Short answer: adding a second deck roughly doubles the cards but more than doubles the work. In simulations our team ran in September 2026, the first solution a solver found for FreeCell Two Decks averaged 148.0 moves (95% range 145.9-150.1, from 352 solved deals), while one-deck FreeCell averaged 62.2 moves (95% range 61.9-62.4, from 9,913 solved deals). That is about two and a half times the moves for twice the cards. The game is not just longer. It is deeper, slower to unlock, and far less forgiving of a careless early move.
The two games side by side
Here are the figures from our own runs, on this site's own rules. Read the ranges, not just the headline numbers.
| Measure | FreeCell (one deck, 52 cards) | FreeCell Two Decks (104 cards) |
|---|---|---|
| Deals tested | 10,000 | 400 (seeds 101-500) |
| Search budget | 200,000 positions, then a second pass at 1,000,000 | 150,000 positions |
| Solved by the solver | 99.7% (99.5-99.8%) | 88.0% (84.4-90.8%) |
| Proven impossible | None | 0.0% (0.0-1.0%) |
| Undecided (budget ran out) | , | 12.0% (9.2-15.6%) |
| So winnable share is | at least 99.7% (99.5-99.8%) | somewhere in 88.0-100% |
| Moves in the first solution found | 62.2 (61.9-62.4), 9,913 solved deals | 148.0 (145.9-150.1), 352 solved deals |
| An Ace playable on move one | 49.0% (48.0-49.9%) | 55.0% (50.1-59.8%) |
One honest caveat before you go further: these two columns did not get the same treatment. The one-deck game was tested on 10,000 deals with a generous budget (200,000 positions, then a second pass at 1,000,000 for anything left over). The two-deck game was tested on 400 deals with a single 150,000-position budget. Smaller sample, smaller budget, much bigger game. That comparison is useful, but it is not apples to apples, and we will not pretend otherwise.
Why 88.0% does not mean 12% of deals are dead
This is the most important line in the whole post. In the two-deck run, 0.0% of deals were proven impossible (0.0-1.0%). Not one deal was shown to be a loss. The 12.0% (9.2-15.6%) that the solver did not crack are undecided: the search hit its 150,000-position ceiling and stopped looking.
A two-deck board has vastly more possible positions than a one-deck board, so the same budget buys far less certainty. The lower solved share reflects a smaller budget spent on a longer game, not deals that have been lost. That is why we report the winnable share as a range, 88.0-100%, and never as a single tidy number.
How we got these numbers
Our team ran the simulations in September 2026 against the exact rules used on this site. The code is Python, written with help from an AI assistant (Claude, by Anthropic) for the code and for drafting; every figure here comes from actually running that code, not from an estimate.
Three things you should know about what solver numbers mean:
- A simple bot's win rate is a floor. It shows what mechanical play achieves, not how often a thoughtful person wins.
- A solver that can see every card is a ceiling. It plays with knowledge no human has at the table, so its success rate is the best case, not your case.
- Undecided deals turn an answer into a range. When a search runs out of budget without a verdict, the truth sits somewhere between "solved so far" and "all of them", and we print both ends.
What the published research says (and does not say)
For the one-deck game there is excellent peer-reviewed work. Charlie Blake and Ian Gent's study in the Journal of Artificial Intelligence Research (Vol. 85, February 2026) puts single-deck FreeCell's winnability at 99.998881% and classic Klondike's "thoughtful" winnability at 81.945% ± 0.084%, see Blake & Gent, JAIR 2026. That paper also notes how enormous this pastime is in everyday American life, citing Microsoft's figure of Windows Solitaire being played roughly 100 million times a day in 2020.
There is no published figure for two-deck FreeCell. It is not in Blake and Gent, and we have not found it elsewhere. Our 88.0-100% range is our own measurement on our own rules, offered as a starting point, not as a settled fact.
It is also worth separating solver math from what real people score. Data published by solitairex.io covering 913,106 games of Classic Solitaire in a single month showed 409,374 wins, a real-player win rate of about 45%, against that 81.945% theoretical ceiling for Klondike. Human results always sit well below what a perfect-information solver manages. Expect the same gap in two-deck FreeCell.
What the second deck actually changes at the table
Numbers are only useful if they change how you play. Here is what 104 cards feel like in the hand.
