Alaska Solitaire: one rung below four-suit Spider
Alaska is Yukon with a twist that Spider players will appreciate. The deal is the same: seven columns, no stock, every card on the table from the start, and the freedom to move any face-up card with everything piled on it. But instead of alternating colors, you build in the same suit, and you may go either up or down by one. A 7 of hearts can take the 6 of hearts or the 8 of hearts.
On our difficulty ladder, Alaska sits in an odd spot. Our simple program, which sees only face-up cards and never undoes, won just 1.5% of 10,000 deals. Only four-suit Spider went lower. But our solver, which sees every card, proved 75.6% of deals winnable, about the same as classic Klondike (76.6%). Alaska's deals are kind; it is the planning that is tough.
Below: where Alaska ranks, the rules here, what to take from our numbers, and a step-by-step account of the study.
Where Alaska ranks
| Game | Simple program won | Proven winnable, every card known | Still undecided |
|---|---|---|---|
| Spider, 2 suits | 5.7% | not measured | – |
| Yukon | 4.5% | 67.0% | 29.6% |
| Alaska (this page) | 1.5% (1.3–1.8%) | 75.6% (71.6–79.2%) | 17.2% |
| Spider, 4 suits | 0.2% | 98.5% (published) | – |
So between 75.6% and 92.8% of Alaska deals can be won with perfect knowledge; 7.2% are proven impossible. Our simple program is weak at this game, and we do not present its 1.5% as how often a person wins. It shows how unforgiving Alaska is to play without looking ahead or undoing.
How Alaska works here
- Layout: seven columns: one card in the first, six to eleven in the others, with the last five of each face up.
- No stock: every card starts on the table.
- Foundations: four, built up by suit from Ace to King.
- Building: same suit, one rank higher or lower.
- Moving: pick up any face-up card and everything on top of it comes too. Only the card you pick up needs to fit.
- Empty columns: only a King, with anything on top of it.
What the numbers tell you
Think in suits, like Spider
Every build in Alaska is a same-suit build. That is good news: every link you make is a "clean" link of the kind Spider players work so hard for. The challenge is that a card has only two possible places to go in the whole tableau.
Remember cards can climb as well as fall
Up-or-down building means a card can climb as well as fall. If the 5 of spades cannot go on the 6, look for the 4. Beginners often forget the "up" option.
Uncover hidden cards above all
Across all 10,000 games our program averaged about 5 cards on the foundations. It stalled mostly because the cards it needed were still face down. Moves that turn a card over should come first.
Expect a hundred-plus moves in a win
Our program's winning games took about 111 moves. Alaska rewards a long, patient game.
Let Undo do your looking ahead
The gap between 1.5% for our program and at least 75.6% for the solver is mostly about seeing ahead. You cannot see face-down cards, but unlimited Undo lets you try a move, see what it reveals, and change course.
Features on spidersolitaire.us
Alaska is free on this site and runs in any modern browser on computer, tablet or phone. No download, account, sign-up or email address. Large clear cards with a zoom control, Hint, unlimited Undo, Restart, Auto Play, and drag-and-drop or tap. Wins, time and moves are tracked for the current game; sound can be switched off; fullscreen and table colors are available.
Behind the numbers: how we ran the study
The spidersolitaire.us team studied Alaska in September 2026 alongside Yukon, to see how same-suit building changes a game that already resembles Spider.
- Started from the code that runs our game. Alaska uses Yukon's deal and pile moves, with building changed to same suit, one rank up or down. Only Kings fill empty columns.
- Rebuilt the game as a program. Alaska shares its program with Yukon, with the building rule switched.
- Dealt numbered hands. Deals 1 to 10,000 for the simple player and 1 to 500 for the solver, each from a fixed seed so anyone rerunning the test gets identical cards.
- Chose two players. A simple program that uses only face-up cards and never undoes (a floor), and a solver that sees every card (a ceiling).
- Ran the deals. The simple player completed all 10,000. The solver was allowed 50,000 positions per deal; unlike the Yukon solver, it made no automatic moves to the foundations.
- Looked for published research. The study we use as our reference, Blake and Gent (2026), does not cover Alaska, so there is nothing to check our figures against.
- Calculated the margin of error. The solver's 75.6% has a 95% range of 71.6% to 79.2%; the simple program's 1.5% has 1.3% to 1.8%.
- Stated the limits. 17.2% of solver deals were left undecided, the solver sees hidden cards, and the simple program is weak. We also saw a small, puzzling pattern: the simple program won slightly more often when more Aces started hidden (1.9% against 1.1%). We report it but would not build any advice on it.
The code behind the ladder
Everything is written in Python 3.11 using just the standard library, in the yukon.py module. The simple player looks up to 300 positions ahead each turn using only visible cards, scoring each position with a large bonus for cards on foundations, a penalty for hidden cards, and small points for good and bad links. The solver is a depth-first search that explores legal moves until it finds a win, proves there is none, or reaches its budget. On 19 worker processes on a 20-core machine, the 10,000 player games took just over a minute and the 500 solver deals about three and a half minutes.
Each 95% range on this page is a Wilson score interval. The team used Claude, an AI assistant from Anthropic, to help write and review the simulation code and draft this text. The AI did not produce any figures; each comes from running the code, and we reviewed them before publishing.
Ladder questions, answered with data
Alaska or Yukon: which rung is higher?
For a simple player, Alaska sits higher: our program won 1.5% here against 4.5% at Yukon. The deals themselves lean the other way, since within the same budget our solver proved 75.6% of Alaska deals winnable against 67.0% of Yukon deals. The only rule difference is building in suit, up or down, instead of alternating colors down.
Why is Alaska hard if most deals can be won?
Because each card has very few places to go, and many cards start hidden. Winning takes planning several moves ahead.
Are some Alaska deals hopeless?
No. Our solver proved 7.2% impossible even with every card visible.
Is it good practice for Spider?
Yes. Both reward keeping cards in suit. Try four-suit Spider when you want the ultimate test.
Where are the other games?
On the games page. Our longer write-ups are on the blog.
By Admin, spidersolitaire.us. Research run September 2026.