PremScout - Fantasy Premier League Predictions

AI-powered Fantasy Premier League predictions uses machine learning to predict player points and automatically builds the optimal formation for maximum points. On average gets a 15-point improvement for the average player.

What Does PremScout Do?

PremScout is your intelligent FPL assistant that:

• Predicts player points using advanced machine learning
• Builds optimal formations (3-4-3, 4-4-2, 5-3-2, etc.) automatically
• Shows player cards with FIFA-style design
• Analyzes player form and recommends transfers
• Updates daily with fresh predictions for each gameweek

Key Features

• Smart AI Predictions - Random Forest model trained on historical FPL data
• Dynamic Formation Builder - Finds the highest-scoring 11-player team
• Interactive Dashboard - Beautiful FIFA-style player cards with detailed stats
• Dark Mode Interface - Sleek, professional design that's easy on the eyes
• Mobile Responsive - Perfect experience on all devices
• Daily Auto-Updates - Fresh predictions every morning

Issues

• Not all players have images due to premier league api not having them (for example player joined team late)

Quick Start - Run Locally

Prerequisites

• Python 3.9+
• Node.js 18+
• Git

Installation & Setup

1. Clone the repository

   git clone <your-repo-url>
   cd PremScout-1

2. Set up Python backend

   # Install Python dependencies
   pip install -r requirements.txt

3. Set up React frontend

   # Navigate to frontend directory
   cd frontend
   
   # Install Node.js dependencies
   npm install

4. Start the servers

   Terminal 1 - Backend (Python Flask):

   # From project root directory
   python3 csv_server.py

   Backend will run at: http://localhost:5000

   Terminal 2 - Frontend (React):

   # From frontend directory
   cd frontend
   npm run dev

   Frontend will run at: http://localhost:5173

5. Open your browser Visit http://localhost:5173 to see your FPL predictions dashboard!

Technology Stack

• Frontend: React + TypeScript + Vite
• Backend: Python Flask
• ML Model: Random Forest Regressor (Scikit-learn)
• Data Source: Official FPL API
• Deployment: Vercel + GitHub Actions
• Styling: Custom CSS with glassmorphism effects

How It Works

1. Data Collection - Fetches live player data from FPL API
2. Feature Engineering - Creates 15+ predictive features
3. Model Training - Random Forest learns from historical performance
4. Formation Optimization - Tests all valid formations to find highest-scoring team
5. Visual Display - Beautiful pitch layout with FIFA-style cards

Automatic Updates

The system updates daily at 9 AM UTC via GitHub Actions:

• Fetches latest FPL data
• Retrains the prediction model
• Generates new gameweek predictions
• Automatically deploys updates

How to Use

1. View Team of the Week - See the optimized 11-player formation on the pitch
2. Click Player Cards - Get detailed stats and predictions for each player
3. Browse Full Table - Sort and filter all 500+ Premier League players
4. Mobile Friendly - Perfect experience on any device

AI Model Features

The machine learning model analyzes these key factors:

• Player form and consistency
• Expected goals/assists/saves
• Minutes played reliability
• Transfer momentum
• Value indicators
• Recent performance trends
• Opposition difficulty

Model Accuracy

• R² Score: ~0.85+ on validation data
• Improves throughout the season as more data becomes available
• Optimized for gameweek-ahead predictions