FreeCell World Records

Speed is the most visible FreeCell record category. Players compete to clear deals as fast as possible, and reported times have dropped dramatically as interfaces have improved. But comparing times across different apps is tricky because auto-move behavior, drag mechanics, and deal difficulty all vary.

As of 2026, the most credible speed claims come from platforms with built-in leaderboards that track times server-side. Self-reported times without replay verification are interesting but harder to confirm.

Win streaks are FreeCell's endurance record. Because nearly every deal is solvable, a skilled and patient player can theoretically win indefinitely. The question is not whether it is possible but how long someone can maintain perfect focus.

The most impressive streaks are those played under strict rules: no restarts, no deal selection, limited or no undo. Under those conditions, community records suggest that streaks in the hundreds are a significant achievement, and streaks above a thousand are exceptional. Try building your own streak in Streak Mode.

No discussion of FreeCell records is complete without Deal #11982. It is the most famous individual FreeCell deal in history — not because someone won it, but because nobody can.

The Deal #11982 story matters because it proved that FreeCell is not trivially solvable. The game sits in a fascinating sweet spot: almost every deal has a solution, but finding it requires genuine skill and planning. That balance is what keeps players coming back decades later. Read the full story on our dedicated Game #11982 page, or explore other notable deals in our famous FreeCell deals guide.

Some of the most impressive FreeCell achievements are not individual records but community-wide projects. Groups of players have systematically worked through massive deal sets to determine which deals are solvable and which are not.

These projects are a testament to the FreeCell community's dedication. They transformed a casual Windows game into one of the most thoroughly analyzed puzzles in recreational mathematics.

You do not need to chase world records to enjoy the competitive side of FreeCell. Personal bests are just as satisfying. Here are the records worth tracking in your own play.

Speedrun play has developed into several distinct categories, each with its own conventions. Understanding the categories helps you read claims critically: a forty-second time on a cherry-picked deal is a very different achievement from a two-minute time on an arbitrary deal.

Specific-deal speedruns are the most visible category. A player picks a known deal — often deal #1 or another low-numbered favourable board — and optimises the completion time. Records in this category often dip below thirty seconds because the solution is memorised and execution is mostly about input speed. Timing typically starts on the first move rather than when the deal loads.

Random-deal speedruns are the harder category. Here the deal is drawn at random at the start of the attempt, and the clock runs until the game is won (or lost). Times are longer because the player cannot rely on memorised lines, and the metric depends on consistency rather than peak performance. A two-minute random-deal average is respectable; a ninety-second average approaches tournament class.

Hard-deal speedruns take a known difficult deal (say, #11982's close relatives or the historically hard deal list) and measure how quickly it can be solved. These are rare because the deals genuinely require thought, which is hard to combine with speed. The people who post times in this category tend to be the strongest FreeCell players in the world.

Marathon speedruns measure time across a set of deals (often the first hundred from the Microsoft set, or a curated mix). This category tests endurance as much as tactical speed and is where sustained per-deal averages matter more than peak performance.

A fifth category worth noting is blindfold-style play, where the player announces their moves verbally or by typing and cannot use the interface to experiment. This category is vanishingly rare but produces extraordinary-looking claims when it happens, because it requires holding the entire game state in memory between moves. Treat blindfold claims with particular skepticism unless they come with recorded video.

A parallel achievement category measures efficiency rather than speed: how few moves does it take to clear a deal? FreeCell's move counter increments on every single-card action, so a tight solution can finish with very few registered moves. Solvers typically find optimal or near-optimal move counts as part of their normal output, so the theoretical minimum for any given deal is known even when the fastest human attempt is not.

For typical deals, the minimum move count sits between seventy and ninety registered moves. Gentle deals can be cleared in the sixties. Harder deals sometimes require well over a hundred moves even under optimal play. Human players rarely match the solver minimum, because finding the truly optimal move order requires exhaustive search that humans cannot perform in a reasonable time. Still, hitting within five or ten moves of the optimum is a meaningful achievement and a common personal goal.

One subtlety: what counts as a “move” differs across implementations. Some count every single card action, including foundation auto-moves. Others count supermoves as a single move regardless of how many cards were shuffled through cells to execute them. Claims about fewest-moves records should specify the counting convention, or they are not comparable.

There is also a less-celebrated category: the most-moves record, which tracks the longest legal completion of a given deal. It sounds perverse but is a genuine puzzle: how many legal moves can you make before the game either wins itself or stalls? Researchers studying game-tree depth have used this metric to characterise FreeCell difficulty, and it occasionally resurfaces in community discussions as a curiosity. The deals with the longest legal move sequences tend to be the hardest ones.

For most players, the practical takeaway is that move count is a useful personal-improvement metric even if it never becomes a competitive record. A move count that drops over time is a reliable signal that your play is getting cleaner, whether or not you are competing for any external recognition.

FreeCell tournaments are informal but established. Several formats have emerged over the years, each with different emphases.

FreeCell Pro tournaments are organised around a specific software implementation that has been popular in competitive circles for years. They typically use a curated set of hard deals, give players long time limits, and track cumulative solve rates. Winners are players who complete the most deals in the set within the allotted time without using restarts.

Online leaderboard tournaments run on specific platforms and track times across shared daily deals or fixed deal sets. Players compete across time zones on the same boards, producing clean comparisons. These tournaments tend to favour speed over depth because fast completion of easy deals is usually the winning strategy.

