History of FreeCell

In 1978, Paul Alfille was a medical student at the University of Illinois at Urbana-Champaign. Like many students of the era, he spent time on the PLATO system — a groundbreaking educational computer network that predated the modern internet by more than a decade. PLATO was built for education, but it also happened to host some of the world's earliest computer games, from multiplayer dungeon crawlers to flight simulators. In that environment of experimentation and play, Alfille set out to solve a problem that had bothered card game enthusiasts for generations.

Most solitaire games deal some cards face-down, meaning your fate partially depends on what you can't see. You might play perfectly and still lose because a critical card was buried in an inaccessible position. Alfille wanted to eliminate that hidden information entirely. He was familiar with Baker's Game, an older solitaire variant that dealt all cards face-up, but found it too restrictive — its requirement to build by suit made too many deals unsolvable.

Alfille's solution was deceptively simple but profoundly effective: keep Baker's Game's open layout with all 52 cards visible, but change one rule. Instead of building tableau columns by suit, players could build by alternating color — red on black, black on red. This single modification dramatically increased the number of solvable deals, taking the solvability rate from roughly 75% (Baker's Game) to over 99.999% (FreeCell).

The game retained four temporary storage spaces — the free cells — where any single card could be parked while you rearranged the tableau. The result was a game where nearly every deal has a solution, and finding that solution requires genuine strategic thinking. No blaming bad luck. No praying for a helpful card to turn over. Just you, 52 visible cards, and your ability to plan ahead.

Alfille programmed FreeCell in the TUTOR language, which was PLATO's native programming environment. TUTOR was remarkably advanced for its time, supporting graphics, touch screens, and networked communication. This allowed Alfille to create a polished graphical card game that players could interact with intuitively. The game spread across PLATO's network of terminals at universities and military installations, earning a small but passionate following among students, professors, and system administrators who spent their evenings trying to crack difficult deals.

For the next decade, FreeCell remained a niche game known primarily to people who had access to PLATO terminals. Several programmers encountered the game on PLATO and wrote their own versions for personal computers — early DOS implementations, Macintosh ports, and versions for various Unix systems. But without a distribution channel to match PLATO's reach, FreeCell stayed an insider's game, waiting for the moment that would bring it to a global audience.

To understand FreeCell, you need to understand the game it evolved from. Baker's Game, created by C.L. Baker in the 1960s, was one of the first solitaire variants to deal all cards face-up. Baker, a mathematician, wanted a solitaire game that tested logical reasoning rather than luck. His design was clever: deal all 52 cards into eight tableau columns, provide four free cells for temporary storage, and four foundation piles where cards are built up by suit from Ace to King.

The critical difference between Baker's Game and FreeCell lies in one rule: how you build sequences on the tableau. In Baker's Game, tableau columns must be built down by suit — you can only place the 7 of hearts on the 8 of hearts, the 5 of spades on the 6 of spades, and so on. This restriction sounds minor, but its consequences are enormous. Building by suit means you have far fewer legal moves at any given point, which means far more deals reach dead ends no matter how well you play.

Research has estimated that only about 75% of randomly dealt Baker's Game layouts are solvable — compared to FreeCell's 99.999%. That 25% gap represents the difference between a game that sometimes feels unfair and one that almost never is. When Paul Alfille changed the building rule to alternating colors (red on black, black on red), he preserved Baker's Game's intellectual depth while making the game dramatically more accessible.

Baker's Game still exists and has its own devoted following among solitaire purists who appreciate its greater difficulty. Some solitaire variants offer Baker's Game as an alternative mode alongside FreeCell, and it remains a favorite among players who have mastered FreeCell and want a stiffer challenge. But it was Alfille's modification — that single rule change about alternating colors — that transformed an academic curiosity into the world's most popular strategic card game.

FreeCell might have remained a footnote in computing history if not for one person: Jim Horne. A programmer at Microsoft, Horne encountered FreeCell in the late 1980s and immediately recognized something special about the game. Here was a solitaire variant that rewarded thinking over luck, where skilled players could expect to win nearly every time. He wrote a Windows version and pitched it for inclusion in Microsoft's growing collection of bundled games.

Horne made a design decision that would prove crucial to FreeCell's legacy: he assigned every deal a number. Using a specific random number generator seeded with integers from 1 to 32,000, each number always produced the same card layout. Game #1 was always the same game. Game #17000 was always the same game. This meant that players around the world could compare notes on specific deals, debate strategies for particular numbers, and challenge each other to solve the trickiest layouts.

