Skip to main content

Fluxx for Robots


Fluxx is a mildly popular card game. All things considered, it’s actually fairly contentious. It has a score of 5.7 on Board Game Geek, but it has a serious cult following. This is a python implementation of version 5.0 of the Vanilla Fluxx card game. The full rules are available here. The core idea is that the players can play cards which change the rules of the game. Each turn, each player draws a certain number of cards and plays a certain number of cards. There are four types of cards. There are Keepers, which are laid on the table in front of the player, Goals which determine how the game is won, New Rules which change the rules of the game, and finally Actions, which are played to the discard pile and perform the action printed on the card.

This particular implementation of the game was built from the ground-up, using no third-party libraries or assets. Card text has been abridged slightly, but retains all of the original’s gameplay meaning.

Fluxx is a registered trademark of Looney Labs – College Park, MD 20741-0763

The goal of this project is to create a playground for advanced AI experiments with a complicated, not-analytically-solvable traditional game. Looney Labs has developed a polished and well-made digital version of Fluxx available here. Which is actually optimized for play by humans, and I suggest that anyone who found this project with the intention of playing the game for fun pay them the $0.99 and get the digital edition. That being said, there is a human-interface portion to this project which can be played by running

Installation and Use:

To play the game, all you need to do is run The construction is fully portable and relies on no third party packages.

For actually interesting uses of this project, a user needs to do the following.

  1. from fluxx.objects import Board
  2. Create an instance of Board, which I will call board.
  3. Use to get a dictionary which contains all of the player-knowable information about the game.
  4.['options'] will show a list of the choices that the player has at the given stage of the game.
  5. Run board.action(option) where option is the index of the desired option in['options'].
  6. Repeat 4 and 5 until a winner becomes apparent. The board.action() will throw a Board.Win exception which contains the number of the player who won.


Full Documentation of the engine and objects can be found at

Related Posts