This is Tower of Hanoi in card form — plan three moves ahead, not one.
Tower of Pisa inverts every instinct from standard solitaire: only the bottom card of each column moves, cards must be placed beneath higher-rank cards, and the goal is to consolidate everything into one column in order. The puzzle logic is identical to the classic Tower of Hanoi peg-and-disk problem, and the same strategic principle applies — smaller pieces must be temporarily relocated to expose larger ones.
Last updated: June 2026
The Tower of Hanoi connection
Tower of Pisa Solitaire (also known as Tower of Hanoi Solitaire) is a direct card adaptation of the Tower of Hanoi mathematical puzzle, first described by French mathematician Édouard Lucas in 1883. In the original puzzle, disks of different sizes sit on pegs and must be moved one at a time, with no larger disk ever placed on a smaller one. The goal is to transfer the entire stack from one peg to another using a third as a staging area.
The card version preserves the core constraint: only one card (the “disk”) moves at a time, and a card placed in a column must be of lower rank than the card it lands beneath. The columns are the pegs. The goal — one complete ordered stack — mirrors the original puzzle’s goal exactly.
Tower of Pisa uses Ace through 9 (36 cards) across four suits. The final tower must have Ace at the bottom through 9 at the top. Because only bottom cards move, this means building from 9 down: first get all 9s consolidated (with the target 9 at the column bottom), then work down to 8, 7, 6, and so on.
The inverted access model
Every other solitaire game gives access to the top of a column. Tower of Pisa gives access only to the bottom. This inversion has a profound effect on how you read the board:
- Cards at the bottom of a column are the most accessible — they can move immediately if a legal destination exists.
- Cards at the top of a column are the least accessible — every card beneath them must be removed first to expose them as a new bottom card.
- A “deep” column (many cards) is not inherently bad. A column with a 9 at the top and eight lower cards below it will eventually deliver those lower cards as the column empties from the bottom. A column with a 2 at the top is temporarily blocked until the 2 moves.
Scan column bottoms, not column tops. The available moves are the bottom cards of all columns. For each bottom card, identify every column it could legally move to (where the column bottom is a higher rank). That set of possible moves is your entire decision space.
Building from the largest card down
The winning sequence is always built from the top of the target range downward. Since the final tower needs 9 at top and Ace at bottom, and only bottom cards move, the 9 must be the last card to reach the bottom of the final column. This means:
- Get all 9s to one column. The target 9 (whichever you designate as the foundation of the final tower) must be at the bottom of that column. Stack the other three 9s beneath it temporarily — they will need to move later.
- Work down to 8s.Once the 9 column is established, move 8s beneath the 9 (the 9 becomes a valid destination because 8 < 9). The target column grows from the bottom: 9 at top, 8s below, then 7s below those.
- Continue through Ace. Each rank is consolidated below the previous in succession. Aces are last because they are rank 1 — the smallest — and must be placed at the very bottom of the final tower.
The Tower of Hanoi algorithm for this is recursive: to move a stack of N cards from one column to another, move the top N-1 cards to a staging column, move the N-th card to the destination, then move the N-1 stack on top. Translating this to card terms: to reach the 8 buried two cards above the 9 in a column, first move the two blocking cards to staging columns, move the 8, then move the staging cards back.
Empty columns as staging areas
Empty columns are the critical resource in Tower of Pisa, exactly as empty pegs are in Tower of Hanoi. Without empty columns, circular dependencies can make progress impossible.
An empty column can receive any single card without rank restriction. This makes it a universal staging area: move the card that is blocking access to your target into the empty column, complete the target operation, then move the staged card to its new home.
Managing empty columns requires planning. If you use an empty column as temporary storage for a low-rank card, you must have a legal destination for that low-rank card before you can free up the empty column again. Specifically, a low-rank card in an empty column can only move to a column whose bottom card is of higher rank. If no such column exists, the card sits in the empty column indefinitely, consuming that staging resource.
You need to move the 6 that is currently at the bottom of column three (available to move) to the bottom of column one, where the bottom card is a 7. But column three’s bottom card is actually a 3, blocking the 6. The 3 needs to move, but no column bottom is higher than 3 (rank 1 Ace is lower; rank 4 Four is higher — but it’s not at a column bottom right now).
Move the 3 to the empty column (legal — empty columns accept any card). Now the 6 is the new bottom of column three. Move the 6 to column one (legal — 6 < 7). The 3 in the empty column can now move to the bottom of column three where the next card up is a 4 (which is > 3). If no such column exists for the 3, you must plan further ahead before taking this sequence.
Look-ahead planning
Because each move can create circular dependencies (moving A to B, then needing A again immediately to move C to the column A just vacated), Tower of Pisa requires looking at least three moves ahead before committing.
A useful discipline: before any move, identify what the state of the board will be after that move and the two most likely follow-up moves. If that three-move sequence leads to a position where no legal move exists, the first move is likely wrong.
The specific danger to watch for is the “single move available” trap: a position where exactly one card is available to move and exactly one destination exists, but that forced move leads to another forced move that undoes the progress just made. This loop structure is recognizable early: if you see two columns with the same two cards at their bottoms in alternating order, you are likely approaching a forced-loop position.
Frequently asked questions
What if the bottom card of a column has no legal destination?
If a column’s bottom card rank is higher than every other column’s bottom card (or there is only one column with no destination), the card cannot move unless an empty column is available. Empty columns accept any rank, so a card with no other destination can always go to an empty column — provided one exists.
Can I move a card to its own column?
No. A card must move to a different column. It cannot be placed back under itself.
Does the suit of the final tower matter?
No. The final tower needs Ace at the bottom through 9 at the top in a single column, but there is no suit requirement. Any suits in any combination can occupy the tower as long as the rank order is correct (lower ranks toward the bottom, Ace at the bottom, 9 at the top). Each rank has four copies across four suits, and all four must appear in the final column.
Is every starting deal solvable?
Not always. Some initial arrangements create dependency loops that cannot be resolved even with perfect play. The frequency of unwinnable deals depends on the initial column distribution. Deals where the highest ranks (8s and 9s) are scattered across many columns in ways that prevent consolidation tend to be hardest.