1. Every rank appears twice, so "which copy?" becomes a real decision
In one-deck FreeCell there is exactly one black Seven you need. With two decks there are two of each card and two full sets of foundations to build. A card buried at the bottom of a deep column is no longer a crisis if its twin is sitting near the top somewhere else. Before you dig, look for the duplicate.
2. Openings are friendlier, endings are harsher
An Ace was playable on the very first move in 55.0% (50.1-59.8%) of two-deck deals, against 49.0% (48.0-49.9%) in the one-deck game. More Aces means a better chance one is loose at the start. That early comfort is misleading, the hard part arrives 100 moves later, when free cells are full and every column is long.
3. Free cells are worth more, not less
With longer columns, an occupied free cell blocks a bigger chunk of the board. Treat filling the last free cell as a decision you should be able to justify out loud, and clear cells back out as soon as a home becomes available.
4. Pace yourself for 148 moves, not 62
A one-deck game is a coffee break. A two-deck game is closer to a chapter of a book. If you play in short bursts, that is fine, the board waits for you, and nothing on this site is timed.
5. Don't race both foundations at the same speed
Sending every low card home feels productive, but low cards in the tableau are the glue that lets you build sequences. Hold a Two or Three back if it is doing useful work on a column.
Games with no counterpart here, and we won't guess
Our whole theme is one deck against two, but some games only exist on one side of that line. Spider Solitaire, Forty Thieves and Diplomat are two-deck games with no one-deck version on this site, so there is nothing to compare them to. Most of our one-deck games, Golf, Pyramid, Yukon and the rest, have no two-deck counterpart measured. Where we have not measured both sides, we say so and leave the number out rather than invent one.
How to play here
Everything on this site is completely free, with no ads of any kind, no pop-ups and no paywalls. There is nothing to download or install, no sign-up, no account and no email address to hand over. Open a page and the cards are dealt. The games exist simply for enjoyment: no gambling, no in-app purchases, and no timer nagging you to hurry.
- Big, clear cards. Large playing cards and readable text, comfortable for older players and anyone with tired eyes, plus a zoom control to make the board bigger.
- Hint when you are stuck, and unlimited Undo so you can test an idea and step back. On a 148-move game, Undo is not cheating, it is how you learn a board.
- Restart the same deal to try a different plan, and Auto Play to finish a game once it is clearly solved.
- Drag and drop or simply tap to move cards, whichever suits you.
- Wins, time and moves are tracked for the current game, so you can see how your own count compares with that 148.0 average.
- Sound can be switched off, fullscreen is available, and you can change the table color.
It runs in any modern browser on desktop, tablet and phone, and every feature is available to everyone. Start at the home page, or browse the full list on the games page.
Frequently asked questions
How many cards are in FreeCell Two Decks?
104, two standard 52-card decks combined, so every card appears twice and there are two complete sets of foundations to build.
Is double-deck FreeCell harder than regular FreeCell?
It is longer and more demanding to plan, yes. Whether it is less winnable is still an open question: our solver proved 0.0% of deals impossible (0.0-1.0%) but left 12.0% (9.2-15.6%) undecided on a 150,000-position budget across 400 deals, so the winnable share sits somewhere in 88.0-100%.
How long does a two-deck game take?
The first solution our solver found averaged 148.0 moves (145.9-150.1, from 352 solved deals). Human play usually takes more moves than that, because we backtrack. There is no clock pressure here either way.
Is there a published win rate for two-deck FreeCell?
Not that we know of. The Blake and Gent JAIR study covers single-deck patience games, and we found no peer-reviewed figure for the two-deck version. The numbers in this post are our own.
Where should I start if two decks feels like too much?
Play a few rounds of one-deck FreeCell first to drill the habit of keeping free cells empty, then move to FreeCell Two Decks. If you would rather ease in with a familiar layout, Klondike and its two-deck relative Double Solitaire make the same jump in a gentler way.
Do I need to create an account to save my stats?
No. There is no account and no sign-up anywhere on the site. Wins, time and moves are tracked for the game you are currently playing.
By Admin, for the freecell-two-decks.com team. More articles on the blog.
Sources
- Blake, C. & Gent, I. (2026), The Winnability of Klondike Solitaire and Many Other Patience Games, Journal of Artificial Intelligence Research, Vol. 85
- SolitaireX.io. Data Insights and Strategies to Improve Your Win Rate
- University of St Andrews Research Portal. The winnability of Klondike solitaire and many other patience games