Discover why nearly every FreeCell deal is solvable and what makes it one of the most winnable solitaire games.
FreeCell occupies a singular position in the solitaire catalogue: it is the only mainstream variant for which the solvability statistics are exact rather than estimated. Every other variant's winnability rate is a probability range derived from sampling — a large but finite number of randomly generated deals are tested by automated solvers, and the proportion that are winnable is reported as an estimate with statistical uncertainty. FreeCell's winnability statistics are exact because the original Microsoft implementation used a specific numbered deal set of 32,000 deals (numbered 1 through 32,000 in the original Windows FreeCell), and exhaustive computational analysis has individually tested every deal in that set, confirming with certainty which are winnable and which are not. The answer: exactly 8 of the 32,000 are unwinnable — an unwinnable rate of 0.025%. The remaining 31,992 are all winnable by at least one legal move sequence.
FreeCell occupies a singular position in the solitaire catalogue: it is the only mainstream variant for which the solvability statistics are exact rather than estimated. Every other variant's winnability rate is a probability range derived from sampling — a large but finite number of randomly generated deals are tested by automated solvers, and the proportion that are winnable is reported as an estimate with statistical uncertainty. FreeCell's winnability statistics are exact because the original Microsoft implementation used a specific numbered deal set of 32,000 deals (numbered 1 through 32,000 in the original Windows FreeCell), and exhaustive computational analysis has individually tested every deal in that set, confirming with certainty which are winnable and which are not. The answer: exactly 8 of the 32,000 are unwinnable — an unwinnable rate of 0.025%. The remaining 31,992 are all winnable by at least one legal move sequence.
This exactness has a compounding consequence: because each of the 32,000 deals has been individually solved, the specific winning sequences for all 31,992 winnable deals are known and catalogued. The solvability data for FreeCell is therefore not a population statistic but a complete census — not "approximately X% are winnable" but "these specific deal numbers are winnable and these eight are not." No other mainstream solitaire variant has this property. Klondike's winnability requires probabilistic estimation because face-down cards create hidden information that makes complete exhaustive analysis of all possible deals computationally intractable. Spider's winnability is similarly estimated from large samples. FreeCell's complete information — all 52 cards visible from the start — makes every deal fully enumerable, which is what makes the exact census possible.
Understanding FreeCell's statistics in detail serves three purposes for the practical player. First, it calibrates win rate expectations precisely: a player who knows their FreeCell win rate over 200 games can assess exactly how close they are to the theoretical 99.975% ceiling (the winnability floor), how much of the gap is strategy rather than deal mathematics, and what win rate improvements are realistically achievable through further strategy development. Second, it contextualises the difficulty of the eight known unwinnable deals: when one of these deals appears by number, the player can immediately recognise that continued play is futile — not because the position looks stuck but because the deal's unwinnability has been confirmed by exhaustive analysis. Third, it illustrates the relationship between complete information, computational tractability, and statistical precision that distinguishes FreeCell from all other mainstream variants.
FreeCell is an eight-column solitaire variant with four free cells — temporary holding positions for single cards — and four foundation piles that must be built from Ace to King in suit. All 52 cards are dealt face-up into the eight columns at the start, giving the player complete information from the first move. The win condition is moving all 52 cards to the four foundations in correct suit and rank order. The free cells and empty columns serve as staging resources: cards can be temporarily moved to free cells to access the cards below them, and empty columns can hold any card or partial sequence temporarily.
The statistical properties that make FreeCell uniquely analysable flow directly from its rules. Complete information (all cards face-up) means the initial state is fully specified — there is no uncertainty about hidden cards to branch over. The move set is deterministic — given a board position, every legal move is enumerable. And the state space, while large, is small enough that depth-first search can exhaustively explore it for any specific deal within seconds on modern hardware. These three properties — complete information, deterministic moves, tractable state space — are the mathematical prerequisites for exact solvability analysis. Klondike lacks the first property; its state space per deal is not fully determined by the visible cards. Spider 4-Suit lacks the tractable state space property at high difficulty; its branching factor makes exhaustive search impractical for many deals. FreeCell uniquely satisfies all three.
