twilight-struggle-py ·

Project outline

We're creating a python port of Twilight Struggle which will be a barebones (command-line, no graphics) implementation of the game. The purpose of this project is to implement a reinforcement learning algorithm that would be a stronger computer player than the natively implemented one.

Getting started

Run the python file within a command window via python struggler.py, or open the Jupyter notebook struggler.ipynb.

The following base commands are provided:

Command	Description
new	Start a new game.
?	Displays help text.
s	Displays the overall game state.
m ?	Shows help on move queries.
s ?	Shows help on game state queries.
c ?	Shows help on card information queries.
dbg ?	Shows help on debugging.
rng on\|off	Toggles automatic random number generation (rng off for debugging).
commit on\|off	Toggles commit prompts.
log on\|off	Toggles game logging.
load <filename>	Loads from the log directory.
quit	Exit the game.
m	Lists all possible moves, along with their respective enum.
m <name\|enum>	Makes the move with the name or with the enum. The name can be abbreviated to the first characters as long as it is unambiguous.
m <m1 m2 m3 ...>	Makes multiple moves in order m1, m2, m3, ...
s <eu\|as\|me\|af\|na\|sa>	Displays the scoring state and country data for the given region.
c	Display a list of cards in the current player's hand.
c <name\|ID#>	Display information about the card with the given name or card index.
c opp	Returns the number cards in the opponent's hand.
c dis	Display a list of cards in the discard pile.
c rem	Display a list of removed cards.
c dec	Returns the number of cards in the draw deck.

This repo is a work-in-progress so expect a constant flow of updates.