CheckBot Alpha

Overview

Checkbot Alpha is a checkers game built using Python's pygame library. The project is structured with different classes such as Board, Piece, Game, and Algorithms to manage the game’s functionality and the bot's move strategies.

GUI

We used the pygame library in Python to create the checkers GUI. Different classes such as Board, Piece, etc., were created for better readability and understanding.

The BoardnPawns file contains two classes: Board and Piece. These classes include methods that:

• Return the valid pieces and kill pieces in the current board position for a given color.
• Handle the movement of the pieces.
• Return the valid spaces for a given piece.
• Include general methods such as drawing the board, filling the initial pieces, drawing pieces on the window, and promoting a piece to a king.

The Game class handles the flow of the game by:

• Managing the selection of a piece and the destination to move that piece.
• Drawing and updating the window after every move.
• Checking for win or tie conditions.

Additionally, the Algorithms class contains the implementations for both the Random algorithm and the Minimax algorithm, which are used for the bot's moves.

Strategies Used for the Bot

Random

• A random piece is selected from the set of valid pieces for that move.
• From the selected piece, a random valid move is chosen.
• These selections are reflected on the board after the bot's next move.

Minimax

• The evaluation score is calculated using the difference between bot’s and human’s pieces along with the difference between bot’s and human’s kings:

  Score = (Bot - Human + BotKings - HumanKings)

• The bot aims to maximize this score, while the human aims to minimize it.

• A search depth of three is used. If it is the human's move, the algorithm assumes the best move is played (lowest score); if it's the bot's move, the highest score is considered.

• The minimax function is traversed recursively:

  • The base case is reached when the depth is zero or when any player wins.
  • A Boolean variable indicates whether it is the bot's move and alternates with each recursive call.
  • For bot moves (bot = true), the maximum score is selected.
  • For human moves (bot = false), the minimum score is selected.

• The function ultimately returns the best move for the bot.

Comparison of the Strategies

It is evident that the minimax algorithm is superior, as it makes calculated moves, whereas the random algorithm simply selects moves arbitrarily.

Algorithm for Finding Valid Spaces for Pieces

• The algorithm first checks whether a piece is moving up or down (i.e., blue or red).
• Two functions, checkLeft and checkRight, are used to inspect diagonal directions:
  • They check for an empty space or a piece of a different color.
  • If a kill is possible, the destination (i.e., the valid move) is stored in a dictionary mapping to the pieces that can be captured or skipped.
• These functions are called recursively to cover all possible moves.

For king pieces, a modified algorithm is required since they can move both upward and downward:

• Two functions, checkKingLeft and checkKingRight, simultaneously check the left/right and top/bottom directions.
• These methods are also called recursively to cover all valid moves for kings.

Screenshots

Start Of Game

First Move Of Bot

King Coin

Bot Win