| Grant number: | 18/19007-9 |
| Support Opportunities: | Scholarships abroad - Research |
| Start date: | February 01, 2019 |
| End date: | January 31, 2020 |
| Field of knowledge: | Health Sciences - Physical Education |
| Principal Investigator: | Sergio Augusto Cunha |
| Grantee: | Sergio Augusto Cunha |
| Host Investigator: | Koen A.P.M. Lemmink |
| Host Institution: | Faculdade de Educação Física (FEF). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil |
| Institution abroad: | University of Groningen, Netherlands |
| Associated research grant: | 16/50250-1 - The secret of playing football: Brazil versus the Netherlands, AP.TEM |
Abstract Football is a very complex sport because it is a passing game between ten players (or eleven if the goalkeeper is included) to try to reach the opponent's goal or to maintain possession of the ball to guarantee the result previously achieved, while the opposing team must protect its goal and try to disarm the opponent. Even with complex personal interactions among twenty-two players on the field, what is now known is that physical, tactical, and technical factors is a challenge that sport and computer scientists have lately discovered using applied Machine Learning techniques to spatiotemporal variables of the analyzed team, the opposing team and the ball, in situations of possession of the ball (attack) and without possession of the ball (defence), when the ball is in play. So, for the interaction of physical-tactical-technical random variables, it is necessary that we have an accurate time-motion analysis system because velocity and acceleration vectors are the basis of physical actions, including technical time series (dribbling) in real tactics situations of the match. Furthermore, one of the most important skills in soccer is dribbling, the ability to take on opposition players and past them the ball quickly. By beating defenders, players can create space for teammates, provide scoring opportunities, and pull defenders out of their preferred positions. As a result, high-level dribbling is one of the most effective attacking skills in soccer, and top professional clubs seek to create and develop players who excel at it. The dribble can be performed by touching the ball with any permitted body part by rule, or even without touching the ball. Then, the aim of the research is looking for how we can find dribbling actions time series from velocity-acceleration time series in real tactical situations for each player on the field using Machine Learning approaches. The hypotheses are that Brazilian players dribble more than Dutch ones, mainly in the lateral parts of the field, and these actions are usually performed by external defenders and wingers. To confirm it, I will follow in loco FC Groningen matches using Inmotio system to compare, in future, with other Dutch teams and with Brazilian teams whose players' position will be also determined by the LPM system. (AU) | |
| News published in Agência FAPESP Newsletter about the scholarship: | |
| More itemsLess items | |
| TITULO | |
| Articles published in other media outlets ( ): | |
| More itemsLess items | |
| VEICULO: TITULO (DATA) | |
| VEICULO: TITULO (DATA) | |