Community challenges are informal but persistent. Someone on a forum proposes a challenge (“beat deal #11982 variants, or solve the entire 1,000-deal block starting at #10,000”) and players self-report. These rarely produce “world records” in any formal sense but they generate a steady stream of collective play and record-keeping.

None of these formats is sanctioned by a central body. Each produces real competition and real records, but the records are contextual. A FreeCell Pro champion and an online-leaderboard top-timer may both be world-class players, and comparing them directly would require running them on the same deal set under the same rules. That kind of head-to-head rarely happens.

A final format worth knowing: variant tournaments. Some competitive circles run tournaments in harder FreeCell variants — two-cell FreeCell, Baker's Game, or Seahaven Towers — precisely because those variants punish the speed-focused styles that dominate standard FreeCell tournaments. The winners of variant tournaments tend to be players with exceptional planning discipline, because the variants do not let aggressive play compensate for poor reading of the board.

Because FreeCell has no governing body, the “best player” list is an informal one assembled from forum reputation, leaderboard rankings, and long-running community memory. A few names recur in historical discussions.

Michael Keller, who coordinated much of the Internet FreeCell Project through the 1990s, is not primarily known for competitive play but for his exhaustive research on the Microsoft 32,000 deal set. His work confirmed that every deal except #11982 is solvable, and he documented many hard-deal solutions that remain reference material today.

Dave Ring, founder of the Internet FreeCell Project, organised the collective effort that verified the Microsoft set and helped establish the community norms around deal identification and solution sharing that tournaments still follow.

Modern competitive players are less publicly identified because online leaderboards typically use screen names. What we can observe is their strategy: top players consistently report long planning phases on unfamiliar deals, strict cell discipline, and an explicit habit of counting movable-run capacity before every multi-card move. They do not rely on intuition. They count.

A few common habits recur across interviews and forum posts from top players. They plan first, play later — opening moves often come after two or three minutes of motionless thought. They treat cells as exhaustible resources and count before filling. They study solver output on deals they lost, looking for the moves their own search did not consider. And they calibrate aggression to difficulty: on gentle deals they play fast, on hard deals they play slow, and they do not confuse the two.

For players chasing their own first records, the lesson is clear: the path to stronger play runs through counting and planning, not through faster clicks or clever tricks. Work on the mastery guide first; the records will follow.

Verification is the hard problem of FreeCell records. Without a central authority, every claim has to be verified some other way, and the community has developed a few informal standards.

Server-side timing is the strongest form of verification. A platform that records your completion time on its own servers, before any client-side claim is made, produces a tamper-resistant record. Most modern leaderboards work this way.

Video recording is the next best thing. A player who records their session and publishes it provides evidence that the claimed time and move count match a real play sequence. Video verification has become standard for high-stakes speedrun claims in most games, and FreeCell is no exception.

Replay files are the traditional verification method. Many FreeCell implementations support move-by-move export, which allows other players to replay and audit a claimed solve. Replays are especially useful for verifying fewest-moves claims, because they show the exact move sequence.

Self-reported claims without any of the above are treated skeptically but not dismissed. They are useful as anecdote and as inspiration but not as evidence. The Research Desk treats any unverified claim as “community-reported” and hedges language accordingly throughout this page.

We also note that different platforms run FreeCell differently — auto-move rules, supermove handling, drag-and-drop speed, undo permissiveness — and these details can shift times and move counts in ways that make cross-platform comparison unreliable. A time set on one site is not trivially comparable to a time set on another.

The internet is full of FreeCell record claims, and most of them are either unverified, out of context, or comparing apples to oranges. A healthy skepticism makes you a better reader of the record-keeping landscape. Here are the questions the Research Desk asks before taking any FreeCell record claim at face value.

Which deal? A thirty-second completion on deal #1 is routine. A thirty-second completion on a random hard deal is extraordinary. Claims that do not specify the deal are not claims, they are anecdotes.

Which platform? Times on platforms with aggressive auto-move behaviour are systematically faster than times on stricter platforms. The claim loses meaning without the platform context.

Which rules? Did the player use undo? Restarts? Hints? A win-streak of ten thousand games means something very different if the player restarts any deal they fear losing.

How was it verified? Was there server-side timing, video, or replay? Or is this a forum post from eight years ago?

We are not suggesting that every claim is false — many FreeCell records are real achievements by dedicated players. We are suggesting that the record-keeping ecosystem lacks the verification infrastructure that sports and competitive gaming take for granted, and that readers should calibrate their credulity accordingly. When we make claims on this site, we hedge them. When other sites do not, you should.

The broader point is that FreeCell is an unusually democratic competitive scene. There is no membership fee, no sanctioning body, and no prize circuit. Anyone with an internet connection and patience can post a time, and anyone with the same patience can verify or dispute it. That openness produces a messy record landscape and also a genuinely inclusive community. Take the records with skepticism, and take the community as it is: decades-deep, obsessive in the best way, and still arguing about whether deal #617 is really that hard.

If you want to participate, the path is simple. Play many deals, track your times and move counts honestly, share your results in community spaces, and accept correction when your numbers do not hold up. Records worth anything are the ones verified by a community willing to look at them carefully, and joining that community is as easy as posting your first honest stat, along with enough context for anyone else to check your work. That is the whole bar to entry, and it is a low one by design.