Inspiration
Michael has been interested in fantasy football since high school, which, paired with his pursuit of a career in data science, was how this project idea was born.
What it does
This program will take current NFL players' college football stats, that was gathered from a few websites using web scrapers and manual input, train a decision forest model, and predict fantasy points of incoming NFL rookie running backs
How we built it
The scrapers were made use a mixture of technologies such as BeautifulSoup, Pandas, and Selenium, to scrape data from the HTML code of a few websites. The model was built using sklearn and the data was imported using Pandas
Challenges we ran into
Trying to interpret the website's HTML code to get the scrapers to grab the data we wanted and making sure the scrapers were grabbing the data of the correct players was tough.
Accomplishments that we're proud of
We are proud that we were able to get the scrapers working as much as possible and that we were able to get our model as accurate as it is in the short time we had
What we learned
We learned that websites surprisingly resist web scraping, which was much harder than we thought.
What's next for Smarter Fantasy
We would like to make the model not only predict running backs, but other positions, and maybe even other sports
Log in or sign up for Devpost to join the conversation.