Design a data driven model for eva‎luating player and team performance in soccer (case study on one of the soccer match)

9/19/2023 12:00:00 AM

There is a lot of money going around in the world of sports. Among sports, football is the first in terms of fans and financial institutions. A lot of data is generated in every football match, also outside the game flow in the player market, and also data role management decisions play an important role. In today's world, both data processing tools have made significant progress and the tools or methods used to deal with these data are multi-faceted and multi-level. The footprint of data science and data analysis is visible in football clubs. In the analysis room of the clubs, data on the position and events of the past games of the home team and the heavily monitored opponent team are used to extract weak points and strong points. In this research, relying on data-driven technology to value the actions of the players and their movement in the playground, a model was presented that can monitor the actions of the players during the match. Much of the current research in the world of football has used data that does not provide the stakeholders with much knowledge of the depth of this data. Now, with the advancement of technology and the use of sensors that record player performance during the match, analysts have provided data that can extract deep information from this type of data. A model that can be compatible with the dynamic nature of football and the complexity of the football game and provide a good view of the game will be among the priorities of analysts. A new change happened in this model: the application of physics and the player's ability during the game. As we know, many factors during the game affect the performance of the players and cause the players to drop during the match. As we can see in the next parts of this research, by applying this change, the condition of the model is improved and the model error is reduced from 0.012 to 0.01.

تحليل عملكرد , ارزش گل مورد توقع , داده فضايي-مكاني , داده موقعيت بازيكنان

performance analysis , Expected goal value , Spatio_Temporal , Tracking Data

Amin Gholamrezaei

Dr. Noorosana