By Online Solitaire Free Editorial Team · Editorial Standards
The endgame in solitaire is the phase of play that begins when the win condition becomes theoretically achievable — when enough cards have been revealed, sequenced, or placed that a complete path to victory exists somewhere in the current board state — and ends when that path is either successfully executed or irreversibly blocked. Most solitaire players have a well-developed opening strategy (uncovering moves first, stock discipline, suit purity) and a developing mid-game strategy (sequence management, empty column planning, completion prioritisation). Far fewer have a deliberate endgame strategy, because the endgame appears to take care of itself: the hard work is done, the cards are where they need to be, the win condition is in sight. This appearance is deceptive.
The endgame in solitaire is the phase of play that begins when the win condition becomes theoretically achievable — when enough cards have been revealed, sequenced, or placed that a complete path to victory exists somewhere in the current board state — and ends when that path is either successfully executed or irreversibly blocked. Most solitaire players have a well-developed opening strategy (uncovering moves first, stock discipline, suit purity) and a developing mid-game strategy (sequence management, empty column planning, completion prioritisation). Far fewer have a deliberate endgame strategy, because the endgame appears to take care of itself: the hard work is done, the cards are where they need to be, the win condition is in sight. This appearance is deceptive.
The endgame is where a significant proportion of otherwise winnable games are lost, for a specific reason: the constraints that make the endgame winnable are different from the constraints that governed mid-game play, and strategies that were correct in the mid-game become incorrect or suboptimal in the endgame. The most common endgame failures are: foundation sequencing errors (placing cards on foundations in an order that blocks subsequent placements); empty column depletion (using the last empty columns for purposes that eliminate the positional flexibility needed to complete the final sequences); and rush errors (playing quickly because the win seems obvious and missing a specific ordering constraint that makes the obvious sequence fail). Each of these failures is preventable through endgame-specific habits that require almost no additional effort but produce measurably better results.
This article covers the specific endgame structure for each major variant, the ordering principles that govern correct endgame execution, the mistakes that cause endgame collapses, and the habits that make endgame wins reliable once the mid-game has produced a winnable position.
In every solitaire variant, the endgame phase has a distinct character defined by what has changed from the mid-game. In Klondike, the endgame begins when all face-down cards have been revealed and all four Aces are either on the foundation or accessible on the tableau. The endgame board is fully transparent — no hidden information remains — which transforms the game from a probabilistic planning exercise into a deterministic sequencing problem. The correct endgame sequence in Klondike can, in principle, be calculated from the current position; the only remaining question is whether the player calculates it correctly or makes an ordering error under the pressure of apparent near-victory.
In Spider, the endgame begins when enough suit sequences have been completed and removed (typically five or six of the eight total) that the remaining tableau is sparse enough for the completion of the final two or three sequences to be planned in full. Spider's endgame is structurally more complex than Klondike's because the final sequences must be built entirely within the tableau before removal, requiring more intermediate moves and more empty column management than Klondike's foundation placements. In FreeCell, the endgame begins when all 52 cards are visible and the foundation build order can be calculated directly; FreeCell endgames that reach this state are almost always won because the complete information and four free cells provide enough flexibility for any ordering problem to be resolved. In Forty Thieves and other stock-intensive games, the endgame begins when the stock is exhausted and all remaining moves must come from the current tableau state.
Foundation sequencing — placing cards on the four suit foundations in the correct order to avoid creating blocks — is the most technically demanding endgame skill and the one that causes the most endgame collapses in Klondike and FreeCell. The sequencing problem arises because the four foundations must advance in coordination: if one suit is placed far ahead of the others, the cards of the advanced suit are removed from the tableau before they have served their purpose as build bases for the adjacent-rank cards of other suits, leaving those cards with no tableau placement and no foundation home.
The coordination principle: at any endgame position, before placing any card on a foundation, check whether the card being placed is needed by any other suit's cards currently on the tableau as a build base. Specifically: a red 7 on the hearts foundation is also a potential build base for a black 6. If there is a black 6 currently on the tableau that has no other build base available, placing the red 7 on the hearts foundation removes the only available reception point for that black 6, potentially stranding it without a destination until the black 6's own suit foundation is advanced enough to receive it directly. In the mid-game, this stranding is recoverable because the stock or further tableau moves can provide alternative destinations. In the endgame, with no stock remaining and all tableau positions revealed, a stranded card may have no recovery path.
