Real-time NFL Analytics with AI-Powered Insights

GameSense is a live sports analytics platform that transforms NFL play-by-play data into intelligent, contextual insights using AI commentary, anomaly detection, and interactive chat features. Built for HackGT 12.

• Real-time play updates via Server-Sent Events (SSE)
• Live scoreboard with quarter tracking and game clock
• Dynamic possession tracking and down/distance display
• Win probability calculations based on game state

• Smart play analysis using Google Gemini 2.0 Flash
• Contextual insights that explain why plays matter
• Professional-style commentary for each play
• Interactive chatbot for game-specific questions

• Anomaly detection for unusual team patterns
• Trend analysis for rushes/passes per drive
• Statistical outlier identification using Z-scores
• Rolling window analysis for performance trends

• Game-aware AI assistant with full play history context
• Smart suggestions for follow-up questions
• Real-time game insights and strategic analysis

• Python 3.8+ with pandas and FastAPI
• Node.js 18+ and npm
• Google Gemini API key (Get one here)

# Install Python dependencies
pip install fastapi pandas uvicorn python-multipart

# Start FastAPI server
uvicorn main:app --reload --host 0.0.0.0 --port 8000

# Navigate to frontend directory
cd sports-insights-app

# Install dependencies
npm install

# Add your Gemini API key to your environment
echo "REACT_APP_GEMINI_API_KEY=your-api-key-here" >> .env.local

# Start development server
npm run dev

Open http://localhost:3000 to view the live dashboard.

Retrieve team information for a specific game.

{
  "home_team": "KC",
  "away_team": "LAC"
}

Server-Sent Events stream for real-time play updates.

data: {
  "quarter": 2,
  "time": 1847,
  "description": "P.Mahomes pass short right to T.Kelce for 12 yards",
  "yards_gained": 12,
  "offense_team": "KC",
  "defense_team": "LAC",
  "down": 2,
  "yards_to_go": 8,
  "yard_line": 35,
  "play_type": "PASS"
}

The system generates contextual commentary using Google Gemini:

• Situational awareness: red zone, 4th down, close games
• Play significance: big plays, scoring drives, momentum shifts
• Professional tone: broadcast-style commentary with proper terminology

// Example: Detecting unusual rush patterns
const detector = new FootballAnomalyDetector();
const result = detector.detectTrendChanges(
  rushesPerDrive, 
  "Rushes per drive"
);
// Returns: trend changes, statistical outliers, sudden spikes

• Rushes per drive by team
• Passes per drive by team
• Completion rates per drive
• Win probability changes based on plays
• Drive efficiency and time of possession

• Modern design with dark/light theme support
• Real-time updates with smooth transitions
• Responsive layout optimized for desktop and mobile
• Accessibility with full keyboard navigation and screen reader support
• Performance optimized with React 18+ and Next.js 15 features

• FastAPI: High-performance Python web framework
• Pandas: Data manipulation and analysis
• Asyncio: Asynchronous streaming capabilities
• CORS: Cross-origin resource sharing for frontend integration

• Next.js 15: React framework with App Router
• TypeScript: Type-safe development
• Tailwind CSS 4: Utility-first styling
• Radix UI: Accessible component primitives
• Lucide Icons: Icon library
• Google Gemini AI: Advanced language model integration

• NFL Play-by-Play Data: 2024 season dataset
• Real-time processing with streaming and caching
• Custom anomaly detection algorithms
• Context-aware AI commentary generation

• Efficient streaming with Server-Sent Events and automatic reconnection
• Smart caching of team and game state data
• Optimized React renders with proper state management
• Error handling for graceful degradation and recovery