A quick prototype of a Tic-Tac-Toe game, utilising bitwise operations for efficient, low-level game state representation and outcome resolution - a performance-conscious approach within a high-level engine (Unity).
Each cell in the 3 x 3 Tic-Tac-Toe grid is assigned a unique binary using powers of 2 (bit flags). The flags are mapped as follows:
Index: 0 1 2 3 4 5 6 7 8
Binary: 256 128 64 32 16 8 4 2 1
Board:
0 1 2
3 4 5
6 7 8
The formula for conversion is: flag = 2 ^ (8 -n), where n is the cell index (0 to 8, left to right, top to bottom).
Winning patterns are precomputed using bitwise masks for fast evaluation. A win is determined by comparing a player's current cell state with each winning mask using bitwise AND.
Visual examples of win patterns:
possible wins
x x x
o o o
o o o
outcome: x x x o o o o o o = 448
o o o x x x o o o = 56
o o o o o o x x x = 7
x o o
x o o
x o o
outcome: x o o x o o x o o = 292
o x o o x o o x o = 146
o o x o o x o o x = 73
x o o
o x o
o o x
outcome: x o o o x o o o x = 273
o o x o x o x o o = 84
Win Detection Code (Unity C#):
bool complete = false;
...
...
...
int[] win_masks = { 448, 56, 7, 292, 146, 73, 273, 84 }; // win masks
int curr_player_slot = (curr_player == 0) ? m_Player0Slot : m_Player1Slot;
for (int i = 0;i < win_masks.Length;i++)
{
complete |= ((curr_player_slot & win_masks[i]) == win_masks[i]);
//if a win condition is met, handle game reset
if(complete)
{
m_WinningPlayerIdx = curr_player;
m_ResetGame = true;
return win_masks[i];
}
}
 |
|---|
A Youtube video of in-game sample. |