In yet another interesting session on the application of artificial intelligence (AI) at The Rising 2021, Megha Gambhir, Co-founder, CEO at Stupa Sports Analytics, discussed the use of AI and analytics in the sports industry, particularly in racket sports.
The global sports analytics market is currently at around $800 billion and is forecasted to touch $6 billion market cap by 2025. Stupa Sports Analytics works on AI-enabled ball tracking and video analytical technology from low-end devices.
Gambhir spoke about how data analytics and AI/ML models are revolutionising the sports industry. “Analytics is already being used immensely in Basketball, Rugby and Cricket. Most other sports are still in a nascent stage of technological advancement. Racket sports is one of them. However, many players and coaches in racket sports are today coming forward to use tech to enhance their performance,” she said.
Use cases of data analytics and AI in racket sports.
- Enhancing the performance of players
- Creating probabilities and predictions for future games
- Real-time graphic overlays during streaming
- Creation of gamification zones for amateurs
“Before a decade, the collecting of data was by excel or by writing down on pen and paper. And, it takes a lot of time to consume the video and create data out of it. This one single activity is being made easier by AI, using computer vision cameras, or IoT devices or wearables,” she said.
The second step is to perform data analytics and extract trends and patterns. “So once we have the data, we can churn it, extrapolate it and then model it with AI, ML, and use data science on top of it and derive a lot of trends and patterns, predictions and deeper analysis which can be then be used,” she said.
- You can track the ball trajectory and ball speed
- You can use data and create a VR around it
- You can get game-specific data from your matches and from your practice
- You can use these data to extrapolate the information as per the KPIs needed
“Many of the matches which you have been analysing or players have been analysing on a day to day basis, becomes a solid historic data over a period of time. Using a lot of AI/ML algorithms you can now derive the probabilities and predictions. For example, you can know if your opponent is hitting from a forehand on maybe a 5th point of the game,” she said.
Next is broadcasting and streaming, where one can use AI/ML-driven insights to tell the audience about exact ball placement, ball speed etc, do real-time graphic overlays during streaming and know exactly what replays should be played according to the audience’ likings.
Gambhir said virtual reality is another area that has recently come up. “Virtual experiences created for people sitting at home and watching are going to be huge. OTT platforms will leverage this to the maximum,” she said.
Future of sports analytics
At the moment, most data analytics platforms are using high-speed cameras. “And that’s a constraint for the market as well. When you require to scale up or you need scalability, you have to be economical as well as it should be easy-to-use for people,” she said.
Gambhir feels the next upcoming thing in sports analytics is the use of mobile and low-end devices This, she feels, will bring scalability for a wider audience or audience at the grassroots level.
“By using AI or VR or AI/ML, you can create a lot of gamification zones at malls and clubs for amateurs to try their hands and know exactly the speed they are playing with or how much consistent they are,” she said