The practical endgame sequencing habit: before placing any card on a foundation, scan the tableau for any card of adjacent rank and opposite colour (the card that would build onto the candidate in a tableau context). If that adjacent card exists and has limited placement options, delay the foundation placement until the adjacent card has been placed or its placement options are confirmed sufficient. This one-second scan adds minimal time to each endgame move and prevents the specific class of endgame collapses caused by premature single-suit foundation racing.
Klondike endgame: the two-rank rule. In Klondike endgames where all cards are visible, the two-rank rule provides a reliable sequencing guideline: keep all four suit foundations within two ranks of each other throughout the endgame. When all four foundations are at rank 7 or below (Ace through 7), placing an 8 on any suit is safe. When one suit is at rank 9 and the others are at rank 6, the 10 of the advanced suit should be held back until the other suits approach rank 8, because the advanced suit's high cards may be needed as build bases for the lagging suits' 9s and 10s. The two-rank guideline is not a rigid rule — specific deals may require temporary deviations — but it provides the correct default that prevents the most common Klondike endgame collapses without requiring complex position-by-position calculation at every move.
FreeCell endgame: free cell reservation. FreeCell endgames that become stuck almost always share the same cause: all four free cells are occupied simultaneously at a moment when a specific sequence of foundation placements requires a temporary staging position that no free cell can provide. The prevention is the same as in mid-game FreeCell play: treat free cells as reserve staging, not convenience staging. In the endgame specifically, as foundations advance and the tableau thins, the temptation to use free cells liberally increases because the board looks nearly finished. Resisting this temptation — maintaining at least one free cell available throughout the endgame — is the most reliable single endgame habit in FreeCell.
Spider endgame: completion priority and column preservation. Spider's endgame involves completing and removing the final two or three suit sequences from a thinning tableau. The endgame challenge is that as sequences are removed, the columns they occupied become empty, and the remaining cards concentrate into fewer columns. The correct endgame priority: identify which of the remaining partial sequences is closest to completion and advance it exclusively, using any available empty columns as staging for the intermediate moves required. Do not attempt to advance two partial sequences simultaneously in the endgame; the empty columns available are insufficient to stage two simultaneous completion sequences, and the attempt typically produces a position where both sequences are partially advanced but neither can be completed without using staging that the other sequence also needs.
Forty Thieves endgame: waste pile order management. Forty Thieves endgame begins when the stock is exhausted and the waste pile contains the remaining unplaced cards in a fixed order. The endgame challenge is that the waste pile cards are accessible only from the top, and the needed cards may be buried under later-drawn cards whose placement is currently impossible. Expert Forty Thieves endgame play involves mapping the waste pile's order — identifying which cards are near the top, which are buried, and whether the buried cards can be placed before their blocking cards — and constructing the foundation placement sequence that allows the waste pile to be cleared in order. Forty Thieves endgames that reach a clean tableau with a small waste pile are frequently winnable even when they appear stuck, because the waste pile's fixed order sometimes aligns with the tableau's reception capacity in ways that require careful sequencing to discover.
Klondike Turn 3 endgame: the final pass discipline. Turn 3 Klondike endgames have an additional constraint: the stock may have limited passes remaining, and the final pass is the last opportunity to access any unplaced stock card. The endgame discipline in Turn 3 is to play every tableau move available before exhausting the final stock pass, because tableau moves that occur after the stock is exhausted have no stock draw to fall back on — every card's placement must come from the current visible tableau. This is the extreme expression of the stock discipline principle: in the Turn 3 endgame, the stock is not just a secondary resource but a finite countdown, and every move that resolves a tableau placement before the stock is exhausted is a move that preserves options rather than consuming them.
Playing faster as the win comes into sight. The most universal endgame mistake across all variants is increased move speed as the win condition approaches. The cognitive experience of near-victory produces a different attentional state than mid-game problem-solving: attention narrows toward the goal rather than maintaining the wide scan that identifies constraints. The moves that cause endgame collapses — a premature foundation placement that strands a tableau card, a final empty column filled when it was needed for one more staging sequence — are almost always moves that were made quickly without the pre-move pause described in the psychology guide. Maintaining the same deliberate move pace in the endgame as in the mid-game is the single most effective endgame habit, because it applies the full attentional check to the specific moves where the check is most valuable. See our psychology guide for the complete framework on attention management across a game session.
