Alaska Solitaire: three deals in four, at the very least

Alaska is Yukon's close relative. Same deal, same absence of a stock pile, same power to move any face-up card along with everything on top of it. The difference is how you build: in Alaska a card goes on another of the same suit, one rank higher or one rank lower. A 6 of clubs can land on the 7 of clubs or the 5 of clubs.

That two-way rule gives you more places to put cards, and our solver's results reflect it. Playing 500 deals with every card visible, it found a win in 75.6% and proved 7.2% impossible, leaving 17.2% undecided. So between roughly three-quarters and 92.8% of Alaska deals can be won with perfect play.

No published study covers Alaska on these rules, so this is, to our knowledge, the first estimate of its ceiling. You can play Alaska free here, in any browser, with nothing to install.

Alaska's numbers

The Blake & Gent (2026) study, the most complete survey of solitaire winnability, does not include Alaska, so we have no outside figure to compare against.

Alaska against Yukon: more wins, and more dead ends

Compared with Yukon, our solver settled more Alaska deals in both directions. It proved more of them winnable (75.6% against 67.0%) and more of them impossible (7.2% against 3.4%), with fewer left undecided. Our reading is that the same-suit rule narrows the options on each move, which probably makes the search easier to finish, while the up-or-down rule adds options back, which would explain why so many deals can still be won.

For a player, the lesson is that Alaska is less about brute choice and more about suits. Once you see which suit each column is heading toward, the game becomes much easier to read.

Why the no-peeking figure is so low

Our simple player won only 1.5% of 10,000 deals. That is far below the ceiling, and it is mostly a weakness of the player, not the game. It searches 300 positions ahead using face-up cards only and never undoes a move. In Alaska, a move that looks good now often mixes a column's suits in a way that blocks it later, and the player has no way to take it back.

Treat that 1.5% as a floor for a very basic strategy, not as how often people win.

Rules as played here

• 52 cards in seven columns: one card in the first; 6, 7, 8, 9, 10 and 11 in the others, with the last five of each face up.
• No stock or waste. Every card is dealt at the start.
• Any face-up card can move, taking every card above it along.
• The moved card goes onto a card of the same suit, one rank higher or one rank lower.
• Only a King, with anything on it, may fill an empty column.
• The four foundations take each suit in order, Ace first and King last.

Playing like the solver

Read the columns by suit

Each face-up group tends to settle into one suit. Spotting which columns are becoming "hearts" or "spades" tells you where each new card belongs.

Check up and down before every move

Beginners build only downward from habit. The up-or-down rule often gives you a second landing spot; check both before deciding.

Dig where the most cards are hidden

The longer columns hold up to six face-down cards. Freeing them early reduces the solver's advantage over you.

Plan around the Kings

Kings are the only cards that can start a new column. A King stuck on top of low cards is the most common reason an Alaska game locks up.

Undo when a column goes mixed

If a group move leaves two suits tangled in one column, step back. That is exactly the mistake our player could not undo.

Alaska on the ceiling ladder

Alaska sits in the upper middle of the games our solver played with every card visible. Here is how it lines up with its neighbors on this site:

• Alaska: at least 75.6% solved, 7.2% proven impossible.
• Yukon: at least 67.0% solved, 3.4% proven impossible.
• Klondike Turn 1: at least 76.6% solved, 2.1% proven impossible.

So Alaska's floor of proven wins sits right alongside Klondike's, even though Alaska has no stock to fall back on. The difference is in the losses: Alaska had the most proven dead ends of the three, which means a stuck Alaska game is somewhat more likely to be truly stuck. If you have tried a few different group moves with Undo and nothing opens, dealing again is a reasonable call.

Inside the test, one step at a time

The solitaire-online.us team ran this study in September 2026 to estimate the ceiling for Alaska as this site plays it.

1. Reading the rules from our game code. The deal shape, no stock, group moves, same-suit up-or-down building and Kings-only empty columns come from this page's game files.
2. Rebuilding the game. Alaska runs on the same program as Yukon, with its own building rule.
3. Dealing the cards. Seeds 1 to 500 for the solver and 1 to 10,000 for the no-peeking player.
4. Choosing the players. A solver that sees every card sets the ceiling; a look-ahead player on visible cards only, never undoing, sets a floor.
5. Running it. Up to 50,000 positions per deal. Unlike the Yukon solver, the Alaska solver used no automatic foundation moves, so every move was chosen by the search.
6. Checking against published research. Alaska is not in Blake & Gent (2026). There is no published figure to test against.
7. Working out the margin of error. Our 95% range for deals solved, 71.6–79.2%, would contain the true value in 19 of 20 repeats.
8. What the numbers cannot tell you. With 17.2% undecided, the ceiling is a range, not a point. The no-peeking figure reflects a weak player, not typical human play.

The code behind the ceiling

The simulations are written in Python 3.11 with the standard library alone. Alaska shares yukon.py with Yukon; a single setting switches the building rule. The shared runner.py spread deals across 19 worker processes on a 20-core machine, and each deal was shuffled from a numbered seed.

The solver is a depth-first search that knows the face-down cards, tries each legal move in turn and backs up from dead ends. Because it makes no automatic moves, a deal counts as impossible only when every reachable position has been tried. The no-peeking player looks 300 positions ahead each turn, scoring positions by foundation cards, hidden cards remaining and correct versus incorrect links. The solver run of 500 deals took about three and a half minutes; the player's 10,000 deals took about a minute. Ranges use the Wilson score interval.

Claude, an AI assistant developed by Anthropic, helped us write and review the code and draft this text. Every number on this page comes from running the code, not from the AI.

Because each shuffle comes from a numbered seed, anyone with the same code gets exactly the same deals and the same verdicts.

By Admin, solitaire-online.us. Study date: September 18, 2026.

Ceiling questions from players

How many Alaska games can be won?

With perfect play and every card visible, between 75.6% and 92.8% of our 500 deals.

Does Alaska have a higher ceiling than Yukon?

By the ceiling, yes: more Alaska deals were proven winnable. It also had more proven dead ends, so it is a little more decisive both ways.

Can I build downward only?

You can, but you would be giving away half your options. Alaska lets you build up or down in suit.

Why is there no stock pile?

Like Yukon, Alaska deals everything at the start. What you see, plus the face-down cards, is all there is.

Which game should I try after Alaska?

Try Diplomat for a two-deck game, or see everything on the games page.