Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/bsamseth/tic-tac-toe
Simple game playing AI for the game Tic-Tac-Toe.
https://github.com/bsamseth/tic-tac-toe
board-game minimax tic-tac-toe tictactoe
Last synced: 23 days ago
JSON representation
Simple game playing AI for the game Tic-Tac-Toe.
- Host: GitHub
- URL: https://github.com/bsamseth/tic-tac-toe
- Owner: bsamseth
- License: mit
- Created: 2018-02-05T10:55:18.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2023-05-09T06:37:17.000Z (over 1 year ago)
- Last Synced: 2024-10-26T14:06:46.666Z (2 months ago)
- Topics: board-game, minimax, tic-tac-toe, tictactoe
- Language: Python
- Size: 36.1 KB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Tic-Tac-Toe AI
This is a simple little game playing "AI" that can play Tic-Tac-Toe (Noughts
and crosses). There are several versions, written in Python, Rust, Java and C++. You
may play against either in a simple command line interface. You will never win,
however, as the AI never plays a move allowing you to do so.
This project was made as a simple demo of game playing computer programs,
illustrating an implementation of the [minimax algorithm](https://en.wikipedia.org/wiki/Minimax).
In addition, an implementation of [Monte Carlo Tree Search](https://en.wikipedia.org/wiki/Monte_Carlo_tree_search)
can be found in [mcts.py](mcts.py), showcasing a quite different approach.
## Run (Python)
To play against the [Python version](tictactoe.py) of the AI, simply run the
program in a terminal. The script allows you to chose the players, and you can
even have different AIs play each other.
Below is a short demo:
```
> python3 play.py
Who is player 1?
(1) You (human)
(2) MiniMax (AI)
(3) MCTS with random rollout (AI)
(3) MCTS with perfect rollout (AI)
Enter choice: 1
Who is player 2?
(1) You (human)
(2) MiniMax (AI)
(3) MCTS with random rollout (AI)
(3) MCTS with perfect rollout (AI)
Enter choice: 2
-|-|-
-----
-|-|-
-----
-|-|-
Enter move (1-9): 5
-|-|-
-----
-|X|-
-----
-|-|-
-|-|-
-----
-|X|-
-----
-|-|O
Enter move (1-9):
```
## Run (Rust)
To play against the [Rust version](tictactoe.rs) of the AI, you can use the provided [Makefile](Makefile):
```
> make run-rust
Best move found in an average of 0.28ms
---+---+---
| |
---+---+---
| |
---+---+---
| |
---+---+---
Enter move [0, 9):
```
## Run (C++)
To play against the [C++ version](tictactoe.cpp) of the AI, you can use the provided [Makefile](Makefile). This
implementation uses a [Bitboard](https://en.wikipedia.org/wiki/Bitboard) representation of the game, for variety. In
the case of Tic-Tac-Toe, it's slower than a simple array, but in games like chess they are essential.
```
> make run-cpp
Guaranteed outcome with optimal play (1 = Cross, -1 = nought, 0 = draw): 0
Found in average time out of 100 times: 0.90 ms
Don't trust the AI's determination? Go ahead, try to beat it. You can even go first.
-|-|-
-----
-|-|-
-----
-|-|-
Your move:
```
## Run (Java)
To play against the [Java version](tictactoe.java) of the AI, you can also use the provided [Makefile](Makefile):
```
> make run-java
Found best move in 1.00 ms on average of 100 times.
-|-|-
-|-|-
-|-|-
Your move:
```