Back to Blog
Web DevAssisted by GPT4

Pretty Site: Ugly Data

I really get into a user experience that makes sense—unfortunately, my data doesn’t always do the same. This is the story of a “junior” frontend dev cosplaying as a full stacker slacker. 🙂

September 6, 2025
Pretty Site: Ugly Data

🎨 Pretty Site: Ugly Data

Excerpt:
I really get into a user experience that makes sense—unfortunately, my data doesn’t always do the same. This is the story of a “junior” frontend dev cosplaying as a full stacker slacker. 🙂


✨ Shiny on the Surface

I’ve been building a fantasy football sports betting site as a way to stretch myself, create new problems, and then (hopefully) solve them. On the surface, everything looked great. The UI was smooth, the flow made sense, and I felt like I had the “pretty site” part nailed.

But then… the data came knocking.


🏈 When Odds Don’t Add Up

My scoring services were failing at exactly the wrong time: when odds were being updated.
That meant wagers were being placed after odds had shifted, making the entire bet moot.

Sports betting already feels like an impossible math puzzle, but when your data pipeline fails you, it’s game over. At one point, a missing float value turned what should have been normal losses into $10k+ fantasy gains.

Yeah. Ouch.


🐳 Docker Jail & Other Fun

I fixed the scoring bug… but in the process, I managed to break my production environment.
So there I was: trapped in Docker jail, wishing I had hair to pull out while my solution rebuilt in prod.

The lesson? Testing is important. Breaking things is too. Both usually happen in the wrong order. 😅


🙈 The Good(?) News

Thankfully, only two users had signed up so far. Maybe they won’t notice… (well, except for the part where I accidentally dropped my database). They’ll definitely notice that.

But hey—onward and upward, right?


🚀 What’s Next

I’m still refining this project, and I’d love for you to check it out:

👉 Visit my portfolio to see the work in progress.
👉 Reach out through the contact page if you’d like an invite code to test the platform.


❤️ Final Thought

I may be a frontend-first dev stumbling through the backend world, but every crash and bug teaches me something new. And at the end of the day, this project is about learning by breaking.

So here’s to broken data, rebuilds in prod, and onward progress.

Stay tuned for more dev chaos!

AI Transparency
Model: GPT4
Prompt Used:
Todays blog title: Pretty Site: Ugly Data Excerpt: I really get into a user experience that makes sense unfortunately my data doesn't always do the same. This is the story of an "Junior" Frontend Dev cosplaying as a Full Stacker Slacker:) I am building a fantasy football sports betting site to create and solve problems I haven't had to build around. Everything on the surface was working great but my scoring services were failing and updating odds at the time of scoring thus making any wagers places moot since the "odds" users wagered on would change. Sports gambling already feels impossible but when your data fails you... I had a missing float that turned losses in to 10k+ gains, what... a... mess... I got it fixed but broke my production environment in the process. Spending a day in docker jail and wishing i had hair to pull out. I'm writing this as my solution rebuilds in prod so stay tuned BUT testing is important and so is breaking things. Thankfully only 2 users have signed up and I don't think they'll notice. whoops, i dropped my db... they will notice. Onward and upward check out my portfolio to see the project and reach out through the contact page if youd like an invite code. Can you help me create a blogpost out of that stream of conscious above please?