Hacker News new | ask | show | jobs
Show HN: Early Shift – Detect trending Roblox mechanics 24-48h before saturation (github.com)
2 points by san10 259 days ago
I built a system that monitors 500+ Roblox games and correlates CCU spikes with YouTube creator activity to detect trending game mechanics before they saturate the market.

How it works: - Polls CCU data via RoProxy (TOS-compliant) - Tracks top 20 Roblox YouTube creators - Detects when: CCU growth ≥25% + YouTube video mentions game + mechanic keyword match (NEW/UPDATE/SECRET) - Alerts studios via Notion

Tech: Python, DuckDB, aiohttp, RapidFuzz. ~600 lines of code.

Use case: Studios can clone trending mechanics before 20+ competitors do the same.

Example: Could have caught Pet Simulator X's merge mechanic 36h before it hit Popular page (hypothetical - system is new).

GitHub: https://github.com/SanchitSharma10/early-shift

Built this in 3 days to explore multi-agent architecture patterns. Roadmap includes BERT classification and multi-signal fusion (TikTok/Twitter).

Open to feedback on the approach!
2 comments
maximilianthe1 259 days ago
Looks like you've committed `.env` file with private info. I suggest you change token immediately and add `.gitgnore`! That's a *security risk*!

It's best practice to keep env files private, but commit something like `template.env` with placeholder values.

Other than that - looks like a must-use tool for every roblox studio!

link
san10 259 days ago
Thanks for catching this. Strange part is I actually had added this to my git ignore last time, so maybe something went wrong in the process.

Highly appreciated!

Thank you for the feedback as well :)

link
NoahZuniga 259 days ago
How many false positives?
link
san10 259 days ago
Ty for the question.

The system is new (built last week), so I don't have reliable production data on false positive rate yet.

Right now my current detection logic is basically: - 25%+ CCU growth vs 7-day baseline - YouTube video mentioning game within 48h - Keyword match (currently using regex and exploring other methods)

If I have to guess where the false positives are going to come from: - YouTuber plays game (drives CCU) but doesn't mention new mechanic - CCU (concurrent users) spike from external event (streamer or holidays) - Generic update videos that don't indicate mechanic type

Next step is running it for 30 days and tracking precision/recall. More of that in the GitHub itself.

Appreciate the question, since it's the main thing I need to validate.

link