Football Analytics

Get information from football games for different competitions and leverage data visualization to statistically analyze and visualize the performance of football teams and players. The scripts are written and intended to use with Python 3.

⚠️ Known limitation — the scrapers no longer return live data. The data-collection scripts (game_data.py, player_in_goals.py) scrape ESPN Argentina pages (espn.com.ar/futbol/numeritos, /comentario, /partido, /resultados). That site has been redesigned since this project was last updated (2018), so the hard-coded CSS classes and URL structure no longer match the live markup and the scrapers will return empty results. The scraping code has been modernized to current library APIs (Selenium 4, pandas 2, matplotlib 3) and the bugs fixed, but reviving the ESPN data source is out of scope. The visualization scripts still work against the sample CSVs included in the repo.

Requirements

• Python 3.8+
• The packages in requirements.txt (pandas, numpy, matplotlib, requests, beautifulsoup4, selenium).
• For the scrapers only: Google Chrome/Chromium plus a matching ChromeDriver available on your PATH. Selenium drives a real browser to collect game IDs.

Getting Started

1. Clone Repo

git clone https://github.com/andrebrener/football_data.git

2. Install Packages Required

Go in the directory of the repo and run: pip install -r requirements.txt

3. Enjoy the repo :)

How to run the scripts

Layout: constants.py and the shared scraper game_data.py live at the repo root; the analysis scripts are grouped under analysis/; standalone graph scripts live in radars/ and other_graphs/*/.

• Scraper core game_data.py reads its settings from constants.py; run it from the repository root (python game_data.py). Output CSVs are written under data/.
• Analysis scripts in analysis/ (player_in_goals.py, penalty_ratio.py, season_analysis.py) import constants/game_data from the root, so run them as modules from the repo root, e.g. python -m analysis.player_in_goals.
• Graph scripts in radars/ and other_graphs/*/ read a CSV that lives next to the script. Paths are now resolved relative to each script's own location, so you can run them from anywhere (e.g. python radars/radars_graph.py).

What can you do?

1. Get Games Data

• Insert Constants in constants.py.
  • Select start & end dates.
  • Select competition from the competition dictionary.
• Run game_data.py.
• A directory named game_data will be created in the repo with a csv file named games_data_{start_date}_{end_date}.csv.
• This csv will contain the data for the games of the selected competition and dates including:
  • Date of the game.
  • Home & away teams.
  • Result of the game.
  • Total shots and shots on goal.
  • Percentage ball possessions.
  • Fouls and yellow & red cards.
  • Team who made the first goal.
  • Team that was 2-0 in any moment of the game. This value is null if no team complies with this condition.
  • Number of penalties per team.

2. Passes Map

This graph shows the average location of the players and the pass distribution of the team during the game. To be able to run passes_map.py, the csv must be in the format as this one.

3. Radars

This graph compares teams/players in many aspects of the game at the same time. To be able to run radars_graph.py, the csv must be in the format as this one.

4. Players Under or Overperforming

This graph shows if the player was over or underperforming compared to what he was expected to. To be able to run over_under_perform.py, the csv must be in the format as this one.

5. Team & Player Analysis

The season_analysis.py script is used to compare teams and team_players_comparison.py compares the players. After running the scripts, you will get a series un graphs with their comparison in different aspects of the game. The options for these graphs are:

• Change titles and subtitles of graphs.
• Define a team/player that you want to be remarked from the rest.
• The maximum number of teams/players in the graph.

6. Get Analysis for Player when on & off the Pitch

• Insert Constants in constants.py.
  • Select start & end dates.
  • Select competition from the competition dictionary.
• Run player_in_goals.py from the repo root: python -m analysis.player_in_goals.
• A directory named player_in_goals will be created in the repo with a csv file named games_data_{start_date}_{end_date}.csv.
• This csv will contain the data for the players of the selected competition and dates including:
  • Name and team of the player. More than one player with exactly the same name in the same team will make that data invalid.
  • Minutes on and off the pitch.
  • Total team goals for and against.
  • Team goals for and against with the player in the pitch.

License

This project is released under the MIT License.