Football Match Predictor

Table of Contents

• Overview
• Features
• Installation
  • Prerequisites
  • Steps
• Configuration
• Usage
• How It Works
  • Data Flow
  • Prediction Algorithm
• File Structure
• API Requirements
• Limitations
• Future Improvements
• Contributing
• License

Overview

This is a Python-based football match prediction system that analyzes recent team performance to predict match outcomes. The system fetches historical match data from the Football-Data.org API, processes this information, and provides predictions based on team form.

Features

• Fetches recent match data for any team in the database
• Calculates team form based on last 5 matches
• Provides prediction with confidence percentage
• Displays recent form for both teams
• Supports over 200 teams across major European leagues
• Simple command-line interface
• Data persistence (saves fetched matches locally)

Installation

Prerequisites

• Python 3.8 or higher
• pip package manager
• Football-Data.org API key (free tier available)
• git installed on your system

Steps

1. Clone the repository:

   git clone https://github.com/yourusername/football-match-predictor.git
   cd football-match-predictor

2. Create a virtual environment (recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install dependencies:

   pip install -r requirements.txt

Configuration

1. Get your API key from Football-Data.org.
2. Create a config/api_keys.json file with the following content:

   {
      "FOOTBALL_DATA_API_KEY": "your_api_key_here"
   }

3. The program comes with a pre-populated data/teams.json containing team IDs and aliases. You can modify this file to add more teams if needed.

Usage

Run the program:

python src/main.py

Follow the on-screen prompts:

1. Select option "1" to predict a match.
2. Enter the home team name (can use official name or any alias).
3. Enter the away team name.
4. View the prediction results.

Example session:

Football Match Predictor
1. Predict match outcome
2. Exit

Select option: 1
Home team: man city
Away team: liverpool

Fetching data for man city and liverpool...

Predicting man city vs liverpool...

Manchester City FC won 3, drew 1, and lost 1 in their last 5 matches
man city form: WWDLW
Liverpool FC won 4, drew 0, and lost 1 in their last 5 matches
liverpool form: WWWLW

========================================
PREDICTION: LIVERPOOL FC is favored
Confidence: 66.7%

Current Form:
- man city: WWDLW
- liverpool: WWWLW
========================================

How It Works

Data Flow

1. Data Collection (fetch_matches.py):
   • Resolves team names to API IDs.
   • Fetches last 5 finished matches from Football-Data.org.
   • Processes and saves match data locally.
2. Prediction Engine (predictor.py):
   • Loads saved match data.
   • Calculates team form (win/draw/loss sequence).
   • Computes performance scores.
   • Generates prediction probabilities.
3. Main Application (main.py):
   • Provides user interface.
   • Coordinates data fetching and prediction.
   • Displays results.

Prediction Algorithm

The system uses a simple form-based algorithm:

• Win = 3 points
• Draw = 1 point
• Loss = 0 points
• Probability is calculated based on the ratio of these points.
• Recent form (last 5 matches) is displayed as a string (e.g., "WWDLW").

File Structure

football-match-predictor/
├── config/
│   └── api_keys.json          # API key configuration
├── data/
│   ├── processed/             # Saved match data
│   └── teams.json             # Team database
├── src/
│   ├── data_collection/
│   │   └── fetch_matches.py   # Data fetching module
│   ├── models/
│       └── predictor.py       # Prediction engine
├── main.py                    # Main application
├── requirements.txt
├── .gitignore
└── README.md

API Requirements

This program uses the Football-Data.org API:

• Free tier allows 10 requests per minute.
• Requires registration to get an API key.
• Rate limiting is handled by the API provider.

Limitations

• Only considers last 5 matches by default.
• Doesn't account for injuries/suspensions.
• No head-to-head history analysis.
• Basic probability calculation.
• Dependent on API availability.

Future Improvements

• Add machine learning model for predictions.
• Incorporate more statistics (goals scored/conceded).
• Add head-to-head historical data.
• Implement GUI/web interface.
• Add league table position consideration.
• Include player statistics.

Contributing

Contributions are welcome! Please open an issue or pull request for any:

• Bug fixes
• New features
• Documentation improvements
• Performance optimizations

License

This project is licensed under the MIT License - see the LICENSE file for details.