Windows 3.1 and the Entertainment Pack (1991–1994)

FreeCell first appeared as part of the Windows Entertainment Pack and the Win32s subsystem for Windows 3.1. At this stage it was an optional download, not a default inclusion. The interface was simple — 256-color graphics, basic card designs, and a no-frills menu bar. But the gameplay was already addictive. Early adopters in corporate offices and university computer labs began keeping handwritten lists of which deal numbers they had conquered.

Windows 95: The Tipping Point (1995)

When Microsoft bundled FreeCell with Windows 95, the game went from a curiosity to a phenomenon overnight. Windows 95 sold over 40 million copies in its first year, and every single one came with FreeCell pre-installed. For millions of people, it was the first time they encountered a solitaire game where skill actually mattered. The timing was perfect: the mid-1990s saw an explosion of home computer ownership, and FreeCell was there on every desktop, waiting to be discovered.

Office workers found it during lunch breaks. Students stumbled across it between assignments. IT administrators played it during late-night server maintenance. FreeCell became a universal shared experience — the game everyone had played but nobody had bought. It generated zero revenue for Microsoft directly, but it helped make Windows feel friendly and personal.

Windows XP Through Windows 7: The Golden Era (2001–2012)

FreeCell's inclusion continued through every major Windows release. Windows XP (2001) brought updated graphics and a cleaner interface. Windows Vista (2007) added a redesigned card set and improved animations. Windows 7 (2009) polished the experience further with better visual effects and smoother gameplay. Through it all, the core game remained identical to what Alfille had designed in 1978 — the same rules, the same four free cells, the same 52 face-up cards.

By the Windows XP era, FreeCell had become so entrenched in workplace culture that IT departments reportedly received requests to remove it from company machines to reduce distractions. Some companies did exactly that. Others recognized that a few minutes of FreeCell between tasks could serve as a genuine mental reset — a low-stakes puzzle that exercises working memory and planning skills without the emotional investment of more complex games.

Windows 8 and the Microsoft Solitaire Collection (2012–Present)

With Windows 8 in 2012, Microsoft made a controversial decision: the standalone FreeCell app was gone. In its place was the Microsoft Solitaire Collection, a unified app that bundled FreeCell with Klondike, Spider, Pyramid, and TriPeaks. The new app looked modern and offered daily challenges, but it also introduced something the original never had: advertisements. Players who had enjoyed decades of ad-free FreeCell now faced video ads between games unless they paid for a premium subscription.

The backlash was significant. Long-time players mourned the loss of their statistics, their familiar interface, and the simplicity of a game that just launched and played without asking for anything. Windows 10 continued with the Solitaire Collection model, and while the app itself is well-made, the ad-supported model drove many dedicated FreeCell players to seek alternatives — including the growing number of web-based and mobile versions that recaptured the clean, distraction-free experience of the original.

Among the original 32,000 numbered deals in Microsoft FreeCell, game #11982 holds a unique distinction: it is the only deal that has been mathematically proven to have no solution. Every other deal — all 31,999 of them — can be solved with the right sequence of moves. This one cannot.

The story of how we know this is itself remarkable. In the mid-1990s, a volunteer effort called the Internet FreeCell Project set out to solve every single one of those 32,000 deals. Organized by Dave Ring, the project was one of the internet's earliest crowdsourcing endeavors — years before that term existed. Volunteers would claim game numbers, attempt to solve them by hand, and report their results to Ring, who maintained a central registry tracking the status of each deal.

The project attracted thousands of participants from around the world. Some were casual players who claimed a handful of deals. Others were obsessives who worked through hundreds, documenting move sequences and sharing strategies for the most difficult layouts. By the late 1990s, every deal had been solved except #11982 and a handful of others that would eventually fall to more advanced solving techniques.

Computer scientists later confirmed what thousands of human players suspected: deal #11982 is genuinely impossible. Exhaustive search algorithms — programs that systematically explore every possible sequence of legal moves — proved that every path leads to a dead end. No matter how brilliant your strategy, this particular arrangement of cards simply cannot be untangled. The proof is absolute: there is no combination of moves, no matter how long or convoluted, that leads to all four foundation piles being completed.

