Klondike Solitaire: what your opening deal is really telling you
Every game of Klondike starts with the same small moment. Seven columns land on the table, seven cards turn face up, and you decide in a glance whether this one looks "good" or "bad". We wanted to know whether that first impression means anything, so we dealt 10,000 games of Klondike Turn 1 under this site's rules and looked closely at what each opening layout contained.
The short answer: the opening deal tells you less than you think. In 88.0% of our deals at least one Ace was hidden face down in the tableau, and on average 1.6 of the 4 Aces started buried. In 8.6% of deals none of the seven face-up cards could move anywhere. Yet a simple computer player that never undoes a move still won 42.3% of all 10,000 games — and its win rate barely moved whether the Aces started hidden or not.
So if you have ever reshuffled a deal because it "looked hopeless", this page is for you. Below are the numbers, the rules as they work here, and what to look for in the first thirty seconds.
The opening layout, by the numbers
| What we looked at | Result | Deals | How we got it |
|---|---|---|---|
| Deals with at least one Ace face down | 88.0% (95% range 87.4–88.7%) | 10,000 | Inspecting the deal |
| Aces face down, on average | 1.6 of 4 | 10,000 | Inspecting the deal |
| Deals where no face-up card can move at the start | 8.6% (8.1–9.2%) | 10,000 | Inspecting the deal |
| Computer player's wins, 2 or more Aces hidden | 41.6% | 5,333 | Rule-based player, no peeking, no Undo |
| Computer player's wins, no Aces hidden | 44.5% | 1,196 | Same player |
| Computer player's wins, all deals | 42.3% (41.3–43.3%) | 10,000 | Same player |
| Deals proven winnable by a solver that sees every card | at least 76.6% | 1,000 | Search with a 150,000-position limit |
The gap between a "bad" opening (two or more buried Aces) and a "perfect" one (no buried Aces) was just under three percentage points: 44.5% against 41.6%. That is real, but it is small next to everything else that happens in a game.
How Klondike works on klondike.nz
The table
One pack of 52 cards. Seven columns are dealt from left to right holding 1, 2, 3, 4, 5, 6 and 7 cards, with only the top card of each column face up. That is 28 cards on the table, 21 of them hidden. The remaining 24 cards form the stock.
What you are aiming for
Move all 52 cards to the four foundations, one per suit, building from Ace up to King.
Legal moves
- Build down the columns in alternating colours: a black 9 on a red 10, a red 4 on a black 5.
- Move any face-up run as a block, as long as its top card fits where it lands.
- When you uncover a face-down card, it turns over.
- Only a King (or a run headed by a King) may go into an empty column.
- Turn one card at a time from the stock. When the stock runs out, you can turn the waste back over and go through again — as often as you like.
- You may bring one card back down from a foundation if you need it to catch a lower card.
If you prefer the harder draw, the same game with three cards at a time is on our Turn 3 page.
Reading the deal before you touch a card
Count the hidden cards, not the Aces
With 1.6 Aces buried in the average deal, looking for Aces is almost always disappointing. A better question is: where are the most hidden cards? The right-hand columns hide six, five and four cards. Freeing those gives you more information than anything else on the table, and it is exactly what our computer player prioritised — after sending a card to a foundation, its first choice was always the move that uncovers a card in the column with the most face-down cards.
A frozen start is not a lost game
Roughly one deal in twelve (8.6%) opens with nothing to do on the table. That feels like a dud. It isn't. With unlimited passes through the stock at Turn 1, every card in the stock is available to you eventually, so a frozen tableau just means your first moves come from the stock instead.
Buried Aces cost less than they look
Our player won 41.6% of deals with two or more Aces hidden and 44.5% of deals with none hidden. A buried Ace is a nuisance, but most of the work in Klondike is sorting the middle cards — the 5s to 9s — into columns that let you dig. The Aces come out on the way.
Look for your Kings
Because only Kings can fill an empty column here, an empty column with no King in sight is wasted space. Before you clear a column, know which King is going in and where it is coming from.
Keep the stock flowing
When our player won at Turn 1, it went through the stock 2.3 times on average. There is no penalty for a second or third pass here, so don't force a poor move just to avoid turning the waste over.