Assuming the win is automatic once the position looks winning. A board position that looks winning — all cards revealed, foundations mostly advanced, sequences accessible — is not automatically winning. A specific ordering of the remaining moves is required, and that ordering may have constraints that are not immediately visible. The most common form of this mistake is in Klondike endgames where all face-down cards have been uncovered: the player assumes the remaining moves are routine and plays without checking foundation sequencing, missing a two-rank imbalance that will strand a specific card in two or three moves. Explicitly confirming that the position is actually winning — not just looking winning — requires one additional scan of the current tableau before committing to the final sequence. That scan takes ten seconds and prevents the most frustrating category of solitaire loss: the game that was won until the last five moves.
Using the last empty column without a specific completion purpose. In any variant where empty columns exist in the endgame, the last empty column is almost always the most valuable positional resource on the board. Using it for a routine build or a convenience staging move — when it might be the only available pivot point for the final completion sequence — is the endgame equivalent of the mid-game mistake of filling empty columns immediately. The expert habit described in our expert strategies guide applies with maximum force in the endgame: identify the specific completion purpose the last empty column will serve before filling it with anything, including an apparently useful sequence.
FreeCell is the best endgame training game because its complete information makes the foundation sequencing problem fully visible and calculable — every endgame constraint is detectable through board examination, with no hidden card uncertainty. Players who develop the two-scan endgame habit (adjacent-rank build base check before foundation placement; full board confirmation that position is actually winning) in FreeCell develop the habit structure that applies immediately to Klondike and Spider endgames. Klondike Turn 1's unlimited recycling makes it the most forgiving environment for developing endgame sequencing habits — endgame ordering errors can be undone and replayed, producing faster iteration on the specific ordering constraint that caused the collapse. Forty Thieves is the most demanding endgame environment in the mainstream catalogue: its single-pass stock, same-suit build constraints, and waste pile order management combine to produce endgames that require explicit sequencing calculation rather than approximate coordination. Forty Thieves endgame proficiency is the strongest evidence of complete endgame strategy development, because it requires all the skills of the simpler variants plus the waste pile order mapping that is unique to stock-exhaustion endgames.
Three habits cover the complete endgame strategy: the foundation sequencing scan (before each foundation placement, check whether the card being placed is needed as a build base by any adjacent-rank card currently on the tableau); the two-rank coordination guideline (keep all four suit foundations within two ranks of each other throughout the endgame, delaying advanced-suit placements when lagging suits need their high-rank cards as build bases); and the endgame pace discipline (maintain the same deliberate pre-move pause as the mid-game, specifically resisting the attentional narrowing that near-victory produces). These three habits together prevent the three specific endgame failure types — foundation sequence blocking, rush errors, and empty column depletion — and make endgame execution reliable once the mid-game has produced a winnable position.
Forty Thieves has the most strategically complex endgame of any mainstream variant because its single-pass stock means that by the time the endgame phase begins — stock exhausted, waste pile fixed — the tableau must provide all remaining placement options from a static set of accessible cards. There is no stock to draw from if a placement gap appears, and the waste pile order is fixed rather than reshuffleable. Every Forty Thieves win requires a correctly executed endgame; many Forty Thieves games that reach the endgame phase with a winnable position are lost to sequencing errors that a foundation coordination scan would have prevented. FreeCell, by contrast, has the most forgiving endgame — its free cells and complete information make most endgame positions that reach the late game straightforwardly winnable with patient sequencing.
No. Not every position that feels like an endgame is actually a winning endgame — some positions that appear to be close to a win have specific sequencing constraints that make the win impossible from that exact arrangement, even though the game was theoretically winnable from the opening deal through a different move sequence. This is the endgame version of the broader principle that some games are lost at the mid-game planning stage rather than the endgame execution stage. The diagnostic process described in our expert strategies guide applies directly to apparent endgame stuck positions: identify the specific blockage type, apply the pattern-specific response, and determine whether the position is genuinely unwinnable or merely requires a non-obvious move sequence to resolve. Expert players convert a meaningful proportion of apparent endgame failures into wins through this diagnostic process; developing players who resign apparent endgame stalemates early give up winnable positions that patient expert analysis would have resolved.