Within the original 32,000-deal Microsoft numbered set, exactly eight deals are unwinnable: deals 11,982 and 146,692 are the most frequently cited (the latter from an extended set), along with deals 164, 166, 454, 655, 1,021, and 6,469 in some versions of the numbered set, with the exact list varying slightly between implementations. The unwinnable deals share a structural property: each contains a confirmed circular dependency — a configuration where at least two cards block each other's movement in a cycle that no legal move sequence can break, given FreeCell's rules. Specifically, the four free cells plus eight columns do not provide enough staging capacity to untangle the mutual blockage in these eight deals.
The deals' unwinnability has been independently confirmed by multiple solvers using different search algorithms, all reaching the same conclusion: no legal move sequence from the initial position leads to a state where all 52 cards are on the foundations. This independent confirmation from multiple solvers using different approaches is what makes the unwinnability verdict certain rather than probable — it is not a statistical estimate but a logical proof by exhaustion. The same exhaustive method that confirms these eight deals are unwinnable also confirms that the other 31,992 deals are all winnable — for each of these, at least one winning path has been found and verified.
For practical play, the implication is straightforward: if a player is using a numbered deal system and encounters deal number 11,982, continued play after confirming the circular dependency is genuinely pointless — not because the position looks difficult, but because the impossibility has been mathematically verified. This is different from the situation in Klondike, where a stuck position might be an unwinnable deal or might be a difficult-but-winnable deal that requires a non-obvious recovery sequence. In FreeCell, the eight unwinnable deal numbers are known, and any other numbered deal is confirmed winnable.
Use the 99.975% winnability floor as a strategy calibration benchmark. If a player's FreeCell win rate over 200 games is 70%, that win rate is 29.975 percentage points below the winnability ceiling. Since fewer than 0.001% of randomly shuffled FreeCell deals are unwinnable, essentially the entire 30% gap is accounted for by strategy failures — games that were winnable but were played to a stuck position through suboptimal move sequencing, free cell mismanagement, or rushed foundation building. This is the most direct application of FreeCell's statistical exactness: the gap between any player's observed win rate and 99.975% is an upper bound on the strategy improvement available, and that improvement is accessible because the deals are winnable. No other variant provides this degree of precision in strategy gap measurement — in Klondike, a portion of every player's losses are intrinsically unwinnable deals that cannot be attributed to strategy; in FreeCell, almost none are.
Track free cell usage as a leading indicator of strategy quality. FreeCell's statistics show that the winning paths for most deals use the free cells as temporary staging positions that are vacated within two to four moves — not as long-term parking for cards that have no immediate home. A player whose average free cell peak occupancy per game is three or four (all cells filled simultaneously) is likely experiencing free cell depletion traps — positions where all four cells are occupied and no legal move can empty any of them, leading to a stuck board. The statistical benchmark: expert FreeCell play reaches four-cell simultaneous occupancy rarely and recovers from it quickly; casual play reaches it frequently and often cannot recover. Monitoring free cell peak occupancy as a per-game metric identifies whether free cell depletion is the primary driver of losses — and if it is, the free cell rationing discipline (never fill the fourth cell without a specific three-to-five move plan for vacating it) is the highest-leverage single strategy adjustment available.
Use solution count variation to calibrate position tightness. Among the 31,992 winnable FreeCell deals, the number of distinct winning paths varies enormously — some deals have thousands of winning move sequences, while others have only one or a handful. Deals with many winning paths are forgiving: multiple different approaches all reach the win state, and minor sequencing errors can be recovered through alternative paths. Deals with very few winning paths are unforgiving: a single wrong early move may foreclose all remaining winning paths, leaving a stuck position that is not an unwinnable deal but is an effectively stuck position for a player who does not know the specific path required. The practical implication: when a FreeCell position develops that feels unusually constrained — where every candidate move seems to create a new problem — the correct diagnosis is not "this deal is unwinnable" (it almost certainly is not) but "this deal has a small solution count and the winning path is narrow." The response is more careful positional analysis, not resignation. Extended undo-based hypothesis testing on positions with narrow solution counts is the correct strategy response — and the algorithmic framing in the algorithms guide explains why: a deal with few winning paths requires deeper search to find any of them.
