Fantasy Football and AI – Week 4

I realize that I forgot to write an update on what our final roster was for week 4. I suppose that is because I am on vacation and writing blog posts isn’t my number one priority.

Week 4, We WON! The AI team is now 2-2. I believe that as we gather more data, include information from waivers, the model is able to get a better sense of who is going to play well. Good data in > good data out. Bad data in > bad data out. This is key for all AI workflows. It is no different than giving a regular employee bad data and expecting them to make something good out of it.

As a reminder, every day, I have a Lambda function that goes out and scrapes stats for player from both ESPN and FantasyPros. This includes a players outlook that is presented on ESPN. The idea here is to give the AI more news about specific players. As this platform expands we could also include other data scraping methods from other news sources or blogs that could help bolster decision making for the model.

This is going to be a short post so I can go back to the beach. Here is the result from Week 4:

For the most part I was pretty happy with the suggestions that the model made. We’re still struggling in the receiver and tight end position. That being said, the bench only had two players who preformed better than what we started. This week came down to the Monday night games where we needed Breece Hall to secure the victory and he put up enough points to beat out the opponent.

Stay tuned for week 5 roster updates and moves which I will be making from Portugal!

Go Lions.