Baker's Solitaire: why a promising deal can still be impossible

Baker's Solitaire looks like FreeCell's twin. The same eight face-up columns, the same four free cells, the same four foundations. You can see every card from the start, so it is tempting to read the deal and judge it at a glance. On klondike.nz, that would be a mistake.

We dealt 10,000 games of Baker's under this site's rules. 49.7% of them had an Ace ready to play on the first move — the deals look just as friendly as FreeCell's. But our solver, which searched every deal to the end, found that only 31.8% could be won. The other 68.2% were proven impossible, no matter how well you play.

The reason is a single rule. Here, only a King may go into an empty column. Change that one rule to "any card", and the same solver wins 76.0% of deals. What the opening layout can't show you is how many of your empty columns will be usable.

A Baker's deal, by the numbers

How Baker's Solitaire is played here

What you see on the table

A standard pack dealt face up into eight columns (four of seven cards, four of six), four free cells for single cards, and four foundations.

The aim

Build each suit on its foundation from Ace up to King.

The rules that make it Baker's

• Columns build down in the same suit: the 7 of hearts goes on the 8 of hearts, nowhere else.
• Any exposed card can go to a free cell.
• Only a King (or a run headed by a King) can fill an empty column.
• Multi-card moves use free cells only; empty columns don't add to how many cards you can move at once.

Classic Baker's Game, the version studied by researchers, lets any card fill an empty column. The Kings-only rule used here makes it a much tougher game.

Reading a Baker's deal

Find the Kings first, not the Aces

With about half of all deals offering an Ace straight away, Aces are not the scarce resource. Kings are. Every empty column you create can only be used by a King, so note where the four Kings sit and how deep they are buried.

Look for same-suit chains

Because building is in suit, a column like 9♠ 8♠ 7♠ is gold, while 9♠ 8♥ is a dead end until it is broken up. Count the natural same-suit pairs in the deal: they tell you how much work is already done.

Watch for low cards under same-suit high cards

If the 4♦ sits beneath the 10♦, the 10 can only move to a free cell, a foundation, or onto the J♦. Deals with several of these traps are the ones most likely to be among the 68.2% that can't be won.

Don't be fooled by a free Ace

A playable Ace on move one is pleasant, but it tells you almost nothing about the outcome: roughly half of all deals have one, while fewer than a third can be won.

What the numbers mean for you

If you lose at Baker's here, there is a better than two-in-three chance the deal was never winnable. That is very different from FreeCell, where our solver found no impossible deal at all in 10,000. Use Undo to test a line, but don't spend all evening on a deal that may have no solution; when a position has no King available for the empty column you need, it is often time to deal again.

When a Baker's deal can be won, the first solutions our solver found averaged about 39 moves, against about 62 for FreeCell.

A first-minute checklist for Baker's

With every card face up, Baker's can be sized up before the first move. Because so many deals on this site can't be won, this checklist is partly about spotting a hopeless deal early and partly about not throwing away a winnable one.

1. Locate all four Kings. Note how many cards sit on each. A King near the top of a column is ready to move into the first empty column you make.
2. Find the low cards trapped in suit. Look for Aces, 2s and 3s buried under higher cards of the same suit. Each one needs a free cell, a foundation move or its exact suited partner to escape.
3. Count your same-suit pairs. Adjacent suited cards in the right order are finished work; mixed columns are work still to do.
4. Choose one column to empty — for a specific King. An empty column that no King can reach is wasted, because nothing else is allowed in.
5. Protect your free cells. Multi-card moves here rely on free cells alone, so each filled cell reduces how much you can shift at once.
6. Decide when to stop. If no King can reach the empty column you need and every cell is full, you may well be in one of the 68.2% of deals that can't be won.

That last point is the unusual one. In most solitaire games we would say "keep going". In Baker's on this site, the maths says a quick, honest look and a fresh deal is sometimes the smarter move.

Play Baker's Solitaire on klondike.nz

Free in any modern browser, on computer, tablet or phone, with no download, no account and no email. Big, clear cards with zoom, Hint, unlimited Undo, Restart and Auto Play; drag and drop or tap. Wins, time and moves are tracked for the current game, and you can mute the sound, play fullscreen or pick a table colour. If you want the gentler cousin, try FreeCell.

Where these numbers come from

The klondike.nz team ran this study in September 2026.

1. Rules from our code. In-suit building, four cells, and Kings only in empty columns, read directly from the game on this page.
2. A program version of the game. Built and cross-checked: for deals 1 to 5 an independent brute-force search reached the same verdict as the main solver.
3. Numbered deals. Seeds 1 to 10,000 for the site's rules and 1 to 2,000 for the classic rule.
4. Looking at each deal. We checked for an immediately playable Ace.
5. Solving. A search with a 200,000-position limit, which was enough to reach a firm answer on every one of the 10,000 deals.
6. Comparison. Under the classic rule our 76.0% matches the published 75.053%, which tells us the solver is sound.
7. Error bars. 30.9–32.8% is the 95% range around our 31.8%.
8. Limits. Our deal numbers are our own; the solutions found are not necessarily the shortest.

Under the bonnet

The code is Python 3.11, standard library only. Baker's runs in freecell.py with the in-suit and Kings-only rules switched on. A shared runner hands deals to 19 worker processes with multiprocessing; the 10,000-deal run took about 16 minutes and the classic-rule run about 3. The solver is a best-first search: it always explores the position with the fewest cards left to place, remembers positions already seen, and only makes foundation moves that can never hurt. Because it finished every deal, each "impossible" verdict is a proof, not a guess. The 95% ranges are Wilson score intervals. Claude, an AI assistant made by Anthropic, helped us write and review the code and draft this text; the AI did not supply any number — every figure is the output of the code, checked against the published study.

Questions players ask

Why do I lose so often at Baker's?

Because most deals here can't be won: 68.2% were proven impossible under this site's rules.

Is Baker's Solitaire the same as Baker's Game?

Almost. Baker's Game in the research literature lets any card fill an empty column; on this site only Kings can. That rule is why our winnable share is 31.8% rather than about 75%.

Does an Ace on the first move mean a good deal?

Not really. About half of deals have one, but fewer than a third can be won.

How long is a winning game?

The solver's first solutions averaged about 39 moves.

Which game should I try if Baker's wears me down?

For something with more wins, try Golf Solitaire, or see everything on the games page.

By Admin, klondike.nz. Study date: 18 September 2026.