FreeCell with two free cells left: naming the exit

What this sitting was about

Fifteen minutes, sixty-odd moves, every card visible from the start. FreeCell does not hide information — it hides consequences. Unlike Klondike, where the next card might be one draw away, every fact FreeCell needs is already on the table from the opening deal. The uncertainty is not in what exists; it is in what a move commits you to. This sitting stalled once when all four free cells filled with “temporary” cards that had no next home, and the rest of the game became a slow argument with my own earlier decisions.

I have played FreeCell long enough to know the shape of a good game: it begins in the first ten moves. The deal gives you four free cells, four foundation piles waiting for aces, and eight columns of face-up cards. What you do in that first minute decides whether the middle of the game is a flow or a rescue. This sitting, the rescue arrived early and stayed.

What actually happened on the table

Free cells are hands, not shelves: name the exit before you park. The phrase sounds like a slogan, but it came from a specific failure. Around move twelve I had a red ten cluttering a column that I wanted for a black jack. There was a free cell open, so I parked the ten, intending to fetch the jack two moves later. Two moves later the jack was blocked by a nine that I had also parked, and the nine was blocked by an eight that had looked perfectly harmless when I moved it. Within four moves, three of my free cells were holding a tidy little staircase of cards that, individually, had all been “just for a second.” There was no cascade long enough to clear them profitably, and no foundation yet open at the right rank. The helpers had become residents.

The fix was not clever. It was slow. I had to refuse a legal cascade — the game offered me a clean run of six cards from one column to another, built by colour and descending rank, and taking it would have felt like progress. But the cascade was built on the same illusion: it used the free cells as a shelf. Moving six cards through three free cells and one empty column is legal only if the empty column stays empty, and making that column empty again required parking two more cards into the very cells I was draining. I counted the exchange three times. Each time it ended with the same number of free cells, plus a rearranged mess. Legality was not usefulness; the move was a bicycle wheel — lots of motion, no new ground.

So I left the staircase alone and went looking for aces. Getting the aces and twos home unlocked cascade bottoms nothing else could reach. The ace of spades sat under a hidden column, reachable only after three specific cards moved in a specific order. The first two times I tried, I ran out of columns. The third time I stopped trying to clear the path and instead built a path: I moved two cards into a foundation that was technically premature (leaving a red queen stranded), and that single stranger-looking sacrifice opened the column the ace was riding on. It felt slow — four moves of housekeeping for one foundation card — and it was the real progress of the whole game. Everything after it flowed the way the opening should have flowed.

There was a second judgement worth writing down. Around move forty, with two free cells left and four columns partly built, I faced a choice between two long cascades. Both were legal. One kept a dark-suited run intact; the other broke it but freed a column that had been locked for ten moves. The tempting move was the first — it preserved the tidy chain I had spent effort building. The correct move was the second, because a locked column is a debt, and the tidy chain was decorative. I made the correct move and the rest of the game stopped being stressful. Locked columns are the actual enemy in FreeCell; everything else is furniture.

Let me also be honest about the moves I did not write down. There were at least three places where I moved a card only because it was movable — the equivalent of tapping buttons on a machine. Each of those moves cost me something later, usually a free cell or a column start. FreeCell punishes idleness even more than it punishes greed, because a legal move you make for its own sake is a commitment you did not intend.

After the game

The win, when it came, was unremarkable — the last few cards slid home one after another the way they always do once the structure is right. I mention this because the memory that stayed with me was not the finish but the moment, around move forty, when I finally understood why the game had felt blocked for so long. It was not the deal. The deal was generous; almost every card had a legal destination from the opening. What blocked the game was me: I had been treating the free cells as a parking lot and the empty column as a bonus, when the empty column is actually the engine of the whole endgame. Every cascade I had refused earlier became trivial once I stopped guarding it. That is the uncomfortable part of FreeCell — the game rarely beats you with a bad deal; it beats you with your own tidy arrangements, the ones that looked like progress at the time and were actually locks.

Judgments to take with you

1. Legality ≠ usefulness on cascade transfers. Count the exchange to its end state before you start, and if the end state is the same number of helpers, the move is decoration.
2. Protect one empty column early; it multiplies every helper. An empty column is a multiplier, not a storage space.
3. If a long run “won’t move,” free a cell first — do not force hope. The cascade that glitters is often the one that rebuilds the same jail in a new order.
4. A premature foundation move can be the cheapest way to open a buried ace. Ugly is not the same as wrong.
5. Locked columns are debts. Clear them before you polish anything.

One line to keep for next time

Before you park, say the next square out loud. If you cannot name the card that is coming to take this one's place, the parking move is a promise you are not ready to keep.