HOW VISIBILITY WORKS
Each row and column in a Skyscraper puzzle contains the digits 1 to N exactly once (like a Sudoku Latin square). The number on the edge is a "visibility clue" that tells you how many buildings are visible from that position, looking along the row or column.
A building is visible if no taller building stands between it and the viewing position. The tallest building (N) is always visible. A building shorter than all previous buildings is never visible. The sequence's left-to-right (or top-to-bottom) nature is the key: visibility depends on position, not just height.
Example for a 4×4 grid (digits 1–4): the sequence 3, 1, 4, 2 has visibility count 3 (3 is visible, 1 is hidden by 3, 4 is visible, 2 is hidden by 4). The sequence 4, 3, 2, 1 has visibility count 1 (only 4 is visible; everything decreases behind it). The sequence 1, 2, 3, 4 has visibility count 4 (every building is visible; each one is taller than the last).
THE EXTREME CLUES: 1 AND N
The most powerful clues are the extreme values: 1 and N (the grid size).
Clue = 1: Only one building is visible — which means the tallest building (N) must be in the first position (closest to the clue). Anything behind it is hidden. This forces the first cell to contain N immediately.
Clue = N: Every building is visible — which means the sequence must be strictly increasing from the clue's perspective. The only strictly increasing arrangement is 1, 2, 3, … N in order. This forces the entire row or column immediately.
Always start with 1-clues and N-clues. They are immediate forced placements. On a 5×5 grid, a clue of 5 fills the entire row; a clue of 1 places the 5 in the first cell.
HIGH CLUES: FORCING THE LARGEST BUILDINGS
A clue of N−1 means all buildings except one are visible. For this to work, the tallest building (N) must be in position 2 or later — if it were in position 1, the clue would be 1. Specifically, N cannot be in the first position, and the second-tallest building (N−1) must be in position 2 (otherwise the clue would be N−2 or less).
More generally, high clue values force the large buildings toward the near end of the row. For a clue of V (where V is large), the N must appear in one of the first V positions, and the sub-sequence of buildings from position 1 to position N's location must be strictly increasing.
This gives you a range constraint on the position of N: clue V means N can't be beyond position V. If V = 3, N must be in positions 1, 2, or 3. Combined with the rule that N is always visible, this directly constrains where the largest building can be placed.
LOW CLUES: WHAT THEY TELL YOU ABOUT THE FRONT
Low clues (2, 3, …) constrain the near positions differently. A clue of 2 means exactly two buildings are visible. The tallest building hides everything behind it. Arrangements that satisfy clue = 2 all share a structure: the tallest building is relatively close to the front, and whatever is in front of it is shorter.
For clue = 2 on a 5×5 grid, the valid placements of N are positions 2 through 5 — but N must be preceded by exactly one visible building, meaning N is the second visible one. If N is in position 2, position 1 can be anything (1, 2, 3, or 4). If N is in position 3, positions 1 and 2 must have a "visible then hidden" structure.
In practice, use low clues to eliminate positions for N (if N were in position 1, the clue would be 1, not 2 — contradiction). And use them to check whether a proposed arrangement satisfies the clue count before committing.
COMBINING ROW AND COLUMN CLUES: THE INTERSECTION TECHNIQUE
Every cell belongs to a row and a column, each with clues on both ends. Cells near corners are doubly constrained — they contribute to both a row constraint and a column constraint simultaneously.
The intersection technique: for each cell, list the valid values from its row constraint, list the valid values from its column constraint, and take the intersection. If only one value appears in both lists, that cell is forced.
This is most powerful for corner cells (constrained by two edge clues each). In a 5×5 grid, the cell at row 1, column 1 is affected by the top clue for column 1, the left clue for row 1, plus the Latin-square constraint. Three constraints on a single cell often combine to force it to a single value.
THE SUDOKU CONSTRAINT AS A TOOL
The underlying Latin-square rule — each digit appears exactly once in each row and column — is as important as the visibility clues. Don't forget to apply standard Sudoku elimination alongside the visibility reasoning.
Once a clue places the digit N in a specific cell, that digit is eliminated from all other cells in the same row and column. This immediately restricts what other cells can contain and cascades into new visibility deductions.
The most effective Skyscraper solving alternates between two modes: (1) visibility reasoning — using clue values to constrain which positions are valid for each height; and (2) Latin-square elimination — using placed digits to rule out values elsewhere. Both modes inform each other, and the cascade is what makes Skyscraper puzzles feel elegant when they resolve.
A PRACTICAL SOLVING ORDER
- Step 1: Process clue = 1 (force N to the nearest cell) and clue = N (force entire sequence).
- Step 2: Process clue = N−1 (N cannot be in position 1; N−1 is in position 2 or near).
- Step 3: For each placed digit, apply Latin-square elimination across its row and column.
- Step 4: Use intersection technique on corner cells — combine row and column clue constraints to narrow down valid values.
- Step 5: For remaining unsolved cells, enumerate valid arrangements for each partially-filled row/column and find the overlap — cells in the same position in every valid arrangement are forced.
Steps 1–3 resolve most easy and intermediate Skyscraper grids. The intersection technique and enumeration are the tools for hard grids. The key is resisting the urge to guess — every Skyscraper placement should be derived from a specific combination of clue values and Latin-square constraints.