prem-elo

A python project for reimaging ranking premier league seasons by using ELO rather than the traditional points system using historical data from premier league matches soruced from Kaggle: https://www.kaggle.com/datasets/evangower/premier-league-matches-19922022.

Features

• Calculates Elo ratings for all Premier League teams in a given season.
• Uses a custom Elo update formula with adjustable sensitivity and K-factor.
• Reads match data from a CSV file (premier-league-matches.csv).
• Plots Elo rating progression for selected teams across all gameweeks.
• Outputs detailed Elo changes and league table after each gameweek.

Requirements

• Python 3.x
• pandas
• matplotlib

Install dependencies with:

pip install pandas matplotlib

Usage

1. Place the premier-league-matches.csv file in the project directory.
2. Run the script:

python calc.py

3. The script will:
   • Process the matches for the season specified by SEASONEND in calc.py.
   • Print Elo changes and league tables for each gameweek.
   • Display a plot of Elo ratings for selected teams (default: Arsenal, Manchester City).

File Descriptions

• calc.py: Main script for Elo calculation and visualization.
• premier-league-matches.csv: Historical Premier League match results. Columns include:
  • Season_End_Year, Wk, Date, Home, HomeGoals, AwayGoals, Away, FTR

Customization

• To change the season analyzed, modify the SEASONEND variable in calc.py.
• To plot different teams, edit the teamname list near the end of calc.py.
• Adjust K_FACTOR and SENSITIVITY in calc.py to experiment with Elo calculation dynamics.

Example Output

• Console output for each gameweek, showing match results and Elo changes.
• Matplotlib plot showing Elo rating progression for selected teams.