Tic Tac Toe
- 8 Devlogs
- 14 Total hours
A Tic-Tac-Toe game featuring multiple bot difficulty levels, from random moves to an unbeatable Minimax AI and a reinforcement learning-based AI.
15m 7s logged
Fixed a bug inside my impossible level
Bug: My AI got stuck on “AI is thinking…”
Cause: I used grid instead of grids.Therefore, JS threw a reference error that grid is not defined and this error was stopping the function.
Fix: I have replaced grid with grids.
Lesson: When something appears stuck or not working as intended then check browser console first.
1h 28m 13s logged
I have changed my UI
Ship #1
I built a Tic-Tac-Toe game with 4 different AI bots, each with its own difficulty level.
The most challenging parts were building the Minimax AI and the reinforcement learning-based AI. I also spent a lot of time fixing bugs and handling different edge cases.
The game has different levels, starting with an easy bot that makes random moves and going all the way to an impossible bot trained using reinforcement learning.
Through this project, I learned a lot about reinforcement learning and solving problems with minimal use of external AI.
It was a fun project to build, and now it’s finally ready to be shipped!
2h 9m 2s logged
Completed the Impossible Level
Finally finished the impossible level using Q-learning. The bot learns from thousands of games and uses the trained Q-table to decide its moves instead of following fixed rules.
And with that, the Tic Tac Toe AI journey is complete. I started with simpler random moves and gradually moved to rule-based AI, Minimax and finally reinforcement learning.
Now it’s time to upload the project.
31m 39s logged
Hard Level is finally completed
I have upgraded the Tic-Tac-Toe AI from rule based decisions to Minimax. Instead of only looking at current board, the bot now stimulates possible future moves, considers how the player could respond, and chooses the move with the best possible outcome.
**Next: Impossible Mode (Reinforcement Learning)
3h 13m 36s logged
Finished the Medium Level bot
I manually built the bot’s decision making logic. It can now look for winning move, block opponents winning move, prioritize the center and corner and fallback to a random move when there is no better option.
Next: Hard Level (Minimax)
1h 20m 23s logged
🤖Easy Mode is now complete!
The bot makes random moves, with win and draw detection working properly.
Next up: Medium Mode — making the bot actually think. 🧠
1h 56m 9s logged
Added a Bot Selection area to the game!
Players can now choose different AI difficulty levels — from random moves to an unbeatable Minimax AI.
Next up: making each bot smarter 🚀
2h 52m 29s logged
Tic Tac Toe — Entrance Screen
I started building my tic-tac-toe game.
I created a responsive screen with full-screen background, glass-style card, title and play button.
Next: I will build the bot selection screen
