Solitaire Strategy: How to Win More Klondike Games

Most people play Klondike by reflex. A card fits somewhere, so they move it; an Ace appears, so it goes up; a foundation will accept a nine, so the nine goes up too. That style of play wins a respectable share of easy deals and loses almost every difficult one. The good news is that Klondike strategy is not especially subtle once someone spells it out — it comes down to a handful of principles that consistently override the obvious move. This guide covers those principles and, more importantly, the reasoning behind them, so you can apply them to positions no article could anticipate.

Underlying almost everything below is one idea: in Klondike, information is the scarce resource, not moves. At the start of a deal, twenty-one tableau cards are face down and the order of the twenty-four stock cards is unknown. You cannot plan a route to victory because you cannot see the map. What you can do is systematically buy information — flip face-down cards — while keeping as many future options open as possible. Every decision should be judged against those two currencies.

Turn Over Face-Down Cards Above All Else

When two moves compete and one of them exposes a hidden card, take the one that exposes the hidden card. It is the closest thing Klondike has to an unconditional rule. A revealed card might be the very card you need; even when it is not, it becomes a piece of knowledge that makes the next twenty decisions better informed. A move that merely rearranges face-up cards changes nothing about what you know.

When several moves each expose a card, prefer the one working on the longest face-down stack. Column seven starts with six hidden cards and column two with one. Cracking column two gains a single card and a column that will soon be empty-ish but not empty; grinding into column seven is where the deal is actually decided. Deep columns are also where deals go to die: if you leave a six-deep column untouched while you tidy the shallow end of the tableau, you will typically stall out with the cards you need still buried.

Empty Columns Are the Most Valuable Asset in the Game

An empty tableau column is not just a place to put a King. It is a manoeuvring space — a workbench where you can temporarily park a run, break a colour deadlock, or unbury a card that is otherwise stuck under three cards of the wrong colour. Players who win consistently treat producing an empty column as a strategic objective in its own right, comparable to finding an Ace.

Two corollaries follow, and both are counterintuitive.

First, do not fill an empty column with the first King you see. Once a King sits there, the space is gone, potentially for the rest of the game, because Kings are almost impossible to move again. Before committing, ask what that King brings with it. A King with a Queen of the opposite colour available — either in the tableau or reachable from the waste — immediately begins a new productive run. A King with nothing behind it just plugs your best asset with dead weight.

Second, prefer a King that is currently blocking something. If a King sits face up on top of face-down cards in another column, moving it to the empty space costs you nothing in flexibility and gains you a fresh reveal. If instead you take a King off the waste, you have spent your column and revealed nothing. Same King, very different move.

The one deal-breaking configuration to watch for early: if all four Kings sit high in the tableau with hidden cards beneath them, and no column can be emptied, you may be in a deal that no amount of skill will win. Recognising that early saves time.

When Not to Play a Card to the Foundation

This is the single biggest gap between casual and strong play. Sending a card to a foundation feels like progress — it is literally how you win — but a card on a foundation is a card that can no longer help you in the tableau. Every card you send up is one less landing spot for the cards still stuck in the columns.

The reliable rules of thumb:

• Aces and twos: always send them up immediately. An Ace can never receive anything in the tableau and can never be built on. A two can only ever hold an Ace, which you would never place there anyway. They are pure dead weight in a column and pure progress on a foundation. There is no realistic downside.
• Threes, fours and fives: usually up, but glance first. These are low enough to be rarely needed as landing spots, but if the matching lower card is stranded in the waste or under a pile, a three left in place can still do work.
• Sixes through Kings: hold them back unless there is a reason. A black eight in the tableau is a home for a red seven. Send it to the foundation and that red seven may sit uselessly on top of a face-down card for the rest of the game.

A practical balance test: try to keep your foundations roughly level. If hearts and spades are up to seven while clubs and diamonds are still on three, you have stripped away most of your high red or black landing spots and the tableau will seize up. Aim to advance foundations in step, and let a suit run ahead only when you can see that the cards it would have caught are already safely placed.

The exception that overrides all of this: play a card up whenever doing so unblocks something. If sending the nine of clubs home is what lets you flip a face-down card or empty a column, send it. Progress in the tableau outranks abstract caution.

