myfootballanalytics

Creator of the repository: Ilias Samathrakis

Table of contents

• About
• Prerequisites
• Installation
• Getting started
• Utilized Skills

About

myfootballanalytics is a user-friendly, in-house developed python-based (available here) tool to analyse and visualise football data in order to assess the performance of football clubs and players and extract meaningful insights.

By using myfootballanalytics, you are able to:

• answer questions such as:

  1. Does Mourinho play the catenaccio of the modern years?
  2. Was Guardiola’s Barcelona better than Heynckes’s Bayern Munich?
  3. Is Robert Lewandowski a ’killer’ striker?

• visualise your findings:

• get files with data for your own analysis
• many more.

Prerequisites

1. Install Python3 (at least 3.6) directly from its website https://www.python.org/downloads/ or via anaconda https://www.anaconda.com/

2. Install package manager for python packages 'pip'

   python -m pip install --upgrade pip
   

3. Install the following external python pachakges using 'pip'

   pip install beautifulsoup4
   pip install requests
   pip install tqdm
   pip install mplsoccer
   

Installation

From pip

Type the command

pip install myfootballanalytics

From Github

1. Clone or download the repository
2. Save it at a convenient directory

Getting started

import myfootballanalytics as mfa

update_id_file = True   ## Updates file containing match ids from https://understat.com/
get_data =       False   ## Scrapes data from https://understat.com/ and saves them in the defined directory
analyze_match =  False   ## Plots the position of all shots of a given match and provides a table with information
analyze_league = False   ## Plots xG figure for the teams of the selected league over the selected seasons
analyze_team =   False   ## Plots xG figure for the selected team over the selected seasons

## Relevant if get_data = True
save_csv_file_leagues = True   ## Saves league’s data to csv file

## Relevant if analyze_league = True
save_csv_file_players = True   ## Saves players’ data to csv file

## Relevant if update_id_file = True
parallel = True    ## Activates multiprocessing for faster results.

## Relavant for all tags
save_dir_path = ""  ## Saving directory. Default: current working directory. Empty string activates the default setting. 

## Relevant if get_data = True
leagues = ['PremierLeague','Ligue1','Bundesliga','LaLiga','SerieA']   ## Available leagues: Ligue1, PremierLeague, Bundesliga, LaLiga, SerieA

## Relevant if get_data = True or analyze_league = True or analyze_team = True
seasons = ['2014-2015','2015-2016','2016-2017','2017-2018','2018-2019','2019-2020','2020-2021','2021-2022','2022-2023']

## Relevant if analyze_match = True
home_team = "FC Cologne"  ## Available teams: all teams
away_team = "Bayern Munich"  ## Available teams: all teams
season = "2022-2023"   ## Available season: all seasons

## Relevant if analyze_league = True
league = "Bundesliga"   ## Choose league to analyze. Available leagues: Ligue1, PremierLeague, Bundesliga, LaLiga, SerieA

## Relavant if analyze_team = True
team = "Arsenal"       ## Available teams: all teams

if update_id_file:
    ids = mfa.FindIDs(parallel,save_dir_path)

if get_data:
    data = mfa.DataUpdater(leagues,seasons,save_dir_path,save_csv_file_leagues)
    
if analyze_match:
    md = mfa.MatchDataLoader(home_team,away_team,season,save_dir_path)
    match_data = md.load_match_data()
    
    match = mfa.MatchAnalyzer(match_data)
    table = match.analyze_match()
    print(table)
    
if analyze_league:
    league_data = mfa.LeagueAnalyzer(league,seasons,save_dir_path,save_csv_file_players)
    
if analyze_team:
    td = mfa.TeamDataLoader(seasons,team,save_dir_path)
    team_data = td.load_team_data()
    
    team_analysis = mfa.TeamAnalyzer(team_data,seasons,team)
    team_analysis.analyze_team()

1. Copy the above python code and paste it within a file named 'run.py'

2. Save it at a convenient directory

   Note:

   • If you installed the tool via pip, the location of 'run.py' file is not important.
   • If you downloaded the tool from github, it is recommended to paste 'run.py' within the same directory (a level above myfootballanalytics directory)

3. Modify 'run.py' based on your needs

4. Run 'run.py' using python3

   python run.py
   

Detailed explanation is provided in Instructions_manual.pdf

Utilized skills

• Data collection: Python (Beautifoulsoup)
• Data Analysis: Python (pandas)
• Data visualization: Python (matplotlib)