A good Klondike win rate is about one win in three for honest one-card play. Above 50% is strong. The often-quoted 80%-plus figures are not normal player records; they come from theoretical or open-information versions of the game.
That distinction matters because Klondike is full of small rule traps. Turn 1 is easier than Turn 3. Unlimited redeals help. Undo changes the result because a mistake no longer has to stay made. Hints may stop obvious mistakes. Restarting losing deals without counting them destroys the number completely. Two players can both say they have a 40% solitaire win rate and mean very different things.
There is also no official Klondike win-rate record in the way there is for a timed sport. No recognised body publishes a deal-by-deal leaderboard for classic Klondike across fixed rules, verified starts, and audited finishes. The best documented high figure belongs to research on winnability, not a human record-holder. It is useful, but it answers a different question: how many deals can be won with enough information and correct play?
So what counts as a good win rate in Klondike?
For Klondike Solitaire (Turn 1), a good win rate starts around one win in three if you count every deal you begin. That is the plain answer. It assumes normal play, not cherry-picked deals, not restarts hidden from the total, and not a solver making decisions with perfect information.
The phrase average solitaire win rate is messier than it looks. Most public figures come from apps, old rule books, forum threads, or studies using different rules. Some count all abandoned games as losses. Some do not. Some allow unlimited undo. Some show hints. Some games auto-move cards to the foundations. A small rule change can move the number more than a player realises.
For a practical good win percentage solitaire benchmark, use bands rather than pretending there is a single world standard. Under 20% usually means you are missing forced moves, freeing the wrong columns, or wasting empty spaces. Around 25% to 35% is decent for casual Turn 1 play. Around 40% to 50% is good if every deal is counted. Above 50% is strong enough that your rules and record-keeping need to be clear before anyone compares it fairly.
Turn 3 is stricter. It restricts access to the stock, so the same player will normally win less often than in Turn 1. Comparing a Turn 1 result with Klondike Solitaire (Turn 3) is like comparing Blackjack and Single Deck Blackjack without saying which rules are in force. The name looks close. The mathematics is not.
Who actually holds the Klondike win-rate record, and what was the number?
No official Klondike win-rate record-holder exists. Not a credible one. There is no public, recognised authority that certifies a named player’s long-run Klondike win percentage under fixed rules, with every starting deal logged and every loss counted.
The best documented high number is therefore not a human record. It is a research figure. Charlie Blake and Ian P. Gent reported results for Klondike winnability in their 2019 work on patience games using the Solvitaire solver. The commonly cited result is about 82% winnability for one-card Klondike under a thoughtful-play model.
The crucial detail is thoughtful. In thoughtful Klondike, the hidden cards are effectively known to the analysis. A human player at the table does not have that information. You can make the right-looking move and still be punished by an unseen card. A solver working with the whole deal is answering a cleaner question: can this deal be won if the player has the necessary information and makes the right choices?
That makes the 82% figure valuable, but it is not a solitaire win rate record. It is closer to an upper-limit study for a defined version of the game. A real player claiming 82% over ordinary hidden-information Klondike would need public deal logs, fixed rules, unedited losses, and enough games for the number to mean something. Without those, the honest answer is boring and important: no verified record.
Why do different sources give different win rates?
Different sources give different win rates because they are often measuring different games while using the same word: Klondike. A printed rule book, a phone app, a solver, and a player’s notebook can all be internally honest and still produce numbers that should not be compared.
| Source type | What it often measures | Why the number changes |
|---|---|---|
| Casual player records | Finished games the player remembers to count | Abandoned deals and restarts are often missing |
| Browser or app statistics | Games under that software’s rule set | Undo, hints, auto-move, and redeals vary |
| Turn 1 Klondike | One-card stock access | More stock cards are reachable sooner |
| Klondike Solitaire (Turn 3) | Three-card stock access | Card order blocks more moves |
| Solver studies | Winnability under defined assumptions | The solver may know hidden cards or search many lines |
Searches for average solitaire win rate often land on pages that mix these categories. That is where the confusion starts. A theoretical winnability figure can sit beside a casual app statistic as if both describe the same achievement. They do not.
Even the word “win” needs care. Does the game count a win when all cards reach the foundations manually, or when the software can finish the remaining forced moves? Does an undo erase the mistake or merely allow better play after seeing information? Does a new deal after 20 seconds count as a loss? These choices are not decoration. They are the measurement.
What makes a real win rate worth comparing?
- Same variant: compare Turn 1 with Turn 1, and Klondike Solitaire (Turn 3) with Klondike Solitaire (Turn 3). Mixing them makes the result useless.
- Every deal counted: a win rate needs wins and losses. Restarted deals, abandoned deals, and “bad starts” belong in the denominator.
- Clear undo rule: unlimited undo is fine if everyone knows it is allowed. It is not the same record as no-undo play.
- Clear hint rule: hints reduce missed tactics. A hinted result can still be fun, but it should not be compared with unaided play.
- Enough games: 10 games tells you almost nothing. A run of 100 games begins to say something. Longer logs are better because Klondike has streaks.
- Fixed redeal rule: unlimited redeals are normal in many digital versions, but the rule must be stated before the number means anything.
- No selective reporting: the most common way to inflate a solitaire win rate is to remember the wins and quietly lose the failed starts.
- Deal evidence: a serious claim should include dates, deal identifiers if available, rules used, and the raw count of wins and losses.
How should I judge my own Klondike results without fooling myself?
