Chess Game with AI Advantage Prediction

A modern chess game implementation with real-time AI-powered position evaluation and a beautiful graphical interface. The game features a professional design, intuitive controls, and machine learning-based position analysis.

Features

Gameplay

• Interactive chess board with legal move validation
• Real-time position evaluation using machine learning
• Visual advantage bar showing current game state
• Move history display (last 10 moves)
• Current player indicator
• Chess coordinates (a-h, 1-8)
• Reset game functionality (R key)
• Cancel selection (ESC key)
• Pawn promotion with visual piece selection dialog

User Interface

• Modern, clean design with professional color scheme
• Smooth hover effects and piece selection highlighting
• Side panel with game information
• Responsive board sizing and centering
• Shadow effects and visual feedback
• Alternating move history background
• Board coordinates for easy move reference
• Centered promotion dialog with piece selection

AI Integration

• Machine learning model for position evaluation
• Real-time advantage prediction
• Feature extraction from board positions
• Gradient-based advantage visualization

Technical Features

• Pygame-based graphical interface
• Python-chess library for game logic
• RandomForest model for position evaluation
• Efficient board state management
• Frame rate limiting for smooth gameplay
• Error handling for resource loading

Requirements

• Python 3.8+
• pygame>=2.5.0
• python-chess>=1.10.0
• scikit-learn>=1.3.0
• numpy>=1.24.0
• joblib>=1.3.0

Installation

1. Clone the repository:

git clone https://github.com/ayush-jadaun/chess-game.git
cd chess-game

2. Set up a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Run the game:

python src/game/main.py

Project Structure

chess-game/
├── src/
│   ├── game/
│   │   ├── __init__.py
│   │   └── main.py           # Main game implementation
│   ├── engine/
│   │   ├── __init__.py
│   │   └── model.py          # Chess engine and ML model
│   └── utils/
│       ├── __init__.py
│       └── constants.py      # Game constants
├── assets/
│   ├── images/              # Chess piece images
│   │   ├── bB.png          # Black bishop
│   │   ├── bK.png          # Black king
│   │   ├── bN.png          # Black knight
│   │   ├── bP.png          # Black pawn
│   │   ├── bQ.png          # Black queen
│   │   ├── bR.png          # Black rook
│   │   ├── wB.png          # White bishop
│   │   ├── wK.png          # White king
│   │   ├── wN.png          # White knight
│   │   ├── wP.png          # White pawn
│   │   ├── wQ.png          # White queen
│   │   └── wR.png          # White rook
│   └── models/
│       └── chess_model.pkl  # Trained ML model
├── data/
│   └── pgn/                # PGN game data for training
├── requirements.txt        # Project dependencies
└── README.md              # Project documentation

How to Play

1. Starting the Game:

   • Run the game using python src/game/main.py
   • The board will appear with pieces in their starting positions

2. Making Moves:

   • Click on a piece to select it
   • Valid moves will be highlighted with circles
   • Click on a highlighted square to move
   • Press ESC to deselect a piece

3. Pawn Promotion:

   • When a pawn reaches the opposite end of the board (8th rank for White, 1st rank for Black)
   • A promotion dialog will appear in the center of the screen
   • Click on the piece you want to promote to (Queen, Rook, Bishop, or Knight)
   • Press ESC to cancel the promotion and choose a different move

4. Game Controls:

   • R: Reset the game
   • ESC: Cancel piece selection or promotion dialog
   • Mouse: Select and move pieces

5. Game Features:

   • The side panel shows the current player and advantage
   • Move history displays the last 10 moves
   • The advantage bar shows the current position evaluation
   • Chess coordinates help identify squares
   • Hover effects show which square you're pointing at
   • Visual feedback for piece selection and legal moves

Model Training

The chess engine uses a machine learning model trained on:

• PGN game data
• Position features (768 binary features representing piece positions)
• RandomForest regressor for position evaluation

Future Improvements

• Add multiplayer support
• Implement game saving/loading
• Add difficulty levels
• Improve AI model accuracy
• Add sound effects
• Implement undo/redo
• Add game statistics
• Create tutorial mode
• Add keyboard shortcuts for moves
• Implement move validation hints

Author

• Ayush Jadaun - GitHub

Support

For support, please open an issue in the GitHub repository or contact the author.