The IE Research Datalab seminar series continued with an insightful presentation by Dae-Jin Lee, entitled “From modeling human growth curves to talent detection in team sports.” Professor Lee used data science to challenge traditional talent scouting, highlighting the overlooked potential of athletes like soccer legend Messi, whose late physical development did not match his talent nor predicted future greatness.
Dae-Jin Lee emphasized the importance of integrating mathematical modelling tools with an understanding of biological growth for precise talent identification. By introducing concepts such as bio-banding, the process of grouping athletes based on growth and maturation rather than chronological age, professor Lee showcased how data and modelling can revolutionize scouting and the analysis of growth related injuries in young athletes. This innovative approach aims not only to level the playing field for late bloomers but also to refine the scouting process by factoring in athletes’ physiological development stages.
A highlight of the presentation was Dae-Jin Lee’s collaboration with a professional team, presenting a unique dataset tracking growth and injury records over 20 years. The study offers new insights into growth patterns and their impact on injury likelihood, marking a pioneering approach in sports science.
The seminar sparked engaging discussions on the intersection of sports science and data analytics, reflecting the series’ aim to foster interdisciplinary collaboration. Professor Lee’s presentation underscored the potential for innovative methodologies in athlete development and scouting, setting the stage for future explorations in the seminar series.
Related articles:
- “Can We Really Predict Injuries in Team Sports?”, Lee, D.J. and Zumeta-Olaskoaga, L. Boletín de la Sociedad Española de Estadística e Investigación Operativa (BEIO) (2022). https://www.seio.es/beio/can-we-really-predict-injuries-in-team-sports/
- “Prediction of sports injuries in football: a recurrent time-to-event approach using regularized Cox models”. Zumeta-Olaskoaga, L., Lee, D.J., et al. (2021). https://link.springer.com/article/10.1007/s10182-021-00428-2