Interestingly, eight other deals were long thought to be unsolvable: #146, #455, #495, #512, #530, #1941, #6182, and #8591. As solver algorithms improved over the years, solutions were found for all eight. Some required over 100 moves and extremely precise play — sequences so unintuitive that no human player had discovered them — but they were all technically winnable, making #11982 the sole holdout.

Want to try your hand at the impossible? You can play Game #11982 here and see for yourself why it defeated every human and computer solver. Or try some of the notoriously difficult (but solvable) deals like #169 or #178 to test your skills.

FreeCell occupies a fascinating position in the intersection of recreational mathematics and computer science. Unlike most card games, where probability and chance dominate the analysis, FreeCell's open-information design means it can be studied as a pure combinatorial puzzle. And the numbers that emerge from that study are remarkable.

How Many FreeCell Deals Exist?

A standard deck of 52 cards can be arranged in 52 factorial (52!) different orders — approximately 8.07 × 10⁶⁷. That's an 8 followed by 67 zeros. To put this in perspective: there are roughly 10⁸⁰ atoms in the observable universe. While that's a larger number, the number of possible card arrangements is still mind-bogglingly vast. If every person who has ever lived (roughly 100 billion) played a unique FreeCell deal every second since the Big Bang (13.8 billion years ago), they would have collectively played about 4.35 × 10²⁸ deals — barely a scratch on the surface of all possible games.

Microsoft's original 32,000 deals, which formed the basis for decades of FreeCell play and research, represent an infinitesimally small fraction of all possible games. Later versions expanded to 1,000,000 numbered deals, but even a million is effectively zero when measured against 52 factorial.

The 99.999% Solvability Rate

The question that fascinates both mathematicians and players is: what fraction of all possible FreeCell deals are solvable? The answer, based on extensive computational research, is approximately 99.999%. This figure comes from studies that generated millions of random deals and applied sophisticated solving algorithms to each one. The unsolvable deals that do exist tend to share certain structural characteristics — configurations where key cards are trapped in positions that no sequence of moves can untangle.

The high solvability rate is a direct consequence of Alfille's design decision to allow alternating-color building. The four free cells provide just enough flexibility to navigate around obstacles, while the alternating-color rule ensures that useful moves are almost always available. It's a beautifully balanced system: challenging enough that winning feels like an achievement, but fair enough that losing feels like a learning opportunity rather than bad luck.

Computational Complexity

From a computer science perspective, solving a FreeCell deal is an NP-complete problem — meaning that while solutions can be verified quickly, finding them efficiently is theoretically difficult. In practice, however, modern FreeCell solvers use heuristics and pruning techniques that solve most deals in milliseconds. The best solvers combine depth-first search with pattern recognition, evaluating positions based on factors like column homogeneity, card accessibility, and foundation progress.

These solver algorithms have practical applications beyond FreeCell. The techniques developed to search FreeCell's game tree — efficiently pruning dead-end branches, managing large state spaces, and balancing exploration with exploitation — have informed research in automated planning, constraint satisfaction, and even protein folding. FreeCell, in its quiet way, has contributed to the broader field of artificial intelligence.

FreeCell's impact extends far beyond the game itself. As one of the three card games bundled with Windows (alongside Klondike Solitaire and Minesweeper), it became a shared cultural experience for an entire generation of computer users. In the 1990s and 2000s, FreeCell was often the first game people played on a new computer — the digital equivalent of breaking in a new deck of cards.

In the workplace, FreeCell became synonymous with office downtime. IT departments joked about removing it from company machines, and more than a few managers walked past employee screens showing suspiciously quick Alt-Tab reflexes. A 2003 survey estimated that American workers collectively spent billions of hours per year playing Windows card games, with FreeCell and Klondike accounting for the vast majority. Rather than viewing this as pure waste, some researchers argued that brief game breaks improved focus and productivity — a mental palate cleanser between cognitively demanding tasks.

FreeCell also served an unexpected educational role. Microsoft originally included card games with Windows specifically to help users learn mouse skills — particularly drag-and-drop operations that were unfamiliar to people transitioning from keyboard-only interfaces. For millions of people in the early 1990s, moving cards around in FreeCell was their first experience using a mouse to interact with objects on a screen. Solitaire taught clicking and dragging; FreeCell added strategic thinking on top.

