Global sports technology firm Sportradar Group AG has created a data-focused project that works as a “foundational model for basketball”, says Behshad Behzadi (pictured), chief product, technology and artificial intelligence (AI) officer at Sportradar Group AG, in an interview with GGRAsia.
Mr Behzadi, a computer scientist, was a co-founder of Google Assistant, a voice-activated virtual assistant for digital devices, and is nowadays applying his know-how at Sportradar.
Sportradar, founded in 2001, describes itself as a global sports technology company, providing data and related services to sports federations, news media, consumer platforms and sports betting operators.
In terms of using a predictive approach in Sportradar’s case, the firm’s model is “for all the basketball movements” of players in a game.
This means that “every second” within a game, the firm has mapped “29 body points for all the players”, he noted.
The company has been doing monitoring of play for five years, and analyses currently “100,000 basketball games a year,” and “not just NBA [National Basketball Association] games”, he stated.
Potential uses of the data include for basketball coaching, and for predictions of game action.
Sportradar utilises AI analysis for a product it calls Synergy Coaching & Scouting. The technology format – used by the NBA, WNBA (Women’s National Basketball Association) and NCAA (National Collegiate Athletic Association) basketball teams – is now available for high school coaching.
It includes “identifying” players even in “high school and college games” where their “hand moves or trend” of play is “very promising”, Mr Behzadi noted.
Mr Behzadi said: “Synergy is our coaching product, and there we use it for scouting … finding players”, he stated, referring to commercialisation of the technology.
The mapping data can also assist with “predictive analysis”. The scientist stated: “There is something we call ‘EPV’, which is ‘expected possession value’. For example, with a player in a certain position on the court, ‘how likely are they to score points, and if so, how many?’.”
The technology can help in identifying strategy, and what might happen in a certain scenario: “what is a turning point in a game, what made a team lose, what set off those events,” he explained.
Sportradar is working with broadcaster NBC in the United States, on “augmenting the experience of the game” for the basketball-viewing public.
Via an “enhanced mode” the broadcaster has for its viewers, Sportradar can offer information relating to “player tracking” and “animations”. The EPV data is also shown in enhanced mode, said Mr Behzadi.
That chimes with the “newer generations’ expectation of the experience” of watching an NBA game, he added.
Bet Concierge, and Bettor Sense
Sportradar also has its own freshly-launched consumer-facing conversational AI for its bookmaking operators, called Bet Concierge. It is described by the brand as “a match assistant using AI to provide match info, smart betting suggestions, and statistical answers, with real-time responses”.
Mr Behzadi told GGRAsia: “You can ask questions such as ‘How often in the last five games’ a certain team or player ‘scored in the last five minutes’ of a game, ‘how many points usually they make… does the lead change often?’ and so on.”
He suggested that without such AI, “it is very hard” sometimes even “impossible” on a betting website, if a consumer “wants to combine two, three different bets”.
With Bet Concierge “you can just say ‘Here’s what I want to bet’; it’s much easier,” suggested Mr Behzadi.
Asked whether an AI assistant for bookmakers posed any ethical risks in terms of gambling behaviour by consumers, the scientist stated ethical risks were “not specific to AI agents or chat bots”.
He noted a bookmaker could show on its website “different fonts or colours and then attract people to something”.
Though he stressed: “At Sportradar, we always think that for our customers [bookmakers], the aim is fan engagement and the fun and experience.”
That, he said, is “a long-term thing”. You don’t want the gambling clients to “lose much or to get addicted,” he added.
Within the company, Sportradar has a “big group working on integrity” topics, said Mr Behzadi.
He stated: “The whole goal is … preventing on one side fraud, but on the other side, preventing betting-related problems.”
The brand has a product for bookmakers that it calls Bettor Sense. It is powered by AI, and promises according to the company’s website, “early and accurate detection of at-risk customers”.
That involves “trying to identify” vulnerable consumers, to enable necessary interventions “at the right time,” observed Mr Behzadi.