Sportfolio

Inspiration

Our inspiration for Sportfolio was the pain points that sports betters all share: losing money. In our search for solutions, we noticed sports betting market problems are surprisingly similar to the stock trader problems. So why don't we employ similar quantitative approaches? We are empowering sports better with their own "financial dashboard"--PrizePicks style.

What it does

Sportfolio acts as a wrapper over PrizePicks that allows them both the quantitative tools of a stock trader as well as the ability to trade like one. Our app offers the ability for a user to "buy and hold" a player's "stock" in the form of placing long term repeating parlays on the same diversified portfolio of players to minimize risk over the long term. Stock traders know that long-term, consistent gains beat aggressive and emotionally-driven strategy.

How I built it

We built Sportfolio on the MERN stack, and with Typescript. Our data is derived from the NBA JSON api from the 2024-2025 season, giving us about a variety of over 1500 games throughout which we analyzed players' performance. We had a data pool of tens of thousands of values that was stored in MongoDB. We also involved a small Python script called pygoogle for scraping articles relevant to the players and helpful to the user for context. It spoke to our application through a FastAPI server that integrated with our Node.js backend. We also introduced caching in order to speed up localized sessions, since we would be placing fantasy bets from past seasons rather than ones in real-time.

Challenges I ran into

Redis caching turned out to be a major roadblock, where our AI coding tools were running into walls. Additionally, a big part of the design process was arguing over the feature list. With our team having one avid PrizePicks user and one finance bro, we conflicted over what aspects of each field would be more strongly represented in our app. Half of our 36 hours was heavy discussion.

Accomplishments that I'm proud of

I am proud to have put myself out of my comfort zone as far as the track I selected. Even though I enjoy sports, I had never seen it as an industry that presented significant needs, since it's for entertainment. However, I saw parallels with my own field of interest, finance, and was able to create insights to solving user problems using other domain knowledge.

What I learned

I learned that being flexible with domain makes for a fun hackathon. I definitely leaned toward other tracks, but competing alongside the teammates I met here at GT forced me to partake in a new experience, and I learned a ton about how the world of sports betting operates. I also learned I need to brush up on GitHub, because the pulling/pushing pipeline lost me some sanity today.