Nim

My solution for the Nim project of the CS50AI course.

About the project

CS50AI exercise link: https://cs50.harvard.edu/ai/2020/projects/4/nim/

The play.py file and the signature of the functions of the nim.py file were already created by the CS50AI staff, so my task as a student was to create an AI capable of playing the Nim game. Using the method of reinforcement learning, making the AI play against itself 10000 times.

Rules of the game

In the Nim game, you have 4 piles with sticks:

• Pile 0: 1 stick
• Pile 1: 3 sticks
• Pile 2: 5 sticks
• Pile 3: 7 sticks

The two players have to remove a certain number of sticks from one pile every turn, the player to remove the last stick will be the loser.

Try it yourself

To run the game yourself, just run the following commands:

git clone https://github.com/BrenoMorim/nim.git nim
cd nim
python play.py