Fair Solitaire
Spider 4 Suits
Every game can be won. We checked.
Game
#
Moves
0
Time
0:00

How to play Spider Solitaire with four suits

Four-suit Spider is the game in full: two complete decks, clubs, diamonds, hearts and spades, and eight runs to build, each in a single suit. The rules are those of the one-suit game; what changes is how rarely a card lands somewhere it can stay. Every game you are dealt here can be won. New to Spider? The one-suit and two-suit games are the way in.

Why four suits is hard

Each suit appears 26 times, two of every card, so there are two runs of each suit to complete. For any card, the two decks hold eight cards one rank higher, and only two of them are of its own suit. Three places in four that a card can go therefore leave a joint behind, a point where one suit meets another and the column stops moving as one. With two suits it was one place in two, and with one suit none.

The game cannot be won without joints, but each one has to be undone before the run beneath it can be finished, and that takes a card of exactly the right suit and rank, or an empty column. With four suits there are rarely enough of either, and meanwhile the stock keeps coming, five rows that bury every joint not yet undone.

The solver's own games show the scale of it. In its first 500 four-suit wins, 31.3% of its card moves put a card on another suit, a median of 65 per game, against 17% with two suits. Its games took a median of 208 card moves, almost twice the one-suit figure, and the first run left the table only after a median of 137 moves. In 276 of the 500 games, not one run was complete when the last row was dealt. Four-suit Spider is won at the end, in a long unwinding of the columns after the stock is gone.

The rules, unchanged

Nothing in the rules changes with four suits: any card on any card one rank higher, only runs of one suit moving together, anything into an empty column, complete runs of one suit leaving by themselves, no deal while a column is empty. The one-suit page sets them out in full, and the two-suit page explains joints and when to make them.

Strategy for four suits

What the opening layout predicts

With one or two suits, the ten face-up cards at the start tell you almost nothing. With four they do, because the moves they offer are the cheap ones, made before the first deal. These are our solver's win rates on all 5,000 four-suit deals, grouped by what each deal held at the start.

Share of deals the solver won, by what the deal held (four suits, all 5,000 deals)
In the dealWon
No face-up card can go on one of its own suit (938 deals)41.5%
One or two can (3,143 deals)46.6%
Three or more can (919 deals)51.0%
Three face-up cards or fewer can go anywhere (1,067 deals)41.2%
Six or more can (1,917 deals)49.6%
No Ace face up (2,137 deals)47.0%
Two Aces face up (707 deals)46.5%
No King at the bottom of a column (2,205 deals)44.2%
Two Kings at the bottom (656 deals)48.9%
Six Kings or more in the stock (540 deals)39.8%

The best sign you can see is how many face-up cards can move at the start, and onto what. When none had a place of its own suit, deals were won 41.5% of the time; when three or more had, 51.0%, almost ten points better. The number of moves and their suits both count: among deals with three opening moves or fewer, those with at least one in suit were won 43.2% of the time (674 deals), those with none 37.9% (393). Face-up Aces, which feel like good news, make no difference: an Ace is the last card of a run and matters only at the end.

The rest is hidden. Two Kings at the bottom of columns are worth nearly five points, and six or more Kings in the stock cost about eight. Even the worst group was won by the solver two times in five, and the solver is not the limit of what can be won, so no opening is a reason to ask for another game.

Why the solver gives up

Our solver played the first 5,000 four-suit deals and won 2,324 of them, 46.5%. Each solution was replayed move by move by the game, and New game deals only from that list. That is far below the 92.8% it reaches with one suit, and the gap says more about the solver than about the deals. It follows the most promising positions first and gives up after 60,000 of them. With four suits few moves are clearly safe, many positions score almost the same, and the way through is long, so it reaches that limit more often than it finds the end.

The published study by Blake and Gent found almost every Spider deal winnable, four suits included, so most of the deals not dealt here are winnable too; we have simply not proved it yet. Our own lists point the same way. A deal number lays out the same ranks in the same places whatever the number of suits, and of the 783 deals among the first 5,000 that the solver could not win with two suits, it won 311 with four, the harder game. Among the four-suit deals, the first it did not win was number 1, the longest run of deals it won in a row was 11, and the longest run it missed was also 11.

One caution about the solution figures on this page: the 500 games are ones the solver won, which tilts them towards deals it found easier, and it does not look for the shortest path, so they describe its games, not the fewest moves a deal needs. The numbers for every game.

Game numbers

Every game has a number, and the same number always deals the same cards. Add it to the address, for example #2, to play or share a specific game. On the one-suit page, when it is on that list too, the same number gives the same layout with every card a spade: a good way to see the shape of a deal that beat you. A number without a known solution gives you a winnable game instead.

Frequently asked questions

Is it possible to win Spider with four suits?

Yes. Every game dealt on this page has been won by our solver and replayed move by move, and the published research finds almost every four-suit deal winnable. It is a long game, though: even the solver, which sees every card, needed a median of 208 card moves.

How many games do experts win?

There is no count we can check, and quoted figures depend on whether Undo was allowed and whether every deal could be won. Our solver, which sees every face-down card and the whole stock, found a way through 46.5% of the first 5,000 deals within its limit. Here you only get those, so a lost game could always have been won.

Why does the solver give up on so many four-suit deals?

It stops after 60,000 positions, and four suits multiply the positions worth trying. With one suit almost every build is safe; with four, three builds in four leave a joint, and many more orders of play must be tried before one works. The deals it gives up on are not proved impossible; they are simply not dealt here.

Why are there two of every card?

Spider is played with two full decks, so each of the 52 cards appears twice and there are two runs to complete in each suit. Either copy of a card can go into either run.

Should I start with four suits?

Only if you already know Spider well. One suit teaches the moves and two suits teaches you to live with mixed runs; four suits asks for both, over roughly twice as many moves. Our solver won 92.8%, 84.5% and 46.5% of deals with one, two and four suits.

Will a King-to-Ace run of mixed suits leave the table?

No. Only thirteen cards of one suit, King down to Ace, leave by themselves. A mixed run stays where it is until you take the column apart at each joint and rebuild it in suit.

What does Hint do when a game has gone wrong?

Hint asks the solver for a way to win from where you are. If it finds none, it says so and offers to take you back to the newest position from which it can, telling you how many moves to undo. With four suits that can be a long way back.