betting experience. Micro-betting provides instant action with a bet resolved in less than a minute on something happening in-game and not necessarily related to the final outcome. In table tennis this could be something as immediate as the type of stroke a player will next use or the number of bounces in a rally.
As odds and numbers change by the second in betting, technology’s rapid and accurate analysis, as well as automated trading capabilities, are vital to operators offering new products to their customers to stay ahead of the competition. The process of data collection to trading can be automated so we can build odds and markets extremely quickly. We are talking about 30-second or shorter cycles from placing the bet to getting the result. This type of betting is particularly interesting for newer generations of bettors with a shorter attention span.
But the opportunities don’t stop at betting. Computer Vision turns every video frame into a digital asset, streaming live performance insights like betting data on screen, and transforming what fans and bettors will see with live graphics, 3D replays, and virtual advertising. This gives the sports fan a far richer, more immersive, experience.
Through this sophisticated AI technology, we’re transforming the popular betting sport of table tennis and providing our operator clients with a significant increase of the amount of data they need to develop new products and services for bettors.
There’s untold potential for Computer Vision, and its application within table tennis is just the beginning as Sportradar looks to integrate the technology across its portfolio of sports and change the way fans watch, engage with and bet on sports.