[Francesco Baldissera
[Data Analytics Full Time, Barcelona March 2020]
Football is a very complex non Zero-sum game to analyze. Beyond the challenges of gathering the right data there is that of choosing the right KPI's to effectively measure success as well as analyzing tactics over a season.
I wanted to find the best success predictor for a Football team in the league. At first, Football might seem like a simple game, but its ramdomness makes it a very interesting game to analize statistically. I belive finding a good KPI to measure success of teams is crucial to understanding the game.
Testing this coaching mantra using statistical analysis and data visualization to understand better how the Atlanta FC high defensive pressure was successfull and what is the best way to interpret the saying.
I used the datasets provided at https://fbref.com/en/comps/22/Major-League-Soccer-Stats I exported their data to CSV and transformed them to JSON to be able to read them properly and merge them.
- Research about the topic to stablish the criteria of evaluation of data sources
- Researched about the different sources
- Data gathering and wrangling
- Exported some data to use on Tableau
- Writing the paper
- Presentation of less than 3 minutes
I used a Trello kanban board to organize my time.
Inside the your-project directory there are two folders:
- 'Data' where all the sources and the data wrangling jupyter notebook files are.
- 'Graphs' where I kept the exports of all my graphs done in Tableau
- Main paper file is in the folder with the title of the project
Include links to your repository, slides and kanban board. Feel free to include any other links associated with your project.