Show HN: Early Shift – Detect trending Roblox mechanics 24-48h before saturation

2 hours ago 2

Detect trending Roblox game mechanics 24-48 hours before market saturation.

Early Shift monitors 500+ Roblox games and cross-references CCU spikes with YouTube creator activity to alert studios when new mechanics go viral - before 20+ competitors copy them.

Example: Caught Pet Simulator X's merge mechanic spike 36 hours before it hit the Popular page.

Current System showing via Mark-Down UML Diagram(v0.1)

┌─────────────────┐ │ RoProxy API │ ← Poll CCU data every 6h for top 500 games └────────┬────────┘ │ ▼ ┌─────────────────┐ │ DuckDB │ ← Store: games table (7-day rolling window) │ (early_shift.db)│ youtube_videos (48h lookback) └────────┬────────┘ mechanic_spikes (historical feed) │ ▼ ┌─────────────────┐ │ Mechanic │ ← Detect: CCU growth >= 25% │ Detector │ + YouTube mention within 48h └────────┬────────┘ + Keyword match (NEW/SECRET/UPDATE) │ ├──────────┐ ▼ ▼ ┌─────────────┐ ┌──────────────┐ │ Notion │ │ DuckDB │ │ Database │ │ mechanic_ │ │ │ │ spikes table│ └─────────────┘ └──────────────┘

Components:

roproxy_client.py: CCU data collection (respects rate limits, TOS-compliant)
youtube_collector.py: Top 20 Roblox creator video monitoring
mechanic_detector.py: Correlation logic + fuzzy matching (RapidFuzz)
notion_writer.py: Studio alert delivery
main.py: Orchestrator (runs every 6 hours)

Tech Stack: Python 3.11, DuckDB, aiohttp, RapidFuzz, Notion API

🚀 Quick Start (5 minutes)

Install dependencies

cd early_shift pip install -r requirements.txt
Run first collection cycle
View detected spikes

python mechanic_detector.py

📊 Setting up Notion Integration

Create a Notion database with these columns:
- Game (Title)
- Growth (Number)
- Current CCU (Number)
- Mechanic (Text)
- Source (URL)
- Detected (Date)
Get your Notion API token:
- Go to https://www.notion.so/my-integrations
- Create new integration
- Copy the token
Share database with integration:
- Open your database in Notion
- Click "..." menu → "Add connections"
- Select your integration
Get database ID:
- Open database as full page
- Copy ID from URL: notion.so/xxxxx/**DATABASE_ID**?v=yyyy
Add studio to Early Shift:

python add_studio.py --name "Your Studio" --token "YOUR_TOKEN" --database "YOUR_DB_ID"

# Install test dependencies pip install pytest # Run unit tests pytest test_mechanic_detector.py -v # Run integration test python test_early_shift.py

See ROADMAP.md for detailed development plan.

Phase 2 (4-6 weeks):

React dashboard UI for historical spike visualization
BERT-based mechanic classification (economy/gameplay/social/cosmetic)
Multi-signal fusion (TikTok + Twitter mentions)
Confidence scoring based on historical accuracy

Phase 3 (8-12 weeks):

Supabase data warehouse migration for scale
Real-time WebSocket alerts
Studio feedback loop (mark spikes as "useful" / "false positive")
API for programmatic access

Current: MVP/Demo phase
Purpose: Demonstrating multi-agent architecture for trend detection
Next steps: See ROADMAP.md for production considerations

Built in 3 days as a technical demonstration of:

Real-time data pipeline design
Multi-source signal correlation
Production-ready code patterns

For growth spikes >50%, subscribe to mobile push notifications:

# iOS/Android app: ntfy # Subscribe to: https://ntfy.sh/early-shift-alerts

# Run monitoring loop (checks every 6 hours) python main.py --production # Or use cron (Linux/Mac) crontab -e # Add: 0 */6 * * * /path/to/python /path/to/early_shift/main.py # Or Task Scheduler (Windows) # Create task to run main.py every 6 hours

Edit roproxy_client.py to customize:

Game list (default: top 500 by CCU)
Poll frequency (default: 6 hours)
Growth threshold (default: 25%)

Edit mechanic_detector.py for detection tuning:

Lookback window (default: 48 hours)
Fuzzy match threshold (default: 82%)
Keyword hints for mechanic extraction

Game | Growth | Current CCU | Mechanic | Source | Published -------------------- | ------- | ----------- | --------------------- | ------------------------------- | ------------------- Pet Simulator X | +38.2% | 142,340 | New merge pets system | https://youtube.com/watch?v=... | 2025-10-01 14:23 Blox Fruits | +31.7% | 287,120 | Dragon Quest v2 | https://youtube.com/watch?v=... | 2025-10-01 09:15 Adopt Me | +26.4% | 156,890 | Pet fusion mechanics | https://youtube.com/watch?v=... | 2025-09-30 21:47

Built by @SanchitSharma10

Total lines of code: ~600
Setup time: <10 minutes
Detection accuracy: Being validated (beta phase)