Python Blackjack Game

A text-based Blackjack game written in Python.

Recently, I've spent a lot of time with pandas in Jupyter Notebooks. This project provided an excellent excuse to revisit Python's core functionality and focus on principles of object-oriented programming whilst building something fun. Along the way: I tried-out the pytest framework (it's great); kept everything modular and extensible; brought dedication to documentation and have lost huge sums of virtual money to the game's dealer.

Credit to Alex Large who inspired this project by starting his own Blackjack game. It got me thinking about how I'd approach the task and before I knew it, this repository was born. Give the game a go, instructions to get started can be found below; one day I'll come back and add a GUI...

Getting Started

1. Clone or download the repository:

git clone https://github.com/Simon-Lee-UK/blackjack-game.git  # Grabs the code from GitHub
cd blackjack-game  # Navigates into the top level of the repository

2. Using the Conda environment and package manager, install the environment required to run the game and associated set of tests. First, create the environment from the environment.yml file:

conda env create -f environment.yml

Next, activate the newly created environment:

conda activate blackjack-dev

3. Finally, run the game package from the repository root:

python -m blackjack

If you're not familiar with the game of Blackjack, an excellent set of rules can be found here: https://bicyclecards.com/how-to-play/blackjack/

Tests

Unit tests use the pytest framework. Run all tests using:

python -m pytest

Extensions

• Extend gameplay with additional Blackjack actions: 'splitting pairs', 'doubling down', 'insurance', etc.
• Add support for multiple human players.
• Add a graphical user interface to the game.
• Refactoring:
  • Update the return values of some methods (especially on the hand classes/subclasses) to either return data that is then e.g. printed by the caller or to return a suitable status code. Aim to protect current data encapsulation and abstraction while facilitating the writing of better unit tests incl. initial-/final-state tests across single rounds with fully mocked-out player and dealer hands.
  • Extend inheritance and polymorphism to the card class? Aces have suitably different behaviour that they may benefit from specialised methods.
• Create StatJack: an agent that repeatedly plays rounds of the game with pseudo-random inputs, generating a large dataset of outcomes linked to initial/intermediate conditions and player actions.
• Train a ML agent to play (basing the reward function on long-term balance as opposed to rounds won); compare to results from StatJack visualising key decision pathways; trained agent could play alongside human players.

License

The code in this project is licensed under the MIT License. See LICENSE.txt for details.