Colour Alternation Planning

Because tableau builds alternate colour, you should think of the tableau as two interleaved half-decks. When you need a home for a red seven, you need a black eight — and there are only two of those in the deck. If both black eights are buried or already on foundations, that red seven is homeless no matter how many other cards you can see.

Strong players run a quick mental audit before big moves: for each card I am about to strand, does a receiver of the opposite colour still exist and is it reachable? This is also why it pays to avoid stacking two same-rank cards of the same colour in ways that consume both receivers. If you have a choice of which black eight to move a red seven onto, choose the one that keeps the other black eight free and, ideally, the one whose column has face-down cards you want to attack.

The same logic applies to Kings and Queens. Two red Kings and two black Kings; two of each colour Queen. If you plant both black Kings in empty columns and both red Queens are already on foundations, those Kings are permanent dead ends.

Managing the Stock in Draw-3

Draw-3 is where stock management becomes a real skill. Because only every third card of the stock is playable on a given pass, the deck is effectively partitioned into three interleaved sets — and the crucial fact is that playing a card changes the partition. Remove one card from the waste and every subsequent triplet shifts by one, exposing a completely different set of cards on the next pass.

Practical consequences:

1. Make a full pass early with no plays if you can afford it, simply to see which cards exist and roughly where. In unlimited-redeal play this reconnaissance is nearly free.
2. Remember the order. You do not need perfect recall — just noting "the ace of spades is near the bottom" or "there are two useful red sevens back to back" is enough to plan a pass around.
3. Use a deliberately useless play to shift the offset. If a card you badly need is permanently landing in the middle of a triplet, taking any playable card earlier in the pass re-aligns the deck and brings it to the top next time.
4. Empty the waste when you can. A waste pile you have picked clean gives you a cleaner offset and fewer buried cards to track.

Where scoring penalises redeals, or the variant limits you to three passes, this all becomes sharper: you should be planning which cards you intend to extract on each pass rather than reacting to whatever surfaces.

Using Undo Honestly

Undo is a legitimate part of digital solitaire and pretending otherwise mostly means playing worse. The useful distinction is between using undo to explore and using it to cheat information.

Exploring is fine and genuinely improves your play: you commit to a line, discover three moves later that it dead-ends, and back out. That is planning under uncertainty, and it is how you learn which moves are traps. Undoing purely to re-flip a face-down card and see something different is a different activity — some implementations reshuffle, some do not — and it turns the game into a search rather than a decision problem. Neither is wrong, but if you are trying to measure your skill, keep undo for exploring lines you could have reasoned about, and accept the flips you get.

What Win Rate Should You Expect?

Klondike deals are not all winnable, so a hundred percent is not the target. Analyses of "thoughtful" Klondike — where every card is visible from the start and the player has perfect information and perfect play — generally put the proportion of solvable draw-1 deals somewhere around eighty percent, often quoted as roughly 79 to 82 percent. Draw-3 with limited passes is meaningfully lower. These figures are approximate and depend on the exact rule set used, so treat them as a ceiling rather than a benchmark.

Real play falls well short of that ceiling, because you cannot see the face-down cards. A casual player in draw-1 typically wins something like a third of games. A player consistently applying the principles above — prioritising deep face-down stacks, protecting empty columns, holding back mid and high cards from the foundations — can commonly reach roughly half in draw-1, with draw-3 running considerably lower. If your win rate sits below that, the fastest single improvement is almost always the foundation discipline: stop sending sixes through Kings up on sight.

A Short Checklist

1. Aces and twos go up immediately, without thinking.
2. Prefer any move that turns over a face-down card.
3. Attack the deepest face-down stacks first.
4. Work toward an empty column, and do not spend it cheaply.
5. Fill empty columns with Kings that unblock something, and ideally that have a Queen ready.
6. Keep the four foundations roughly level.
7. Before stranding a card, confirm an opposite-colour receiver still exists.
8. In draw-3, track the stock and shift the offset deliberately.
9. Accept that some deals are simply lost, and deal again.

If any of the terminology here is unfamiliar, our complete Klondike rules guide defines every term and covers the variations you will meet.