Attempting to find the value of passes depending on whether or not they will end up being a goal for/against within a minute using classification models. performs training and predicting on games from the 2018 world cup provided by statsbomb.com and implements a leave one out strategy to avoid predicting on training data.
main.py is all the actual python code needed in one file player_stats.csv is the result of main.py for every player after applying the pass value model.
dependencies
-
numpy
-
pandas
-
sklearn
-
seaborne
-
requests
-
math