Apply the statistical insight about multiple winning paths to undo strategy. The fact that 31,992 deals all have at least one winning path — and most have many — means that in any FreeCell game that has not yet reached a confirmed dead-end position (circular dependency confirmed, no legal moves remaining), a winning path exists. This is categorically different from the situation in Klondike or Forty Thieves, where a stuck position might genuinely have no winning path because the deal is intrinsically unwinnable. In FreeCell, a stuck position that has not confirmed a circular dependency is almost certainly a position where the winning path has not yet been found — which means the correct response to any stuck FreeCell position (except confirmed dead-ends) is more systematic positional analysis using undo-based branching, not resignation.
Treating any stuck FreeCell position as unwinnable. Given that only eight deals in the numbered set are unwinnable, and that randomly shuffled FreeCell deals are unwinnable with probability below 0.001%, almost every stuck FreeCell position is a winnable deal that the player has played into a difficult branch of the move tree — not an intrinsically unwinnable deal. The correct response to a stuck FreeCell position is to apply the three-pattern structural diagnostic (circular dependency check, key card burial assessment, resource exhaustion confirmation) before resigning. If no structural blockage pattern is confirmed, the position is almost certainly still winnable, and the stuck feeling reflects a planning horizon limitation rather than deal structure. Extended undo-based backtracking is the correct tool for navigating out of the stuck branch — not resignation.
Filling all four free cells before the endgame. FreeCell's statistics on free cell usage show that free cell depletion — filling all four simultaneously with no plan to vacate them — is the single most common proximate cause of stuck positions in losing games. The four free cells are staging resources, not parking spaces: their value comes from being available for future use, not from the immediate convenience of parking a card that has no immediate home. Every card placed in a free cell without a specific plan for its destination within three to five moves increases the probability of free cell depletion and, therefore, of a stuck position. The statistical consequence of free cell mismanagement is not immediately visible — the board does not lock immediately when the fourth cell is filled — but it forecloses future move options progressively until the stuck position becomes unavoidable.
Ignoring the solution path in favour of the shortest visible path. Because FreeCell's statistics confirm that all but eight numbered deals are winnable, there is always a winning path from the starting position. But the winning path is not always the most visually appealing path — it is frequently the path that requires accepting a temporarily worse-looking position (more cards in free cells, no foundation progress for several moves, an empty column filled before it was optimal) in exchange for a better structural position several moves later. Players who optimise for the most immediately appealing path — the one that moves the most cards to foundations in the fewest moves — frequently find that this path leads to a free cell depletion trap or an empty column shortage in the endgame, converting a winnable deal into a player-limited stuck position. The statistical fact that the deal is winnable does not mean every path through it is winning; it means at least one path is, and finding that path may require counter-intuitive intermediate moves.
Pyramid Solitaire provides the sharpest statistical contrast with FreeCell: where FreeCell has fewer than 0.001% unwinnable deals, Pyramid has approximately 30–50% unwinnable deals — a difference of five orders of magnitude in intrinsic difficulty. A player who tracks win rates in both games simultaneously develops an immediate intuition for the difference between deal-structure challenge (Pyramid) and strategy-skill challenge (FreeCell). A session of 20 Pyramid games where the player wins 8 and a session of 20 FreeCell games where the player wins 8 look identical in the score column but represent very different situations: the Pyramid player may have played close to optimally on a difficult deal sample, while the FreeCell player almost certainly has significant strategy improvement available given the near-100% winnability of the deals. TriPeaks Solitaire provides a third reference point: TriPeaks has a high winnability rate (approximately 80–90%) and a high strategic win rate (75–85%), which is closer to FreeCell's statistical profile than Pyramid's but still includes a meaningful unwinnable deal population that FreeCell effectively lacks. For the complete cross-variant win rate context, see our win odds comparison guide. For the algorithmic framework that made FreeCell's exact solvability statistics possible, see our algorithms guide.
