Simplex: A Simple, Extensible Game for Exploring Machine Learning
As I’m exploring machine learning, I have found it helpful to have a simple game I can use as a context in which to test ideas. This blog post describes one such game. I call it “Simplex”. The game is trivially simple, but can be extended to make it more challenging.
The game works as follows:
- There are two players (‘a’ and ‘b’)
- Each player gets only one move which is made simultaneously with the other player
- Each player has one marble
- Each marble weighs one unit
- For each move, players choose how many marbles to place on his/her side of a scale
- The winner is the player with the most weight on their side of the scale
You can read, run, and edit an implementation of this game in Clojure (and Python) here.
In future posts, I’ll examine:
- Different approaches to playing this game (including machine learning strategies)
- How these approaches change when the game changes