Alaska Solitaire: Yukon's deal with a same-suit twist
Alaska uses exactly the same layout as Yukon: seven columns, no stock, every card on the table from the start, and the freedom to move any face-up card along with everything piled on top of it. Only one thing changes, and it is how cards build. In Alaska a card goes on another card of the same suit, one rank higher or lower. So the 7 of clubs can take the 6 of clubs or the 8 of clubs.
That up-or-down freedom sounds generous, but the same-suit rule narrows each card's options to two specific cards. Our computer player's wins took 110.9 moves on average, about 14 more than in Yukon, and in a typical game it only got 5.1 cards home. A solver that could see every card did far better, winning at least 75.6% of deals.
What an Alaska game costs in moves
| What we measured | Alaska | Yukon, for comparison | Deals |
|---|---|---|---|
| Moves in a won game (our computer player) | 110.9 (95% range 108.9–112.9) | 96.9 | 151 and 445 wins |
| Cards on the foundations at the end, all games | 5.1 of 52 | 7.4 | 10,000 each |
| Games won, no Undo, no peeking | 1.5% (95% range 1.3–1.8%) | 4.5% | 10,000 each |
| Solver that sees every card: solved | 75.6% | 67.0% | 500 each |
| Solver: proven impossible | 7.2% | 3.4% | 500 each |
| Solver: undecided within its limit | 17.2% | 29.6% | 500 each |
Put together, somewhere between 75.6% and 92.8% of Alaska deals can be won with perfect play. Our blind, no-Undo player gets a tiny share of that, so read its 1.5% as a floor and nothing more.
Alaska rules on solitaire-online.nz
- Deal. Seven columns of 1, 6, 7, 8, 9, 10 and 11 cards. The single card in the first column is face up; in the others, the bottom five are face up.
- Building. Place a card on one of the same suit that is one rank higher or one rank lower.
- Moving piles. Pick up any face-up card with everything on top of it, as long as the card you pick up fits where it lands.
- Turning over. When a column's last face-up card goes, the next one turns face up.
- Empty columns. Only a King, and whatever is on it, can fill one.
- Foundations. Build each suit from Ace to King. No stock, so everything you need is somewhere in the columns.
Where the extra moves come from
In Yukon, a red 6 can go on either black 7. In Alaska, the 6 of hearts has exactly two homes: the 5 or the 7 of hearts. If both are buried, the 6 waits. Digging those cards out costs moves, which is why wins run longer and our player struggled more.
Shorter games, better odds
Map each suit before you start
Look along the face-up cards for same-suit neighbours. Two or three ready-made links are a good start.
Use up and down together
A pile can run downwards (8, 7, 6 of spades) or upwards (6, 7, 8 of spades). When one direction is blocked, the other is often open, and that flexibility is how long piles get unpicked in Alaska.
Dig where the Aces are
On average 1.6 Aces start face down. Unlike Yukon, our player actually did slightly better when two or more Aces were buried (1.9% against 1.1%), a quirk of how deals fall. Either way, getting at the Aces early keeps the foundations moving.
Back up and try the other direction
Our solver shows most deals are winnable. When you stall, back up and try the other direction.
Play on any device
Alaska runs in any modern browser on phone, tablet or computer. Nothing to download, no account and no email. Tap or drag to move, zoom in on long columns, and use Hint, unlimited Undo, Restart and Auto Play whenever you like. Your moves and time for the current game are counted. Sound, fullscreen and table colours can be changed to suit.
What an Alaska session looks like
An Alaska game starts like Yukon: dig out the face-down cards using the big move. The difference appears once the cards are showing. Because each card only has two possible homes, you spend more time looking for the right card and less time moving piles around.
That slower middle is why wins take longer. But it is also where Alaska rewards patience, because once a same-suit run is in place it is very tidy and easy to send home.
Across all games our player managed about five cards home on average, a sign of how much planning Alaska asks for. The solver's results tell the other side of the story: most deals can be won by someone who sees the way through.
How we did the research, step by step
The solitaire-online.nz team ran this study in September 2026 to see how Alaska's same-suit rule changes the length of a game and the share of deals that can be won.
- We began with this site's game code. The same deal as Yukon, but cards build in the same suit one rank up or down, any face-up card moves with what is on top of it, and only Kings fill empty columns.
- The game as a program. Alaska and Yukon share one simulation, so the only difference between them is the building rule.
- Numbered deals. 10,000 deals for the computer player and 500 for the solver, each from a numbered seed.
- Two players. A computer player that only sees face-up cards, never uses Undo and looks up to 300 positions ahead; and a solver that sees every card.
- The run. 10,000 games for the player; up to 50,000 positions per deal for the solver.
- Published figures. We did not find a published winnability figure for Alaska, so these results are our own.
- Error bars. A 95% range contains the answer 19 times in 20 on fresh deals. For moves per win it is 108.9 to 112.9, fairly wide because there were only 151 wins to average.
- Limits. Our player is weak at Alaska, so its win rate is only a floor. The solver left 17.2% of deals undecided, so its answer is a range. Move counts cover won games only.
The technical side
The simulation is written in Python 3.11 using the standard library only, in yukon.py with Alaska's up-or-down same-suit rule switched on. The computer player looks up to 300 positions ahead over the visible cards before each move, rewarding cards on the foundations and penalising cards still hidden. The solver is a depth-first search limited to 50,000 positions per deal; unlike our Yukon solver it makes no automatic foundation moves at all, so every move it considers is its own choice. Using runner.py across 19 worker processes, the player's 10,000 games took about a minute and the solver's 500 deals about three and a half minutes.
Error bars are Wilson score intervals. Claude, an AI assistant made by Anthropic, lent a hand writing and reviewing the Alaska code and drafting these paragraphs. The numbers are the program's output, not the AI's, and the team checked them before they went live.
By Admin, solitaire-online.nz. Study date: 18 September 2026.
Short answers before your next deal
Is Alaska a good game for a short break?
It can be, as long as you do not mind a slower middle. There is no stock to go through, so a game is just the seven columns. Our player's wins averaged 110.9 moves, a little longer than Yukon but far shorter than a two-deck game like Diplomat. If you only have a few minutes, a deal or two of Alaska fits well, and the game keeps your place in the browser if you need to stop.
What separates Alaska from Yukon?
Only the building rule: Alaska builds in the same suit, up or down, where Yukon builds down in alternating colours.
How many moves should I expect for an Alaska win?
Our player's wins averaged 110.9 moves, about 14 more than Yukon.
Can most deals be won?
Yes, with good play. Our solver won at least 75.6% and proved only 7.2% impossible.
Can I play on a tablet?
Yes, in any modern browser. Tap a card and then its destination.
Where to after Alaska?
Yukon for the original, Baker's Solitaire for another same-suit challenge, or anything on the games page.