Talent discovery with artificial intelligence: the YAFA model

This work concerns the integration of artificial intelligence into talent screening in sport in Turkey. Through a new model developed by the Artificial Intelligence Idea Academy (YAFA), the aim is to determine the sporting aptitude of children aged 5 to 12 through scientific tests and measurements.

YAFA processes multi-dimensional data with AI-assisted algorithms: branch-specific technical skill tests, visual and auditory reaction measurements, anthropometric and motor analyses. The purpose is to make athlete selection more objective and fair. The system was first applied in the province of Şırnak. It stands out as the first AI-based athlete screening model in Turkey developed with domestic algorithms.

Discovering ability early matters for the development of sport and for widening the pool of athletes in a serious way. The limits of traditional methods made it unavoidable to bring technology into this field.

The model analyses several data sets through artificial intelligence, defined by experienced coaches for each age group. Branch-specific technical skill tests apply drills of varying difficulty set for each sport. Visual and auditory reaction measurements record how quickly children respond to environmental stimuli. Anthropometric measurements assess height, weight, and arm and leg length. Performance tracking follows the development of endurance, speed and balance over time.

All of this data is analysed with image processing and AI algorithms, and the sport each child is most suited to is determined objectively.

The Şırnak pilot: the first step of the project was carried out in Şırnak in cooperation with the International Talent Scouts Football Academy (UYAFA). In screenings held at Şırnak City Stadium from 11 to 15 November, roughly 5,000 children aged 5 to 12 went through preliminary selection; 128 of them were found suitable for the data collection stage. The process was coordinated by Bahtiyar Can Vanlı, an experienced coach who has worked with the Turkish national team.

The data gathered is compared not only within Turkey but with athletes of the same age group abroad, which allows a child’s potential to be measured against an international level. The findings support both individual development tracking and the building of a scientifically grounded selection system for clubs.

The system also aims at equality of opportunity by reaching children in rural areas, and at removing the favouritism and subjective selection of traditional club-based scouting.

Developed in consultation with the Ministry of Youth and Sports and the Turkish Football Federation, the model is planned to spread across the country. Cooperation with clubs, companies and angel investors at home and abroad supports its continuation. In this sense the model is not only a screening system: it is a whole transformation project that brings a scientific, AI-based approach to how Turkey discovers athletes.