Using AI and Python to predict NFL Matches

Utilize power of AI to make NFL score predictions

Introduction

Today we will leveraging the power of Claude Sonnet 3.5 to generate predictions for the upcoming Week 9 of NFL games.

Given NFL current season data from pro football reference, the idea here is we will not write any code ourselves instead we will ask Claude to write all code. We will simply run the code to ensure it executes properly and then examine the generated predictions.

Walkthrough

First, lets get the data we will be making predictions on. For simplicity I will only be using data for the current NFL season scraped from -

2024 NFL Weekly League Schedule | Pro-Football-Reference.com

The benefit of this data is that it also contains upcoming fixtures which should make our workflow much easier.

The first prompt I will ask claude is- “Using requests and beautiful soup extract the data from this table”. Table-

The code generated is

from bs4 import BeautifulSoup
import pandas as pd

def extract_table_data(html_content):
    """
    Extract data from an HTML table with the id 'games' into a DataFrame
    
    Parameters:
    html_content (str): HTML content containing the table
    
    Returns:
    pandas.DataFrame: DataFrame containing the table data
    """
    # Parse HTML with BeautifulSoup
    soup = BeautifulSoup(html_content, 'html.parser')
    
    # Find the table
    table = soup.find('table', id='games')
    
    # Get headers
    headers = []
    for th in table.find('thead').find_all('th'):
        # Get header text, if empty use the data-stat attribute
        header = th.text.strip() if th.text.strip() else th.get('data-stat', '')
        headers.append(header)
    
    # Get rows
    rows = []
    tbody = table.find('tbody') if table.find('tbody') else table
    
    for tr in tbody.find_all('tr')…