Shakashaka Strategy Guide: Master the Triangle Puzzle in 7 Days

Shakashaka is played on a rectangular grid. Some cells are black (walls) and cannot contain triangles. The remaining white cells are where the puzzle takes place. In each white cell, you may place a triangle in one of its four corners, or leave it empty. A triangle occupies exactly half of a cell, filling a right triangle from one corner. The four possible orientations point to the top-left, top-right, bottom-left, or bottom-right corner. When you place a triangle, the other half of that cell becomes part of the white (unshaded) space. The goal is to place triangles so that every connected region of white space (the halves of cells not occupied by triangles, plus any fully empty cells) forms a perfect rectangle or square. Black cells and triangles act as boundaries between these rectangular white regions. Numbered clue cells (which are always black) tell you exactly how many of their four orthogonal neighbors contain a triangle.

Clue numbers appear in black cells and tell you how many of the four orthogonal neighbors (up, down, left, right) contain triangles. A clue of 0 means none of the four neighbors has a triangle, so all adjacent cells are either empty or black. A clue of 4 means all four neighbors contain triangles. Intermediate clues require more thought. A clue of 1 means exactly one of the four neighbors has a triangle, but you do not know which one without additional information. However, if three of the four neighbors are already known to be empty (perhaps because they are forced by other constraints), then the fourth must contain the triangle. Clue of 2 is particularly common and often the most informative when combined with spatial analysis. If two neighbors of a "2" clue are already known to contain triangles, the other two must be empty. Conversely, if two neighbors are known to be empty, the remaining two must contain triangles. Corner and edge clue cells have fewer than four neighbors, which makes their clues even more constraining.

Begin by finding clue cells that immediately force their neighbors. A clue of 0 next to the grid edge means all its neighbors are empty. A clue of 4 forces triangles in all four neighbors. Process all of these forced placements first, as they create the foundation for further deduction. Next, look for **rectangularity constraints**. If a white region is forming near a black cell or the grid edge, analyze what shape it needs to become rectangular. A white area that is 3 cells tall and 2 cells wide must remain exactly that shape: any additional white cell appended to the wrong side would break the rectangle rule. **Edge analysis** is another powerful technique. Along the grid border, white cells are already bounded on one side. If a white cell on the edge cannot form a rectangle without a triangle in an adjacent cell, that triangle placement is forced. Work the edges of the grid early, as they provide the tightest constraints.

The **rectangle deduction method** is the most powerful advanced technique. When you identify a partially formed white region, calculate what rectangle dimensions would complete it. If only one rectangle shape is possible, every cell in that rectangle is determined. If two shapes are possible, look at the boundary cells they share: those cells must be white regardless of which shape is correct. **Parity analysis** applies to larger grids. Consider the total area of white space in a region and the possible rectangle dimensions that could fill it. A region with an odd-numbered total area cannot be filled by 2-cell-wide rectangles, which constrains the possible orientations of triangles around its boundary. Finally, **global connectivity checking** ensures you have not created isolated white cells or L-shaped regions. Periodically scan the grid for white cells that do not belong to any rectangular region. These indicate that a triangle placement nearby is incorrect or that additional triangles are needed to complete the rectangle.