Pick one rule set and keep it boring. If you are playing Turn 1 with unlimited redeals, undo, and hints, write that down once, then do not change it halfway through the run. Count every deal you start. If you quit after the opening layout looks grim, count it as a loss.
Start with 100 games. It is not a perfect sample, but it is large enough to cut through the noise of a lucky evening. Record wins, losses, and nothing more at first. You do not need a spreadsheet full of card movements. You need a clean denominator. If you win 34 of 100, your win rate is 34%. If you win 12 but quietly skipped 20 bad starts, you do not have a number.
After that, look for patterns rather than excuses. Are you opening tableau cards early? Are you preserving empty columns for kings with real follow-up, not just any king? Are you moving cards to the foundations too quickly and losing flexibility? Klondike punishes tidy-looking play when it blocks the next hidden card.
Use the records discussion as a guard rail. The 82% research figure is not there to make your 30% feel poor. It shows how many deals may be winnable under stronger information and analysis than a normal player has. Your job is smaller and more interesting: make fewer avoidable mistakes across real hidden deals.
Frequently asked questions
Yes. If you saw the layout and chose to stop, it belongs in your results. Otherwise you are measuring only the deals that looked promising. The fair exception is a misclick, interruption, or rule change before the game really began. Be consistent, because consistency matters more than the exact policy.
100 games is a sensible first sample. It will not prove your true level, but it is enough to smooth out a few lucky or brutal streaks. Ten games tells you almost nothing. If you want a cleaner picture, record several blocks of 100 and compare the blocks, not the individual evenings.
The common misconception is that undo automatically makes a result worthless. The correction is that undo only makes it a different category. A no-undo win rate and an undo-allowed win rate should not be compared as if they are the same thing. For learning, undo is useful; for measuring, write down that you used it.
You uncover a black 9, but both red 10s are buried under face-down cards in the same tableau pile. The move you need exists in the deck, yet you cannot reach it in time. That kind of blockage is why Klondike results never reflect skill alone. A good player reduces waste; they do not remove luck.
It depends on rules: draw count, redeals, undo, hints, and whether abandoned games are counted. Klondike Solitaire (Turn 1) with unlimited redeals is not the same test as Klondike Solitaire (Turn 3). If you both use the same rules and count every started deal, the comparison is meaningful enough for friendly rivalry.
FreeCell is the cleaner comparison because its cards are visible from the start, so planning carries more of the weight. Klondike hides information, which means two players can make sensible early choices and still be separated by buried cards. That does not make Klondike shallow. It means its win rate mixes judgement with uncertainty.
Count hints as assistance; the exception is when every player in the comparison is allowed the same hints. Hints can stop obvious mistakes, so they usually raise results over time. There is nothing wrong with using them on Veena Games, especially while learning, but label the record as hint-assisted if you are tracking seriously.
Most people stop recording the dull games. They remember the clever wins, forget the quick dead ends, and reset games that look awkward. Once you write every started deal down, the flattering gaps disappear. That drop is not failure. It is the first honest baseline you have had.
A win rate is the number of games won divided by the number of games counted, usually shown as a percentage. If you win 37 out of 100 recorded games, your win rate is 37%. The hard part is not the arithmetic. It is deciding which games count before you start playing.
Play one rule set, such as Klondike Solitaire (Turn 1), for a fixed block of games. Count every started deal. Mark whether you used undo or hints. Do not remove losses because you were tired. After the block, divide wins by total games. Then start a fresh block and see whether the pattern holds.
Nobody knows precisely, because hidden information and the huge number of possible deals make an exact answer difficult. The best-known published computer work gives a lower bound around 82% for draw-one Klondike under standard assumptions, not an exact total. Human win rates are lower because players do not know where every hidden card sits.
Microsoft included Solitaire in Windows 3.0 in 1990, and that version was Klondike. It put the game in front of office workers, students, and home users for decades. The rules were older than Windows, but that distribution made Klondike the default meaning of “solitaire” for many players.
The cost is false confidence. If you restart bad-looking deals, overuse hints without marking them, or avoid harder variants, the percentage rises while your judgement stalls. A useful win rate should make decisions clearer: which moves you miss, which rules you prefer, and whether you are improving over the next 100 games.
Final thoughts
If you want a number to chase, chase your own clean number. Not a claim from a forum. Not a solver’s winnability ceiling. Not a percentage from an app whose rules you cannot see. Take 100 Turn 1 deals, count every one, and write down the result without editing the ugly parts.
A good record has a small pleasure to it. It makes you more honest before it makes you better. You start noticing the losses you used to explain away: the early foundation move that trapped a needed card, the empty column filled because it felt productive, the stock cycle you rushed because the game looked stuck. Those are fixable. A vague win rate is not.
Do not be embarrassed by a modest percentage. Klondike is partly a game of hidden information, and some deals give you very little room. The useful question is not “Am I near the record?” because there is no official Klondike win-rate record to be near. The useful question is sharper: under the rules I actually play, am I making better decisions than I was 100 games ago?
-
Klondike
Klondike Turn 3 Rules: What the Three-Card Draw Changes
Klondike Turn 3 keeps the classic tableau and foundations, but the three-card stock draw makes planning stricter and wins harder.
18 August 2026
-
Klondike
Is Using Undo Cheating? Solitaire's Oldest Argument
Undo is cheating only if your agreed solitaire rules ban it; casino cheating is a different matter, with law, money, and victims involved.
16 August 2026
-
Klondike
Foundation Timing in Klondike: When Not to Play the Ace
A practical guide to Klondike foundation timing: play Aces freely, slow down on low cards, and use board clues before feeding suits upward.
15 August 2026