FreeCell: the thinking player's solitaire
Most solitaires make you gamble. Klondike hides half the deck face down and Golf can deal you a course with no path through it, but FreeCell puts all 52 cards on the table face up before your first move. There is no stock, no waste, no hidden information at all — just eight columns, four empty cells, and a position that is almost always solvable by someone patient enough to find the way. That's the game's famous promise and its quiet threat: when you lose at FreeCell, you usually can't blame the cards.
Where FreeCell comes from
FreeCell's ancestor is a suit-building game described by C. L. Baker and popularized when Martin Gardner featured it in his "Mathematical Games" column in Scientific American in June 1968. A decade later, medical student Paul Alfille made one change — building the columns in alternating colors instead of in suit, which makes far more deals solvable — and programmed the result in 1978 on PLATO, the University of Illinois' pioneering computer-education system. It was one of the first solitaires anyone ever played on a screen.
The version the world knows arrived when Jim Horne implemented FreeCell for Microsoft. It shipped in an Entertainment Pack before becoming one of the standard Windows games from Windows 95 onward, complete with its famous 32,000 numbered deals. In 1994, volunteers in the Internet FreeCell Project set out to solve every one of them by hand — and managed all but a single stubborn layout, deal #11982, which remains unsolvable to this day. That project gave FreeCell a reputation no other solitaire has: the game where winning is the norm and every loss demands an explanation.
The objective
Move all 52 cards onto the four foundations, one per suit, built upward from Ace to King. Win the maneuvering war in the columns and the foundations take care of themselves.
Setup
- The tableau is eight columns dealt from a single shuffled deck, entirely face up — the first four columns get seven cards, the last four get six.
- Four free cells sit above the tableau, each able to hold any single card, parked there for as long as you like.
- Four foundations start empty and collect each suit in order, Ace first.
How moves work
- Build the columns down in alternating colors. A card moves onto a card one rank higher of the opposite color — a red 6 on a black 7.
- Any card can go to an empty free cell, and a card in a cell can come back down onto any legal column (or go straight to its foundation). Cells hold exactly one card each.
- Any card can start an empty column. Unlike Klondike, you're not restricted to Kings — an empty column is a wildcard space, the most valuable real estate in the game.
- Aces go up as they're freed, and each foundation builds upward in its own suit. Double-click a card here to send it to its foundation.
- Formally, you move one card at a time. Moving a whole ordered run is really a chain of single moves through cells and empty columns, so the number of cards you can relocate at once is limited: (1 + empty free cells) × 2 for each empty column. With all four cells free and no empty columns you can move a run of five; fill your cells and that shrinks to one. This board performs the shuffling for you as a single "supermove," but it enforces the same limit — and moving into an empty column costs you that column's doubling.
Strategy that actually matters
- Plan before you touch anything. All the information is on the table, so the opening minute of a FreeCell game is reading, not moving. Find the Aces and 2s, see what's on top of them, and work out the first few exchanges before you start.
- Free the Aces and 2s early. Low cards buried under three layers of face cards are the deals that go wrong. Most of your early moves should be excavation work toward them.
- Treat free cells as scaffolding, not storage. Every occupied cell cuts your biggest-move size and your options. Park a card only when you know how it's coming back down — a full set of cells with no plan is how FreeCell games die.
- Work toward an empty column. An empty column is strictly better than a free cell: it holds a whole run, doubles your supermove capacity, and accepts any card. Emptying one column, even at some cost, often cracks the whole deal open.
- Don't strip one suit to the foundations too fast. You may still need that black 5 in play to hold a red 4. Sending cards up the moment they're legal can leave the tableau without the in-between cards it needs; keep the foundations roughly level until the endgame is clearly won.
- Build on high cards reluctantly. A run stacked on a King goes nowhere until the game is nearly over. When you have the choice, build on cards you'll eventually be able to move as a unit.
Common variations
Baker's Game is the original: identical layout, but columns build down in suit rather than alternating colors — noticeably harder, with around one deal in four unsolvable. Eight Off, an older cousin, deals eight cells and also builds in suit. Relaxed FreeCell drops the supermove limit and lets any ordered run move regardless of free space, which turns near-losses into wins and suits players who find the capacity rule fussy. You'll also meet two-cell and one-cell FreeCell — the standard game with fewer cells, played as a deliberate challenge by players for whom four cells has become too safe. This page plays the classic rules: four cells, alternating colors, honest supermove math.
Frequently asked questions
Is every FreeCell deal winnable?
Very nearly. Of Microsoft's original 32,000 numbered deals, exactly one — #11982 — is impossible, and computer analysis of much larger deal sets puts the unsolvable rate around one in 80,000. This page deals a fresh random shuffle each game, so the deal in front of you is almost certainly solvable. Whether it gets solved is another matter: skilled players win the large majority of their games, but everyone walks into a dead end sometimes.
Why won't the game move my whole stack at once?
Moving a run is shorthand for moving its cards one at a time through spare cells and empty columns, so the size of the run you can move is capped at (1 + empty free cells) × 2 per empty column. If the board tells you there isn't enough room, it's not being pedantic — there genuinely is no sequence of single moves that would relocate that stack. Free up a cell or clear a column and try again.
Can you actually lose at FreeCell?
Yes, in two ways. The obvious one is gridlock — all four cells full and no legal move anywhere, which this board announces when it happens. The subtler and far more common one is reaching a position that's merely unwinnable: moves still exist, but the low cards you need are buried past retrieval. Undo is your friend in both cases; stepping back three moves and choosing differently is a legitimate part of learning the game.
How is FreeCell different from Klondike?
Klondike is a game of managed luck: half the deck is face down and the stock feeds you cards in an order you can't control. FreeCell is a game of pure planning — nothing is hidden, nothing is random after the deal, and there's no stock at all. If Klondike is poker, FreeCell is chess: the deal decides the puzzle, not the outcome.
What counts as a good game?
Winning at all is the real score, and a long-term win rate above 80% marks a strong player. If you want a second axis, watch your move count: tight play solves most deals in well under 120 moves, while a 200-move win usually means the deal got solved twice — once wrong, then once right. Speed matters least; FreeCell rewards the player who stares.