~~# NOTE: Only works on my machine. ~~
We didn't deploy it onto Streamlit Cloud until after the hackathon ended ๐ so we ran out of time to debug it for prod. MongoDB ended up being the end of it all ๐ญ๐ญ๐ญ๐ญ
Enjoy the fluff below, which assumed we did successfully get our project running outside of "my machine".
Dev update, post judging this has been fixed! For ConUHacks IX please disregard this and all other changes made past 11h 2025-02-02 in regards to project grading and awards
๐ง Inspiration
It all started when Skye and Alex casually threw out an idea that completely derailed our carefully planned hackathon projects. Instead of something reasonable (whatever that means), we dove headfirst into the DRW Open Data Challenge and decided to build a tool for analyzing Magic: The Gathering (MTG) deck play style archetypes.
Because, letโs be honestโour team is full of nerds (except for Ben who hasn't played MTG), and we were way too hyped about making data-driven MTG strategies a reality.
๐ What Does It Actually Do?
Our project helps players build effective strategies by analyzing relationships between archetypes using past tournament data. Instead of relying on gut feelings or anecdotal evidence, MTG-Versus lets you break down matchups scientifically.
Planned Features & Future Upgrades
- ๐ฎ User-added win/loss ratios โ Let the community contribute real-world results!
- ๐ Expanding archetype relationships โ More decks, more data, better insights!
๐ How We Built It
- Python (Backend & Data Processing)
- MongoDB Atlas (Because databases are funโฆ sometimes)
- Streamlit (For a clean and simple Web UI)
- GitHub (Version control, chaos management)
- OpenAI (Easter egg feature: A card generator for fun!)
- Caffeine (lots of it)
๐ง Challenges We Ran Into
- Ben vs. Python Environments on MacOS โ Python won. Ben still holds a grudge.
- Getting data into MongoDB โ One of us literally called it โfucking annoying.โ
- Streamlit CSS Shenanigans โ Injecting CSS to go beyond default Streamlit styles felt like hacking the Matrix.
- The 24-hour time crunch โ We felt every second of it.
๐ Accomplishments Weโre Proud Of
- We got MongoDB working despite wanting to throw our laptops out the window.
- We actually built a functional UI (shoutout to Streamlit for making it easy).
- We learned new technologies under extreme time pressure and still walked away with something cool.
๐ค What We Learned
- MongoDB Atlas authentication is a headache (seriously, why?)
- Ben still hates Python environments on MacOS
- Hackathons are equal parts adrenaline, suffering, and magic.
๐ฎ Whatโs Next for MTG-Versus?
- โ Keep the site alive and build a community around it.
- โ Expand the database to include more matchup relationships.
- โ Let users add their own win rates, custom cards, and archetype data.
Final Thoughts?
Would we do it again? Probably. Would we fix MongoDB first? Absolutely.
Built With
- mongodb
- streamlit
Log in or sign up for Devpost to join the conversation.