NFL Elo Rating System

This project is a Python program that implements an Elo rating system for NFL teams. The program calculates and updates team ratings based on game results, predicts game outcomes, and provides a betting strategy by assigning confidence points to each game prediction.

Table of Contents

1. Overview
2. How It Works
3. Features
4. Usage
5. Dependencies
6. File Descriptions
7. Contributing

Overview

The Elo rating system is a method for calculating the relative skill levels of players (or teams) in zero-sum games such as chess. This project applies the Elo system to NFL football teams, allowing for dynamic rating updates after each game. The program can calculate the probability of a team winning, predict weekly game outcomes, and provide confidence points for betting.

How It Works

The program uses an Elo rating algorithm adjusted for the NFL:

1. Expected Score Calculation: The expected score of a team is calculated based on its Elo rating relative to its opponent's rating.

   E_A = 1 / (1 + 10^((R_B - R_A) / 400))

2. Margin of Victory (MOV) Adjustment: The K-factor, which determines how much ratings change after each game, is adjusted based on the margin of victory. Larger victories result in larger rating changes.

   Adjusted K = K * (MOV / (MOV + 1)) * log(MOV + 1.5)

3. Rating Update: After each game, ratings for both teams are updated based on the expected score and the actual result using the Elo formula.

Features

• Calculate Win Probability: Computes the probability of a team winning against another based on their current Elo ratings.
• Add Game Results: Updates team ratings after a game result is input.
• Display Current Ratings: Prints the current Elo ratings for all teams.
• Predict Weekly Game Outcomes: Provides game outcome predictions for a week and assigns confidence points to each prediction based on the likelihood of winning.
• Save and Load Ratings: Elo ratings are saved to and loaded from a CSV file (elo_scores.csv).

Usage

1. Run the Program: Execute the nfl_elo.py script in your Python environment.

   python nfl_elo.py

2. Commands:

   • 1: Calculate probability of team winning - Input two teams to see their win probabilities.
   • 2: Add game result - Input the result of a game (team names and scores) to update the Elo ratings.
   • 3: Display current ratings - View the current Elo ratings for all teams.
   • 4: Get week prediction by points - Predict the outcome of weekly games and assign confidence points.
   • 5: Save and exit - Save the updated Elo ratings to elo_scores.csv and exit the program.

3. Input Validation: The program includes validation to ensure that valid team names are input.

Dependencies

The program requires the following Python libraries:

• pandas
• math

Install them using pip if you haven't already:

pip install pandas

File Descriptions

• nfl_elo.py: The main program file containing all the functionality for the Elo rating system.
• elo_scores.csv: A CSV file that stores the initial Elo ratings of NFL teams. This file is updated with new ratings after games are played.

Key Functions in nfl_elo.py

• read_file(file_name): Reads Elo ratings from a CSV file into a dictionary.
• calculate_expected_scores(rating1, rating2): Calculates expected scores for two teams based on their Elo ratings.
• calculate_new_ratings(rating1, rating2, result, mov): Calculates new ratings after a game result.
• get_mov_adjustment(mov): Adjusts the K-factor based on the margin of victory.
• calculate_probability(team1, team2, ratings): Computes the probability of a team winning against another.
• add_game_result(team1, team2, team1_score, team2_score, ratings): Adds a game result and updates the Elo ratings.
• get_week_prediction_by_points(ratings): Predicts weekly game outcomes and assigns confidence points.
• save_ratings(ratings): Saves the updated ratings to the CSV file.

Contributing

Contributions to improve this project are welcome! Feel free to fork this repository, make changes, and submit a pull request. Please ensure your code adheres to the project's style and passes all tests.