FreeCell Two Decks: a long game won by planning
FreeCell Two Decks takes the open-table idea of FreeCell and scales it up. There are 104 cards, all face up, in ten columns, with eight free cells and eight foundations. There is no stock and nothing hidden, so every decision you make is based on complete information. That makes it a game of strategy rather than luck. What decides a game is whether you can hold a plan together across a very long sequence of moves.
How much does planning pay off? We had a search program play 400 deals on this site's rules. It found a full winning line in 88.0% of them (352 of 400; 95% range 84.4%–90.8%) and proved none impossible (95% range 0.0%–1.0%). The other 12.0% (95% range 9.2%–15.6%) were deals where it ran out of thinking budget before finishing. In other words, the only thing that stopped it was how far it could plan.
We did not run a greedy player on this game, so there is no greedy-versus-planner pair to show here. What the figures do show is that on this table, the limit is how much planning you do. It is not the deal.
At a glance: what the solver found
| Measure | Result | 95% range | Deals |
|---|---|---|---|
| Solved with a 150,000-position search | 88.0% | 84.4%–90.8% | 400 |
| Proven impossible | 0.0% | 0.0%–1.0% | 400 |
| Ran out of search budget | 12.0% | 9.2%–15.6% | 400 |
| Length of first solution found | 148.0 moves | 145.9–150.1 | 352 solved |
| An Ace playable on move one | 55.0% | 50.1%–59.8% | 400 |
Compare the one-deck game: the same kind of solver, with a bigger budget, solved 99.7% of 10,000 FreeCell deals (95% range 99.5%–99.8%). Two decks do not make deals less fair. They make the plan longer, and a longer plan takes more thinking to find.
The rules as played here
- Cards: two standard decks, 104 cards, every one face up.
- Tableau: ten columns, four with eleven cards and six with ten.
- Free cells: eight, one card each.
- Foundations: eight, two for each suit, built up from Ace to King.
- Building: down in alternating colors.
- Empty columns: accept any card or legal run.
- Stock: none. The whole game is on the table from the start.
Thinking ahead on a 104-card table
Map the Aces before you move
In 55.0% of our deals an Ace could go home at once. In the rest, spend the first minute finding all eight Aces and the 2s. Choose the dig that frees two low cards at once over the one that frees only one.
Plan for about 150 moves
The solver's first winning line averaged 148.0 moves, more than double one-deck FreeCell's 62.2. A person usually needs more. Think in stages: first an empty column, then the low cards, then long runs.
Eight cells fill faster than you think
Doubling the cells also doubles the cards competing for them. A quick grab that parks a card "for now" is how an easy deal goes bad. Before using a cell, name the card that will take the parked card back out.
Choose which twin to build on
Every card has a twin. A red 7 might go on either black 8, and the choice decides which column you can empty later. A planner picks the copy that leaves the more useful column open.
Undo is your look-ahead
Our solver tries lines and backs out of dead ends. With unlimited Undo you can do the same: test a line, and if it jams, step back to the last position you liked.
Short-sighted habits that cost games
- Filling cells in the opening. Eight free cells can be full within a dozen moves, and from then on you move one card at a time.
- Racing one foundation. Rushing one suit's pile up can strand low cards of the other color that you still need as landing spots.
- Dropping a single card into an empty column. A gap is worth most when a long run moves into it.
- Restarting instead of rewinding. None of our deals was proven impossible. A stuck game usually needs a different plan from an earlier point, not a new deal.
Playing on spidersolitaireonline.us
FreeCell Two Decks is free, runs in any current browser on computer, tablet or phone, and needs no account. Fullscreen and zoom keep ten busy columns readable. You get Hint, unlimited Undo, Restart and Auto Play, drag or tap to move, a counter for wins, time and moves, and settings for sound and table color.
How we ran the study
The spidersolitaireonline.us team ran this test in September 2026 on this site's own two-deck FreeCell rules. The program is written in Python. An AI assistant, Claude by Anthropic, helped write and review the code and draft this page. Every figure comes from running the code.
- Rules taken from the site's code: ten columns of eleven and ten, eight cells, eight foundations, alternating colors, any card in a gap.
- A program that only makes legal moves on those rules.
- Separate deals for tuning and testing. We tuned one scoring setting on deals 1 to 8, then measured on deals 101 to 500, which played no part in the tuning.
- A solver rather than a bot. Every card is visible, so the solver sees what you see. It searched up to 150,000 positions per deal, always exploring the most promising position first.
- Every deal sorted into solved, proven impossible or undecided.
- No published figure to check against. Blake and Gent (2026) cover one-deck FreeCell but not this game.
- Margins of error are 95% Wilson score intervals. 400 deals is a smaller sample than most of our tests, and 12.0% were undecided, so the true winnable share lies somewhere between 88.0% and 100%.
Questions about FreeCell Two Decks
Is every deal winnable?
We proved none impossible in 400 deals, and 88.0% were solved outright. The rest needed more search than we allowed.
Is it harder than one-deck FreeCell?
It is longer. The solver's solutions averaged 148.0 moves against 62.2, and it needed more search to finish.
How many cells should I keep free?
As many as you can, for as long as you can. In one-deck FreeCell, going from four usable cells to two cut the solved share from 99.7% to 74.0%. The same principle applies here.
What should I play next?
Start with one-deck FreeCell if you are new. For a two-deck game where the stock hides cards, try Forty Thieves. For the biggest greedy-versus-planner gaps on the site, go to Spider. Everything else is on the home page.
By Admin, spidersolitaireonline.us (Solitairen OOD). Research run September 2026.