An Analysis on the Effectiveness of Cooperation in A Soccer Team

We design methods as well as models to explore how the complex interactions among the soccer players on the field impact their success. The data used in this paper is provided by problem D in 2020 ICM, which made up a native soccer team named ’the Huskies’.We perform network analysis on the passing-ball network to identify dyadic and triadic configurations. We further determine that the best formation for the Huskies is the ‘4-3-3‘ formation and the most suitable lineup is D4, D5, D6, D9, M1, M3, M5, F2, F3, F6, which contains two dyadic configurations and one triadic configuration.We present a linear model containing 11 indicators extracted from the data and use it to measure the effectiveness of teamwork. To determine the coefficient of each indicator, we use multiple linear regression to fit the difference of scores of the two sides in each game. By comparing the fitted values with the real values and performing a sensitivity test, we verify the accuracy and stability of our model.By analyzing macro data and comparing the indicators of the winner and the loser of each game, we advise the Huskies’ coaches on psychological construction, coaching team and team structure, etc to improve the Huskies’ performance next season.Finally, we generalize our findings above and provided suggestions on designing effective teams, which are selecting the right members, creating smooth communication networks, etc. We also propose a few new aspects that are not included in my models but are of great importance in evaluating team performance.