Artificial Intelligence (AI) is creeping into every walk of life these days and the world of sport is no different, with data and the use of science behind the numbers becoming more prevalent. But that is not just at the elite level.
Could AI help uncover the next superstars in South Africa’s talent-rich Gauteng province by analyzing and exposing the most promising players in the region?
Swiss tech company Talnets AG have partnered with the Gauteng Development League (GDL) to implement their technology to provide young players with the opportunity not only to be seen by local clubs but also win a contract overseas.
It follows a successful one-year trial period where the technology features the youth teams of Mamelodi Sundowns, Kaizer Chiefs and Orlando Pirates, the three biggest clubs in South Africa.
Talnets AG’s technology uses AI to analyze a match based on over 60 different data points from video of GDL matches and turn this into useable numbers to create a detailed profile for each player over time.
Loading...
This ‘player card’ becomes a CV for the player which lives online and is publicly shareable to scouts, coaches and clubs around the world.
But numbers never tell the whole story and the platform also creates short video highlights of the player that is assigned to their card and gives scouts the opportunity to visually see their actions as well as read their numbers.
Fungai Mapondera, originally from Zimbabwe but now based in Switzerland as CMO with Talnets AG, explains the concept to FORBES AFRICA.
“It is about extracting the data from the match video and assigning the different data points based on what the player is doing in the game,” Mapondera says. “In essence, you end up having a digital player card with all your data from all your games, plus the highlights from all those games, in one place.
“It makes it easy for you as the player to share that profile with anybody that may be scouting you.”
There have already been successes. One player was scouted using the system and has had trials in Spain. And there are many others who are set to have their talent exposed to the world through the platform.
Behind it all is a meeting between technology and sport, something that has been rapidly accelerating in recent years, and is set to develop even further into the future as the cost of such platforms come down and therefore become more widely available.
Pin-pointing exactly what makes for useable and readable data has been important.
“There are so many things you can analyze in a game,” says Talnets’ Chief Executive Officer Darko Stanoevski. “Our focus is on individual performance analysis, which requires a bit more of a specific approach.
“That framework could be simple things like goals and goalkeeper saves, to more intricate detail like the different kinds of passes, passive passes, passes that break the (defensive) line and so on.
“Grounded on the FIFA Football Language Model and built with top football experts, we have developed our framework with over 60 different data points to help us spot and identify young talents in an efficient, transparent and scalable way. Analyzing a full game based on this framework would give us around 600 data points for one team, and then it will be, on average, around 15-20 data points per player.”
This data collection used to be done manually by an analyst, taking days or even weeks. But that has been now cut down to a few hours.
“The AI is an enabler to lower time dramatically, to automate this entire analysis process and make it more scalable and more efficient,” Stanoevski says.
“It gets players visibility and access to clubs, to scouts and coaches that have access to the platform, and even more importantly, enabling the players themselves to share their profiles with anyone, so they can build a ‘CV’ and pursue their dream.”
Loading...