It's one of the most common — and most reasonable — questions players ask: was that game even possible to win? The honest answer is that it depends entirely on which game you're playing, and that's worth understanding properly. There are really two different questions hiding inside "is this winnable?"
The first is: is this particular deal solvable at all? That's a fact about the shuffle — some arrangements of cards simply cannot be cleared by anyone. The second is: did I play well enough to win the deals that were winnable? That's about your decisions. Most frustration comes from confusing the two — blaming the deal for a misplay, or blaming yourself for a deal that never had a solution. Here's the truth, game by game.
The short answer
- Guaranteed winnable: Sudoku (by logic) and Mahjong (by construction) — every deal, always.
- Almost always winnable: FreeCell and Eight Off — losses are nearly always a misplay, not a bad deal.
- Depends on the shuffle: Klondike, Spider, Yukon, Scorpion, Forty Thieves, Pyramid, TriPeaks and Golf — skill matters, but some deals are simply impossible.
- Can require a guess: Minesweeper. And 2048 has no "deal" to be unwinnable — it's pure skill.
Winnability at a glance
| Game | Is every deal winnable? | The honest detail |
|---|---|---|
| Sudoku | Yes — guaranteed | Unique solution, solvable by pure logic at every difficulty. No guessing, ever. |
| Mahjong | Yes — guaranteed | Built by reverse-construction, so a full solution always exists (you can still strand yourself). |
| FreeCell | Almost always | Only #11982 of the classic 32,000 is unsolvable. A loss is nearly always a misplay. |
| Eight Off | Almost always | Eight free cells make it, like FreeCell, winnable on the vast majority of deals. |
| Klondike | No — depends | ~80% of Draw 1 deals are theoretically winnable; Draw 3 is harder. Some deals are impossible. |
| Spider | No — depends | 1-suit is forgiving; 2 and 4 suits much less so. Not pre-screened. |
| Yukon | No — depends | All cards face up and no stock, so it's skill-heavy — but a fair share of deals can't be won. |
| Scorpion | No — depends | Build-down-in-suit with buried cards; many deals are winnable, some are not. |
| Forty Thieves | No — hard | Two decks, strict building, one card at a time — a low win rate even with good play. |
| Pyramid, TriPeaks, Golf | Often not | Stock-limited, so the cards you need may be unreachable. Good play wins the marginal ones. |
| Minesweeper | Not always | The first click is always safe, but some boards reach a point where a guess is unavoidable. |
| 2048 | N/A — pure skill | There's no deal to be "unsolvable"; reaching the target tile is down to technique. |
Winnability, game by game
Klondike
Classic Klondike is part skill, part luck. With perfect play, roughly 80% of Draw 1 deals are theoretically winnable — but real win rates are far lower, because perfect play means seeing every consequence in advance, which almost nobody does. Draw 3 is harder, since only every third stock card is initially reachable. And crucially, some deals are genuinely unwinnable no matter how flawlessly you play. So if you lose a Klondike game, it could be a bad deal — or it could be a line you'd take back with hindsight. Our how to win Klondike and Draw 1 vs Draw 3 guides cover how to give yourself the best chance.
Spider
Spider winnability depends heavily on how many suits you play with and on the choices you make. One-suit deals are forgiving; two suits and especially four suits demand careful sequencing and leave far less room for error. Spider deals are not pre-screened for solvability, so a given hand may or may not be beatable — and even a winnable one can be lost by committing to the wrong order of plays early. The Spider strategy guide and 1, 2 and 4 suits breakdown explain how to keep your options open.
FreeCell
FreeCell is the standout: almost every deal is winnable. Because all cards are dealt face-up, there's no hidden information — the puzzle is fully open from the start. Of the classic 32,000 numbered Microsoft deals, only game #11982 has been proven unsolvable. That single exception is famous precisely because it's so rare. The practical upshot is liberating: a FreeCell loss is nearly always a misplay, not a bad deal. That makes it the ideal game for practising clean, deliberate decision-making — see the FreeCell strategy guide and our FreeCell vs Klondike comparison. You can also play any specific deal by its number on the FreeCell deal numbers page — including the famously unsolvable #11982.
Eight Off
Eight Off is FreeCell's close cousin, and it shares the good news: with eight free cells instead of four, the vast majority of deals are winnable. The extra space gives you more room to manoeuvre, so — as with FreeCell — a loss is much more often a planning slip than an impossible deal. The catch is that those eight cells are easy to clog; spend them carefully and keep one cascade working toward empty.
Yukon and Scorpion
Yukon and Scorpion both deal most cards face up with no stock to fall back on, which makes them skill-heavy but not guaranteed. In Yukon you can move any face-up group regardless of order, which opens up tangled positions; in Scorpion you build strictly down in suit, so a card buried under the wrong sequence can lock a deal up. Deals aren't pre-screened in either game, so a fair share simply can't be won — but good order of play rescues many that look hopeless. See the Yukon rules and Scorpion rules to get the most out of each.
Forty Thieves
Forty Thieves is the serious one. Two full decks, building down in suit, and only one card moved at a time mean many deals are unwinnable and even the winnable ones demand patience — historically it's one of the lowest win-rate solitaires. That difficulty is the appeal: a cleared Forty Thieves board is genuinely earned. The Forty Thieves rules explain how to give yourself the best chance.