The ceiling: how many deals can be won at all?
A computer player that only sees what you see is one end of the scale. At the other end we ran a solver that is allowed to see every hidden card. Across 1,000 deals it proved 76.6% winnable and 2.1% impossible, and ran out of search budget on the remaining 21.3%. So somewhere between 76.6% and 97.9% of Turn 1 deals can be won with perfect knowledge.
The published figure from the most thorough study to date, by Blake and Gent (2026) in the Journal of Artificial Intelligence Research, is 90.48%. That sits comfortably inside our range. Between the 42.3% a no-Undo player manages and the roughly 90% a perfect player could manage lies a lot of room for skill, patience and the Undo button.
For a sense of real players: solitairex.io reported 913,106 Klondike games in the month before 16 July 2025, with 44.8% of them won.
Playing on klondike.nz
The game is free and runs in any modern browser on a computer, tablet or phone. There is nothing to download or install, and no account, sign-up or email address is needed. The cards are large and clear, with a zoom control if you want them bigger. You get Hint, unlimited Undo, Restart, Auto Play to finish a solved game, and you can drag cards or simply tap them. The game keeps track of wins, time and moves for the current game. Sound can be switched off, and there is a fullscreen mode and a choice of table colours — handy for a quiet game on the ferry, at the bach, or over a cuppa at smoko.
Where these numbers come from
The klondike.nz team ran this study in September 2026 to answer one question for players on this site: how much does the opening layout matter?
- Rules from our own game. We read the rules straight from the code that runs this page: seven columns, one card per turn, unlimited passes through the stock, Kings only in empty columns, alternating-colour building, and one card allowed back off a foundation.
- A program copy of the game. We rebuilt Klondike as a program and checked over 2,000 games that it only ever makes legal moves.
- Numbered deals. Every deal is shuffled from a numbered seed (1 to 10,000), so the exact same deals can be dealt again and the test repeated.
- Looking at the deal. For each deal we counted the hidden Aces and checked whether any face-up card could move.
- Two players. A rule-based player that sees only face-up cards and never undoes a move played all 10,000 deals. A solver that sees every card played 1,000 deals, with a limit of 150,000 positions each.
- Comparing groups. We split the player's results by how many Aces started hidden.
- Checking against published work. Our solver range (76.6–97.9%) contains the published 90.48%.
- Honest limits. The rule-based player is a floor, not a typical human; the solver is a ceiling no human can reach, because it sees face-down cards.
Under the bonnet
The simulations are written in Python 3.11 using only the standard library. The Klondike code lives in a module called klondike.py; a shared runner splits the deals across 19 worker processes with Python's multiprocessing, so the 10,000-game run took about 14 seconds and the 1,000-deal solver run about 17 minutes. The player follows a fixed priority list (foundation moves first, then moves that uncover hidden cards, then Kings to empty columns, then the stock). The solver is a depth-first search that tries every legal move, remembers positions it has already seen, and only proves a deal impossible when it has run out of moves to try. The 95% ranges are Wilson score intervals, a standard way to put error bars on a percentage. We used an AI assistant, Claude by Anthropic, to help write and review the code and to draft this page. None of the figures came from the AI: every number here comes from running the code, and we checked the results against the published study before publishing.
Questions players ask
Should I start a new game if no Aces are showing?
No. In 88.0% of deals at least one Ace is hidden, so it is the normal state of affairs, and our player won 41.6% of the deals with two or more buried Aces.
What does it mean if I can't make any move at the start?
It happens in about 8.6% of deals. Start turning the stock; at Turn 1 with unlimited passes, you will see every stock card.
How many times can I go through the stock?
As many as you like. Winning games in our test used 2.3 passes on average.
Can every deal be won?
No. Our solver proved 2.1% of deals impossible even with every card visible. Most deals can be won with perfect play, though far fewer are won in practice.
Where can I learn more about the people behind this site?
Our about page explains who we are. For something with no hidden cards at all, try FreeCell, or browse every game on the games page.
By Admin, klondike.nz. Study date: 18 September 2026.