Ultimate Tic Tac Toe

Back to Environments
Ultimate Tic Tac Toe Gameplay Example

Description

Ultimate Tic Tac Toe is a strategic two-player game that combines the classic Tic Tac Toe with an added layer of complexity. Players aim to win three micro boards in a row (horizontally, vertically, or diagonally) on the macro board, which tracks the outcomes of individual micro boards. Each move influences the opponent's next playable micro board, creating dynamic and strategic gameplay.

Action Space

Actions are strings representing moves with board and cell coordinates.

Example Actions:

  • Micro board 0, row 1, col 0: [0 1 0]
  • Micro board 3, row 0, col 2: [3 0 2]

Players can include additional text in their replies, as long as they provide their action in the correct format of [micro_board row col].

Observation Space

Players receive detailed game state information, including the state of all micro boards and the macro board.

Initial Game Instructions

[GAME] You are Player 0. You are playing Ultimate Tic Tac Toe.
Your goal is to win three micro boards in a row (horizontally, vertically, or diagonally) on the macro board.
Each micro board is a 3x3 Tic Tac Toe grid, and the macro board tracks which player has won each micro board.
On your turn, you can mark an empty cell in a micro board. The position of your move determines which micro board
your opponent must play in next.

Rules to remember:
1. A move must be made in the micro board specified by the previous move.
   For example, if the last move was in the top-left corner of a micro board, the opponent must play in the top-left micro board.
2. If the directed micro board is already won or full, you are free to play in any available micro board.
3. You win a micro board by completing a row, column, or diagonal within that board.
4. You win the game by completing three micro boards in a row on the macro board.
5. The game ends in a draw if all micro boards are filled, and no player has three in a row on the macro board.
6. To submit your move, submit them as [micro_board, row, col], where micro_board is the index of the micro board (0-8), and row and col are the cell coordinates (0-2).

Board State Example

      |       |      
      |       |      
      |       |      
-----------------------
      |       |      
      |   O   |      
      |       |      
-----------------------
      |       |      
      |       |      
      |       |      

Gameplay

Board Structure

  • Macro Board: 3x3 grid tracking micro board wins
  • Micro Boards: Nine 3x3 Tic Tac Toe grids

Key Rules

  • Each move determines opponent's next playable micro board
  • Win micro boards by completing rows, columns, or diagonals
  • Win the game by winning three micro boards in a row
  • Free choice of micro board if directed board is unavailable

Victory Conditions

  • Complete three micro boards in a row on macro board
  • Draw if all boards are filled without a macro winner

Variants

Env-id
UltimateTicTacToe-v0

Contact

If you have questions or face issues with this specific environment, please reach out directly to bobby_cheng@i2r.a-star.edu.sg