The Internet FreeCell Project (1994–2000) was a pioneering example of crowdsourced problem-solving that predated the term "crowdsourcing" by a decade. Years before Wikipedia, before citizen science platforms like Folding@home, thousands of volunteers coordinated online to systematically solve all 32,000 Microsoft FreeCell deals. The project demonstrated that large-scale collaborative efforts could accomplish what no individual could do alone, and it did so purely for the satisfaction of completing the challenge — no prizes, no funding, just collective curiosity.

Today, FreeCell remains one of the most popular solitaire card games in the world. Its appeal hasn't faded because its core design is genuinely excellent: perfect information, deep strategy, and the satisfying knowledge that nearly every deal is solvable if you play well enough. Whether you first discovered it on a PLATO terminal, a Windows 95 desktop, or a modern browser, the game is the same — and that timelessness is its greatest achievement.

While most people play FreeCell casually, a dedicated competitive community has emerged around the game, pushing the boundaries of how fast and how consistently deals can be solved. FreeCell speedrunning — completing deals as quickly as possible — has become a niche but passionate pursuit with its own leaderboards, techniques, and culture.

Top speedrunners can solve many standard deals in under 30 seconds, with exceptional runs on favorable layouts dipping below 15 seconds. These players rely on pattern recognition developed over thousands of hours of play, instantly identifying common card configurations and the move sequences that resolve them. They process information about all 52 cards within the first few seconds of seeing a deal, mentally mapping out the critical strategic decisions before making their first move.

Beyond raw speed, some competitive players focus on consistency — maintaining the highest possible win rate across hundreds or thousands of consecutive random deals. A skilled player might maintain a win rate above 99% over a large sample, losing only to the rare genuinely difficult layouts. The streak challenge — how many consecutive games you can win — adds another dimension, where a single loss resets your count and the psychological pressure of a long streak becomes part of the challenge.

The competitive community has also driven the development of analysis tools. Players review their completed games to identify suboptimal moves, similar to how chess players analyze their games with engines. Some platforms offer post-game analysis that compares your move sequence to the optimal solution, highlighting where you made unnecessary moves or missed more efficient paths. These tools have helped raise the overall skill level of the community and given new players a structured way to improve their gameplay.

Competitive FreeCell may never rival chess or esports in scale, but its community embodies something appealing: the idea that even a simple card game, played with enough dedication and analytical rigor, can become a deep skill-based pursuit. The gap between a casual player and an expert is enormous, and bridging that gap requires genuine study — learning techniques, developing intuition, and accepting that every loss is a lesson.

The launch of Apple's iPhone in 2007 and the subsequent explosion of smartphone apps created an entirely new chapter in FreeCell's history. Touch-screen interfaces proved naturally suited to card games — tapping a card to move it or dragging it across the screen feels more intuitive and satisfying than clicking with a mouse. FreeCell apps appeared in app stores almost immediately and quickly climbed the charts, introducing the game to a generation that may never have used Windows desktop games.

Mobile FreeCell brought its own design challenges. Smartphone screens are much smaller than desktop monitors, requiring creative solutions for displaying all 52 cards in a readable layout. The best mobile implementations use carefully sized card overlaps, pinch-to-zoom functionality, and landscape mode to give players enough visual clarity to make strategic decisions. Portrait mode on a phone is a tight squeeze for eight columns of cards, but clever UI design has made it work surprisingly well.

The rise of modern web technologies — HTML5, CSS3, and JavaScript frameworks — enabled another shift: browser-based FreeCell that requires no download at all. Sites like PlayFreeCellOnline.com offer the full FreeCell experience directly in a web browser, complete with features like unlimited undo, auto-complete for solved positions, statistics tracking, and customizable settings. These web versions work across devices — desktop, tablet, and phone — without requiring separate apps for each platform.

Web-based FreeCell has also addressed one of the key frustrations of the Microsoft Solitaire Collection era: advertisements. Many browser-based versions offer clean, ad-light experiences that prioritize gameplay over monetization. For players who remember the simplicity of the original Windows FreeCell — launch the game, play the game, close the game — this return to basics is a welcome change.

What's remarkable about FreeCell's journey across platforms is how little the game itself has changed. The rules Paul Alfille designed in 1978 are identical to those used in every modern implementation. Four free cells. Eight tableau columns. Four foundation piles. All cards face-up. Build down by alternating color. The technology surrounding the game has transformed — from PLATO terminals to Windows desktops to smartphones to web browsers — but the game at the center of it all remains perfectly, stubbornly unchanged. That is the ultimate proof of a great game design: it needs no updating.