The draft was approaching. I knew web scraping and how to do machine learning regressions using scikit-learn. It seemed only natural to try and game out my picks in one way or another.

I scrape the website https://www.pro-football-reference.com/ for this project.

There's a massive tutorial to explain all this that I wrote located here:

http://www.kingzephyr.com/2019/09/13/fantasy-football-with-scrapy-and-scikit-learn-part-1/

http://www.kingzephyr.com/2019/09/13/fantasy-football-with-scrapy-and-scikit-learn-part-2/

I am assuming Python 3 is installed on the system, as well as Git.

TLDR: Assuming everything works as expected, the full script should be

cd ~
mkdir project
cd project

python -m venv football_venv
source football_venv/bin/activate

git clone https://github.com/athompson1991/football.git
cd football

pip install -r requirements.txt

mkdir script/data

scrapy crawl passing
scrapy crawl receiving
scrapy crawl defense
scrapy crawl kicking

cd script
python run_script.py

Navigate to an empty directory.

cd ~
mkdir project
cd project

Now make your Python environment and activate said environment.

python -m venv football_venv
source football_venv/bin/activate

Clone this repo to your local machine, then cd into the directory

git clone https://github.com/athompson1991/football.git
cd football

Install requirements

pip install -r requirements.txt

Now for the fun part. The project requires a directory to store the data, so make the directory in the proper location

mkdir script/data

Scrape all the data

scrapy crawl passing
scrapy crawl receiving
scrapy crawl passing

This scrapes the website and saves our "database" as csv files to the directory we just made.

Now simply run the script to crunch all the numbers

cd script
python run_script.py

That should do it! Once the script runs all the way through, there should be a power_rankings.csv file in your script/data directory. This means that the script successfully fit and tuned 21 regression models, then used those models to produce a power ranking.