Sportradar and Amazon have successfully developed a computer-vision-based Soccer Goal Predictor, calling soccer goals before they have happened.
Sportradar and Amazon Build the Future of Sport Engagement
Sportradar, a global provider of sports betting and sports entertainment products and services, and technological giant Amazon have embarked on an exciting new venture, leveraging big data and machine learning (ML) to create a new soccer goal predicting algorithm. Once launched, it will allow fans to enjoy an even better engagement and immersion with the sport they love.
Commenting on this milestone in a blog post, Amazon ML Solutions Lab, Amazon’s machine learning offshoot, has confirmed the successful development of a solution the company describes as a “computer vision-based Soccer Goal Predictor.”
The solution analyzes soccer games visually and flags situations that are more likely to lead to a goal in near real-time, with a nominal millisecond delay. The predictor should boost fan engagement and allow broadcasters to deliver on an enhanced experience, Amazon has said.
According to the blog post, Soccer Goal Predictor will end up calling a goal two seconds before a score has been realized. Commenting on the breakthrough, Sportradar Group CTO Ben Burdsall said that Sportradar and Amazon ML Solutions Lab deliberately chose to tackle a computer vision problem, as it was both challenging and highly-rewarding to consumers and the industry.
To power up the solution, the team tapped into Amazon SageMaker which eventually made it possible to track matches. Thanks to the achievement, Sportradar and Amazon expect to have numerous new business opportunities.
“After today, I have no more skepticism about the potential of computer vision in innovating our business,” Burdsall added.
Delivering New Opportunities for Business
Sportradar and Amazon’s project seems not only promising, but a true game-changer for the sports entertainment industry and in particular, the in-play segment. According to the companies, Amazon ML Solutions Lab’s Soccer Goal Predictor is capable of calling a future goal with an estimated 75% accuracy.
To power this, the company uses an I3D model in Amazon SageMaker which can analyze an individual attack and determine whether it would lead to goals or not. This will automatically allow broadcasters to signal when audiences should be paying a close eye to the game.
Soccer is just a start for the companies, though. If successful, the model can be re-purposed to carry out the same vision-based analytical work for other sports, opening up new venues of possibilities for fans, businesses and sports at a time when sports betting is expanding at a rapid pace.
Sportradar has an extensive portfolio of sports partners that goes well beyond the soccer bastion, but most recently, the company added the Indian Super League (ISL). In November, the company announced that, moving forward, it would be providing due diligence solutions to US customers through its Intelligence & Investigation Services (I&I) platform.