Alaska Solitaire: one deck, one suit to follow, very few landing spots
On a site built around the difference between one deck and two, Alaska is a useful extreme. It uses a single 52-card deck, and it builds in suit, one rank up or down. Put those together and each card has at most two places on the whole table where it can rest: the next card up and the next card down of its own suit. With two decks, that number would double. With one, it does not.
That scarcity shows in our numbers. A program that looks 300 positions ahead each turn, using only visible cards and never undoing, won 1.5% of 10,000 deals (95% range 1.3%–1.8%). A depth-first search that is allowed to know the face-down cards proved 75.6% of 500 deals winnable (95% range 71.6%–79.2%). We did not simulate a two-deck Alaska, so we cannot tell you what doubling the deck would do to this game.
At a glance: single deck, same-suit building
| Measure (one 52-card deck) | Result | 95% range | Sample |
|---|---|---|---|
| Search that knows hidden cards: proven winnable | 75.6% | 71.6–79.2% | 500 deals |
| Same search: proven impossible | 7.2% | 5.2–9.8% | 500 deals |
| Same search: budget ran out | 17.2% | 14.1–20.8% | 500 deals |
| Look-ahead program, visible cards, no Undo: won | 1.5% | 1.3–1.8% | 10,000 deals |
| Cards home when the program stopped | 5.1 of 52 | 4.9–5.2 | 10,000 deals |
| Moves in the program's wins | 110.9 | 108.9–112.9 | 151 wins |
With perfect knowledge, somewhere between 75.6% and 92.8% of deals can be won. The program's 1.5% is a floor for a weak player without Undo. Alaska is not covered by Blake and Gent (2026), so there is no published check.
Why a second deck would matter here (and why we can't measure it)
Elsewhere on this site, doubling the deck gives every card a twin and every suit two foundations. Our one-deck Klondike program won 42.3% (41.3–43.3%, 10,000 deals); the same kind of program won 81.5% of two-deck Solitaire Double deals (80.7–82.2%, 10,000 deals). Alaska's same-suit rule makes landing spots even rarer than Klondike's, so a second copy of each card would in principle help a lot. But no two-deck Alaska exists on this site and we ran none, so that stays a reasoned guess, not a figure.
Alaska rules on this site
- Deck: one standard 52-card deck.
- Tableau: seven columns: one card, then six to eleven, the last five of each face up.
- Stock: none.
- Foundations: four, by suit, Ace to King.
- Building: same suit, one rank higher or lower.
- Moving: any face-up card, with everything on top of it. Only the lifted card must fit.
- Empty columns: Kings only.
Making the most of a single deck
Count a card's two homes
Before you move a card, name both cards it could sit on. If one of them is face down and the other is about to be buried, that card may have nowhere to go later.
Build both ways
The 9 of diamonds can take the 8 or the 10 of diamonds. Going up is easy to forget, and it is half your options.
Uncover first
The program usually stopped with about 5 cards home and key cards still face down. Group moves that turn a card over should come before anything cosmetic.
Don't over-read the Aces
About 1.6 of the four Aces start face down. The program won 1.9% (1.6–2.3%, 5,352 deals) with two or more hidden and 1.1% (0.8–1.4%, 4,648 deals) with fewer. Small numbers; we would not plan around them.
Undo is your search
From 1.5% to at least 75.6% is the value of seeing further. Unlimited Undo lets you test a pile move and back out.
Playing on freecell-two-decks.com
Alaska is free in any current browser on computer, tablet or phone, with no download or sign-up. You get Hint, unlimited Undo, Restart, Auto Play, zoom and fullscreen, drag or tap moves, a wins, time and moves counter, and sound and table-color settings.
How the figures were made
The freecell-two-decks.com team ran these simulations in September 2026 on this site's own Alaska rules.
- Rules from the code: Yukon's deal and pile moves, with building changed to same suit, one rank up or down, and Kings only in gaps.
- A Python program shared with Yukon, with the building rule switched.
- Numbered deals: seeds 1 to 10,000 for the program, 1 to 500 for the search.
- Two players. A 300-position look-ahead on visible cards with no Undo, and a depth-first search that knows every card, with up to 50,000 positions per deal and no automatic foundation moves.
- No two-deck version and no published figure. Blake and Gent (2026) do not include Alaska.
- 95% ranges: Wilson score intervals for percentages, normal intervals for averages.
Claude, an AI assistant by Anthropic, helped write and review the code and draft this page. Every figure comes from running the code, and the team checked them.
Alaska questions
Is Alaska harder than Yukon?
For short look-ahead play, yes: 1.5% here against 4.5% in Yukon. With full knowledge, more Alaska deals were proven winnable (75.6% against 67.0%).
Are some deals lost from the start?
Yes, 7.2% were proven impossible even seeing every card.
Is there a two-deck version?
Not on this site. For a same-suit two-deck game, try Forty Thieves.
What next?
Our home game, FreeCell Two Decks, shows what 104 open cards feel like. Every game is on the games page.
By Admin, freecell-two-decks.com (Forty Media). Research run September 2026.