Pyramid, TriPeaks and Golf
Pyramid, TriPeaks and Golf are a different breed. They depend heavily on the shuffle and on a limited stock, and many deals simply cannot be cleared no matter how well you play — the cards you need may be buried where you can never reach them in time. That isn't a flaw; it's the nature of these games, and a big part of why a fully cleared board feels so satisfying. Skill still matters a great deal — good order of play wins many marginal hands — but accept that some deals were never going to fall. Our Pyramid, TriPeaks and Golf strategy guides help you win as many of the winnable ones as possible.
Sudoku — guaranteed solvable by logic
Here CardHearth makes a firm promise. Every CardHearth Sudoku has a unique solution and can be solved by pure logic, with no guessing required. Whatever the difficulty, there is always a logical next step to find — you'll never reach a point where the only option is to flip a coin. If you're stuck, the answer is a technique you haven't spotted yet, not bad luck. Learn the methods in our solving techniques guide and see how the tiers differ in the difficulty levels guide.
Mahjong — guaranteed solvable by construction
Our Mahjong games carry a similar guarantee. Every CardHearth Mahjong layout is built by reverse-construction — the tiles are placed by working backward from a fully cleared board, so a complete solution is guaranteed to exist. That said, "solvable" is not the same as "automatic": you can still strand yourself by removing the wrong matching pair and burying a tile you needed. The deal is always beatable; winning it is up to your choices. The Mahjong strategy guide shows how to avoid painting yourself into a corner.
Minesweeper — sometimes you must guess
Minesweeper is the honest outlier among the puzzles. Your first click is always safe, and the great majority of a board is solved by pure logic — but some positions genuinely reduce to a 50/50 guess, where two squares are equally likely to hide a mine and nothing on the board can tell them apart. That's a known property of standard Minesweeper, not a flaw in the deal. The Minesweeper strategy guide shows how to push the guessing as late and as rare as possible.
2048 — no deal to be unwinnable
2048 is different in kind: there's no shuffle or deal that could be "unsolvable". New tiles appear in random spots, but reaching the 2048 tile (and beyond) is a matter of technique — keep your biggest tile cornered and build in one direction. If you fall short, it's the method to refine, not the luck to blame; the 2048 strategy guide has the corner-and-snake approach.
What CardHearth guarantees
We think you deserve to know exactly where you stand, so here's what we promise:
- Sudoku is always solvable by logic. Unique solution, no guessing, ever — at every difficulty.
- Mahjong is always solvable by construction. Each layout is built backward from a cleared board, so a full solution always exists.
- Unlimited undo on every game. Explore a line, and if it dead-ends, take it back and try another — that's the fastest way to learn whether a deal was the problem or your plan was.
- Replay the same deal. Because every deal is numbered, you can run the exact same hand again to test a different approach, or share it with a friend. If you beat it on the second try with smarter play, you've just proven the deal was winnable all along.
Together, undo and replay are the honest way to settle the "was that winnable?" question for games like Klondike and Spider, where deals aren't pre-screened — and for FreeCell, they make it almost certain you can find the solution that was there the whole time.
Frequently asked questions
Are all Klondike games winnable?
No. With perfect play, roughly 80% of Draw 1 deals are theoretically winnable, but a meaningful share are genuinely impossible no matter how you play. Draw 3 is harder still. Real-world win rates are also much lower than the theoretical ceiling, because perfect play is hard — so most losses are a mix of an unlucky deal and an imperfect line.
Is FreeCell always winnable?
Almost. Nearly every FreeCell deal is solvable. Of the classic 32,000 numbered Microsoft deals, only game #11982 has been proven unsolvable. So in practice a FreeCell loss is nearly always a misplay rather than a bad deal — which makes it the best game to practise clean, deliberate play.
Do CardHearth Sudoku puzzles ever need guessing?
No. Every CardHearth Sudoku has exactly one solution and can be solved by pure logic, step by step, with no guessing required. If you are stuck, there is always a logical next move to find — never a coin flip.
Is 4-suit Spider always winnable?
No. Four-suit Spider is the hardest setting, and its deals are not pre-screened for solvability, so some are genuinely impossible. Even winnable ones are easy to lose by committing to the wrong order of plays early. One-suit Spider is far more forgiving, which is why it is the place to learn.
Are Mahjong solitaire boards always solvable?
On CardHearth, yes. Every Mahjong layout is built by reverse-construction — tiles are laid down by working backward from a fully cleared board — so a complete solution is mathematically guaranteed to exist. You can still strand yourself by removing the wrong pair, but the board itself is always beatable.
Why are some Klondike games impossible to win?
Because the shuffle can bury cards in a way no sequence of legal moves can untangle — for example, a card you need early is trapped under cards that can only move after it. That is a property of the deal, not your play. With perfect play roughly 80% of Draw 1 deals are winnable; the rest were lost before the first move.
Keep going
Want a game where every deal is guaranteed beatable? Try Sudoku or Mahjong. Prefer a near-certain solve that rewards careful thought? Play FreeCell. Or test your skill against the luck of the shuffle in Klondike, Spider, Pyramid, TriPeaks and Golf. Whatever you pick, our guides hub has a strategy walkthrough to help you win more of the deals that can be won.