What is the best strategy for maximising FreeCell win rate given its statistics?Three strategy habits produce the largest win rate improvements relative to FreeCell's statistical ceiling. The first is free cell rationing: never fill the fourth free cell without a specific three-to-five move plan for vacating it, because free cell depletion is the most common proximate cause of stuck positions in losing games. The second is undo-based hypothesis testing rather than resignation on stuck positions: given that only eight numbered deals are unwinnable, any stuck position that has not confirmed a circular dependency is almost certainly still winnable, and the correct response is deeper positional analysis via undo branching rather than resignation. The third is counter-intuitive path acceptance: the winning path on narrow-solution-count deals frequently requires moves that look worse in the short term — accepting this and evaluating moves by their structural consequences three to five moves ahead rather than their immediate appearance is the highest-level FreeCell strategy application and the primary determinant of performance above 80% win rate.How many FreeCell deals are unsolvable?In the original Microsoft 32,000-deal numbered set, exactly 8 deals are unsolvable — an unsolvable rate of 0.025%. The best-known are deals 11,982 and 146,692. Extended numbered sets and randomly shuffled implementations produce unwinnable deals at a rate below 0.001% — fewer than one deal in every thousand. This rate is the lowest in the mainstream solitaire catalogue by a factor of hundreds: Klondike has an unwinnable rate of approximately 9–21%, Spider 4-Suit approximately 45–60%, and Forty Thieves approximately 40–60%. FreeCell's near-zero unwinnable rate is the property that makes it uniquely suitable as a skill measurement environment — essentially all variation in win rate across players and sessions is attributable to strategy quality rather than deal mathematics.Can every FreeCell game be solved with the right strategy?Almost every FreeCell game can be solved with correct strategy — specifically, all but the eight confirmed unwinnable deals in the standard numbered set and the equivalent rare cases in randomly shuffled implementations. This is categorically different from all other mainstream solitaire variants, where a non-trivial proportion of deals are intrinsically unwinnable regardless of strategy quality. The practical interpretation: when a FreeCell player loses a game that was not one of the eight known unwinnable deal numbers, the loss is attributable to strategy rather than deal mathematics. This is simultaneously the most demanding and the most instructive aspect of FreeCell statistics: every loss except the eight known exceptions is a diagnosable strategy failure with a specific cause — free cell mismanagement, premature foundation racing, rushed endgame sequencing, or insufficient undo-based exploration of the winning path — and therefore a specific strategy lesson that can be extracted and applied to future games.
FreeCell is unique because it is the only solitaire variant with exact solvability statistics. Unlike other solitaire games, where winnability rates are estimated based on sampling, every FreeCell deal is theoretically solvable. This means that if you play optimally, you can win almost every game, as long as you encounter one of the eight known unwinnable deals. This precision allows players to develop strategies based on the certainty that most games can be won.
In FreeCell, each player has eight tableau piles, four free cells, and four foundation piles. The main rules include: you can move cards between tableau piles if they are in descending order and alternating colors; you can only move one card to a free cell; and you can only move a stack of cards to an empty tableau if it consists of a single card. Additionally, you can only build foundations in ascending order by suit. Understanding these rules is crucial for developing effective strategies.
Common mistakes in FreeCell include mismanaging free cells, which should be used strategically to temporarily hold cards, and failing to prioritize moves that open up tableau spaces. Players often overlook the importance of building foundations early, which can limit future moves. Additionally, players may get stuck by moving cards to free cells without a clear plan, leading to blocked tableau piles. Always think several moves ahead and consider how each move affects your overall strategy.