Ed and Andrew Frazelle, a father and son team, who are the owners of RightChain, advanced supply chain optimization, planning, and analytics software company based in Atlanta. At the same time, Frazelles are tennis lovers and were intrigued to see if their planning concepts could be applied to the sport.
Ed Frazelle contacted Craig O’shannessy who runs match analysis company Brain Game Tennis, who has been working with Djokovic, among other pros, since 2017. As Loudin points out, “he analyzes their patterns of play and helps them understand both how to improve their own performance as well as which strategies they should employ against specific opponents.”
O’Shounessy’s partner in his work is Warren Pretorius, the CEO of Tennis Analytics, “who developed a model of video analysis that utilizes manual tagging, which he pioneered in 2013.” His method is to chart matches over 25 key indicators and then “combines data analysis and visualization to extract match information and generate keywords on indexed video.”
Frazelle says that he met with O’Shounessy and Pretorius at Wimbledon and that, “we literally started running data that night.” It turned out that RightChain’s AI apps help companies Colgate, Caterpillar, Ford, Baxter, and Coca-Cola simplify their supply chains by breaking the process down into 25 components. Loudin gives an example where forecasting utilizes “an A.I.-based algorithm to craft and continually update a unique model for each product. Network optimization uses an algorithm that determines where to place distribution centers based on a multitude of user-defined criteria.”
To apply his methodology to tennis, Frazelle decided to break down a tennis ball’s journey from end to end in a similar way. As he explains, “For tennis, we changed the fields to focus on the destination and origin of the ball. It’s a very formal coordinate system that maps the tennis court to a level of detail not previously available.” (In this case, each service area is divided into 12 sub-zones, and the backcourt is divided into eight such zones.)
Analytics of just the tennis play by itself is quite one dimensional, and as O’Shounessy explained, the A.I. can find repeatable patterns, measure rally lengths, and determine where precisely a player hit a ball. “The technology offers us extra layers and patterns for a more detailed analysis. It’s one thing to tell a player what’s happening, and another to show them with tables and graphs. The graphs that Ed provides cut the data up to multiple ways and easily lead our eyes to where the real keys of winning live.”
O’Shannessy also said that one of his toughest sells to players has been convincing them that consistency of play — the long rallies that occur in practice — “was overrated, something that video analytics can’t quite prove but A.I. can.”For his part, Pretorius added that “Instead of looking at data in isolation, with A.I., they now can get the story of their play evolution.”
In the end, Novak Djokovic won the 2019 Wimbledon tournament, with O’Shannessy adding that the use of AI is “just the start of where the technology can take the sport.”
- The Black Box Problem in LLMs: Challenges and Emerging Solutions
- Alex Ratner, CEO & Co-Founder of Snorkel AI – Interview Series
- Circleboom Review: The Best AI-Powered Social Media Tool?
- Stable Video Diffusion: Latent Video Diffusion Models to Large Datasets
- Donny White, CEO & Co-Founder of Satisfi Labs – Interview Series