AI-Powered Drafting Assistant for League of Legends Esports
Built for the Sky's the Limit - Cloud9 x JetBrains Hackathon
DraftOracle is an AI-powered tool that helps esports teams make optimal draft decisions in League of Legends. By analyzing historical match data from the GRID Esports API, it provides real-time recommendations for picks and bans based on:
- Champion Synergies - How well champions work together
- Counter Matchups - Lane advantages and phase-specific counters
- Meta Analysis - Current patch tier lists and pick/ban priorities
- Player Comfort - Individual player champion pools and performance
| Feature | Description |
|---|---|
| 🎯 Pick Recommendations | Get optimal champion suggestions with reasoning |
| 🚫 Ban Recommendations | Smart bans targeting opponent weaknesses |
| 📊 Draft Analysis | Full draft breakdown with win probability |
| ⚡ Real-Time WebSocket | Live updates during pick/ban phase |
| 📈 Meta Snapshots | Current patch tier lists and statistics |
POST /api/v1/draft/recommend/pick - Get pick recommendations
POST /api/v1/draft/recommend/ban - Get ban recommendations
POST /api/v1/draft/analyze - Analyze draft composition
GET /api/v1/meta/current - Get current meta snapshot
GET /api/v1/meta/champions - Get all champions
WS /ws/draft/{session_id} - Real-time draft session
- Python 3.11+
- PostgreSQL 15+
- Redis 7+
# Clone the repository
git clone https://github.com/your-team/draftoracle.git
cd draftoracle
# Create virtual environment
python -m venv venv
source venv/bin/activate # Linux/Mac
# or: .\venv\Scripts\activate # Windows
# Install dependencies
pip install -e ".[dev]"
# Copy environment config
cp .env.example .env
# Edit .env with your settings
# Run database migrations (coming soon)
# alembic upgrade head
# Start the API
python -m src.api.maindocker-compose up -dThe API will be available at http://localhost:8000.
The modern React frontend is located in frontend/.
cd frontend
npm install
npm run devThe UI will be available at http://localhost:5173.
The API will be available at http://localhost:8000
Once running, access the interactive API docs:
- Swagger UI: http://localhost:8000/docs
- ReDoc: http://localhost:8000/redoc
import httpx
response = httpx.post(
"http://localhost:8000/api/v1/draft/recommend/pick",
json={
"draft_state": {
"blue_picks": [{"champion_id": 1, "role": "top"}],
"red_picks": [],
"blue_bans": [10, 20, 30],
"red_bans": [40, 50, 60],
"patch_version": "26.02"
},
"our_side": "blue",
"role_needed": "mid",
"top_k": 5
}
)
recommendations = response.json()["recommendations"]
for rec in recommendations:
print(f"{rec['champion_name']}: {rec['score']:.2f}")
print(f" Reasoning: {', '.join(rec['reasoning'])}")const ws = new WebSocket("ws://localhost:8000/ws/draft/my-session-123");
ws.onmessage = (event) => {
const data = JSON.parse(event.data);
console.log("Update:", data);
};
// Make a pick
ws.send(JSON.stringify({
type: "pick",
side: "blue",
champion_id: 1,
role: "mid"
}));
// Request recommendations
ws.send(JSON.stringify({
type: "request_recommendations",
side: "blue",
role: "jungle"
}));┌─────────────────────────────────────────────────────────────┐
│ Frontend │
│ (Future: React/Vue) │
└─────────────────────────┬───────────────────────────────────┘
│
┌───────────┴───────────┐
│ FastAPI + WS │
│ REST & WebSocket │
└───────────┬───────────┘
│
┌─────────────────────┼─────────────────────┐
▼ ▼ ▼
┌─────────┐ ┌─────────────┐ ┌─────────────┐
│ Synergy │ │ Counter │ │ Draft │
│ Engine │ │ Engine │ │ Optimizer │
└────┬────┘ └──────┬──────┘ └──────┬──────┘
│ │ │
└───────────────────┼──────────────────────┘
│
┌──────────┴──────────┐
│ PostgreSQL + Redis │
│ Data & Cache │
└──────────┬───────────┘
│
┌───────────────┼───────────────┐
▼ ▼ ▼
┌─────────┐ ┌─────────────┐ ┌─────────┐
│ GRID │ │ Data Dragon │ │ Celery │
│ API │ │ API │ │ Tasks │
└─────────┘ └─────────────┘ └─────────┘
draftoracle/
├── src/
│ ├── api/ # FastAPI application
│ │ ├── routes/ # REST endpoints
│ │ ├── websocket/ # WebSocket handlers
│ │ ├── middleware/ # Rate limiting, auth
│ │ ├── schemas.py # Pydantic models
│ │ └── main.py # App entry point
│ ├── services/ # Business logic
│ │ ├── grid_ingestion.py
│ │ ├── data_dragon.py
│ │ ├── synergy_engine.py
│ │ ├── counter_engine.py
│ │ └── draft_optimizer.py
│ ├── ml/ # Machine learning
│ │ ├── models/
│ │ ├── features/
│ │ └── training/
│ ├── db/ # Database
│ │ ├── models.py
│ │ ├── session.py
│ │ └── migrations/
│ └── tasks/ # Background jobs
├── tests/
├── config/
├── docker/
└── scripts/
All configuration is managed via environment variables. See .env.example for available options:
| Variable | Description | Default |
|---|---|---|
GRID_API_KEY |
GRID API authentication key | Required |
DATABASE_URL |
PostgreSQL connection string | postgresql+asyncpg://... |
REDIS_URL |
Redis connection string | redis://localhost:6379/0 |
API_PORT |
API server port | 8000 |
ML_EMBEDDING_DIM |
Champion embedding dimension | 128 |
# Run all tests
pytest
# With coverage
pytest --cov=src --cov-report=html
# Specific test category
pytest tests/unit/
pytest tests/integration/- GRID Esports API - Historical match data, draft sequences, player stats
- Riot Data Dragon - Champion metadata, patch information
- Phase 1: Research & Planning
- Phase 2: Core Backend (Services, DB, API)
- Phase 3: ML Models (Embeddings, Win Predictor)
- Phase 4: Real-Time System (WebSockets, Drag & Drop)
- Phase 5: Testing & Documentation
- Phase 6: Frontend Application (React + Vite, Glassmorphism UI)
Built for the Sky's the Limit Hackathon by passionate esports fans.
This project is licensed under the MIT License - see the LICENSE file for details.
🎮 Powered by GRID Esports Data 🎮
