Solitaire variants — a guide for Klondike players
Klondike is the variant most people mean when they say “solitaire”, but it is one of dozens. This article covers five of the most popular variants, what makes each distinctive, and where to play them. Our site focuses on Klondike — what follows is a guide for the curious. If you finish a few hands of Klondike and want to see what else exists, this is the map.
Klondike
Klondike is the game on our site, so this section is brief. Seven tableau columns of ascending size, four foundations, a stock pile of twenty-four cards, and a waste pile. You build down on the tableau in alternating colors — a black eight on a red nine — and build up on the foundations by suit from ace to king. The game ends when all 52 cards reach the foundations or no legal move remains.
The defining feature of Klondike is incomplete information. Twenty-one of the twenty-eight tableau cards start face-down. You uncover them as you play, which is what makes the game both accessible and luck-dependent. Around 80 to 90 percent of deals are theoretically winnable; the rest cannot be won regardless of skill.
Most solitaire sites default to draw-three, where you flip three cards at a time from the stock. The draw-one variant — one card at a time — is the original and the easier of the two. Our site offers draw-one. Play Klondike, or read the rules article for the full setup.
FreeCell
FreeCell deals all 52 cards face-up into eight tableau columns at the start. No stock, no waste, no hidden cards. Four free cells at the top of the board can hold one card each, temporarily, while you maneuver. The goal is the same as Klondike: build four foundations up by suit from ace to king.
What makes FreeCell distinctive is the absence of hidden information. Because every card is visible from move one, the deal’s solvability is decidable in principle. With strong play, nearly every deal can be won. Computer solvers have confirmed that of the 32,000 numbered deals in the classic Windows FreeCell, only one — Deal #11982 — is provably unwinnable. Among random deals more broadly, the rate of unsolvable hands is about one in 84,000.
The four free cells are the heart of the game. Each cell holds a single card you have temporarily set aside while moving others. Used well, they let you reorganize the board into any sequence you need. Used carelessly, they fill up and you stall. Most experienced players win 80 percent or more of FreeCell hands; casual players win around 30 percent. The gap between those two numbers is pure skill.
FreeCell rewards planning and punishes haste in a way Klondike does not. There is no shuffle to blame when you lose — if the deal could be won, you can replay it and find the path.
If you like tactical depth and you do not mind losing as a learning experience, FreeCell is the variant most likely to deepen your card-game thinking. Play FreeCell at cardgames.io.
Spider
Spider uses two decks — 104 cards in total — dealt into ten tableau columns. There is no separate foundation row. Instead, you complete sequences of ace through king of a single suit directly on the tableau, and each completed sequence is removed from the board. The game ends when all eight complete sequences have been built and removed.
You build down on the tableau ignoring color — a six can go on any seven of any suit — but only complete same-suit sequences ace-through-king clear from the board. That distinction matters: it is easy to build down by mixing suits, hard to keep an entire suit sequence intact long enough to remove it.
Spider comes in three difficulty levels based on how many suits are in play. One-suit Spider uses 104 cards all of the same suit and is the most accessible variant — somewhere around 80 to 90 percent of deals are winnable with strong play. Two-suit Spider uses two suits and drops winnability to roughly 20 to 35 percent. Four-suit Spider — the standard “expert” version — uses all four suits, and win rates often sit in the single digits even for skilled players.
The defining feature of Spider is patience. Games run longer than Klondike — fifteen to thirty minutes is typical at four-suit — and a single mistake can lock you out of a win that was reachable five moves earlier. The skill curve is steep but rewarding.
Spider also shares Klondike’s hidden-information dimension. The cards in the lower rows of each tableau column start face-down. You uncover them by clearing the cards above, which gives Spider some of Klondike’s luck-of-the-shuffle character, but with substantially more room for strategy. Play Spider at cardgames.io.
Pyramid
Pyramid deals 28 cards into a triangular arrangement of seven rows, with one card at the top and seven at the bottom. A standard stock and waste hold the remaining 24 cards. The goal is to remove every card from the pyramid and the stock by pairing cards that sum to 13.
The pairing mechanic is the entire game. Twos pair with elevens (jacks), threes with tens, fours with nines, fives with eights, and sixes with sevens. Aces equal one, jacks equal eleven, queens equal twelve, kings equal thirteen — so kings remove alone, with no partner needed. A four of hearts paired with a nine of spades clears both cards. A king of clubs clears by itself. There are no suits to track, no sequences to build, no foundations.
Difficulty comes from blocking. Only fully exposed cards — those not covered by the row below — can be paired and removed. If the eight you need is buried under three other cards, you have to clear those three before you can reach it. Many deals trap key pairs in positions no path can unblock.
Pyramid is the hardest of the popular variants by winnability. Around 5 percent of deals are winnable under typical rules with three stock passes, dropping closer to 1 or 2 percent under strict single-pass rules. Games are short — often three to five minutes — which makes the low win rate feel less punishing. A loss costs only a few minutes before you can try again.
If quick games and a memory-and-arithmetic feel appeal to you, Pyramid is worth a few hands. Play Pyramid at cardgames.io.
Yukon
Yukon is Klondike’s closest cousin. Seven tableau columns are dealt similarly to Klondike — column one gets one card, column two gets two cards, and so on — but the deck does not stop there. The remaining cards are dealt face-up across the rightmost columns, so the entire deck is on the tableau from the first move. There is no stock and no waste.
The other big difference is movement freedom. You can move any group of consecutive face-up cards from one column to another, regardless of whether they form a valid sequence. A black six with a red two underneath it can move as a pair onto a red seven, even though the two does not “belong” with the six. The receiving card just needs to be the right rank and color for the moving group’s bottom card.
This freedom changes the game’s character. Yukon has no shuffle luck — every card is visible — but the moves available at any moment are far greater than Klondike’s. You can dig through buried cards more aggressively, but you can also lock yourself into dead ends faster.
Winnability with optimal play is estimated to be similar to Klondike, somewhere in the 70 to 80 percent range, but practical win rates tend lower because Yukon rewards careful planning more than Klondike does. If you like Klondike but want a game where every loss is your own fault, Yukon is the closest match. Play Yukon at cardgames.io.
Why Klondike
Our site offers only Klondike. That is a choice, not a limitation we plan to fix.
Three reasons. First, Klondike is the variant most people mean when they say “solitaire”. For the audience we built this site for — people who want bigger cards and a calmer playing surface, often without a strong card-games background — landing on a familiar game is more important than offering variety. Variety adds a learning curve the senior demographic specifically did not ask for.
Second, Klondike sessions are short. Five to twenty minutes is typical, which fits the relaxed-pause use case better than Spider’s thirty-minute four-suit grinds or FreeCell’s heavier thinking load. We optimized for “a hand between cups of tea”, not for the strategic enthusiast.
Third, doing one thing well beats doing five things adequately. The big-card accessibility design, the WCAG 2.2 AAA contrast work, the keyboard navigation, the win celebration — each took deliberate effort, and spreading that effort across five variants would have diluted all of them.
We may add other variants over time. For now, we are starting by doing Klondike well. If you want to try the others, the links in this article go to a site that has covered them for years. Play Klondike is what we offer, and the FAQ covers the most common questions about our specific setup.