FreeCell is, in its own way, one of the best games ever designed. I am not aware of any other puzzle game which does such a good job of balancing three factors:

  1. The game should be based on random seeds.
  2. Almost all random seeds should be solvable.
  3. The search for that solution should be rewardingly deep.

To expand on that third point: puzzle games are approachable via exhaustive search, so the fun comes in coming up with rules that let you prune that search while leaving enough scope for judgment that the pruning doesn’t become routine. In a lot of my favorite puzzle games, that judgment can be expressed in the form of theorems about the game: I love Slitherlink because I was still discovering new theorems about it after playing it for months. Nurikabe doesn’t lend itself to as many theorems as Slitherlink does, but both games have a real pleasure in the interplay between pattern matching to apply rules locally, using a more global judgment to figure out where to apply your rules to solve the game most quickly, and figuring out where to try an exhaustive search if your rules don’t suggest any obvious next move.

FreeCell is different, though: there aren’t really any theorems per se other than figuring out what size stacks you can move in the presence of a given number of open spaces. Instead, it all comes down to pure positional judgment: what sorts of moves are going to increase the organization of the game board in a way that actually helps solve the game. And this lends itself to all sorts of wonderful tensions: do you try to increase organization by moving as many cards as possible to the foundations? Do you try to increase organization by increasing the number of sequences that you have built up on the board? The former is the obvious strategy; the latter, in general, turns out to be better, but if you go too far in that direction then you can end up with sequences that are too big to ever move them, leading in turn to a search for ways in which to cleverly break them up to move them by components. Or then there’s the desire to maintain options: normally, you want to maintain as many options as possible, which leads to interesting consequences like not always playing a card to the foundation if that will leave your stacks uneven, but you can’t leave options open forever: eventually you’ll have to do something that reduces your options while increasing your organization, and hope that it turns out well. (And FreeCell is a game where an undo button is necessary, I think.)

In fact, this latter tension is present at the very beginning: you’ll almost immediately have to play a card up to a cell, which decreases your options, and you’ll rarely have four cells available to you again; so there’s this delicate balancing act where having two cells filled is generally fine, you’ll sometimes go for long stretches with three filled (and you’ll constantly be temporarily spiking up to all four being filled), but you’re playing with fire with three filled, and sometimes even keeping two filled for a long period of time is too much. Also, the organized cards are covering unorganized cards; and frequently, there’s a card you need down in that unorganized section, buried deeply enough that your organization is hurting you rather than helping you. So sometimes you’ll explore path after path, find them not quite working, push one of them to the limit, and then finally things will cascade out into happiness. Sometimes even that doesn’t seem to work, so you’ll have to fall back to a less common technique; I’m always pleased and surprised when I think I’m stuck and realize that I can cut deeply by playing down two of the suits at the expense of the other two suits.

And this is all in the context of a game that is based on a random seed, almost always solvable, but where a decent proportion of the time the random seeds still give you (or at least give me) interesting games and where the straightforward games go through quickly enough and pleasantly enough that you don’t mind. It’s an amazing balancing act.


Despite how much I enjoy and respect FreeCell, though: I don’t always feel good about myself when I’m playing it. There’s been some number of attacks against “addictive” games this year; many of those attacks feel to me like they’re coming from a polemical basis that is routed in a political position that I disagree with, but still: if some games really are addictive, then that on the face of it sounds like a bad thing.

So: what contributes to making a game addictive? Intermittent rewards are a classic technique: our brains are wired to respond to intermittent rewards even more strongly than reliable rewards, and games frequently use that to keep us playing in hopes that things will turn out better next time. (Or, alternatively: if we just got lucky, then that reminds us why we want to keep on playing as well!) Sometimes this is transparently manipulative, e.g. in the case of random loot drops, and I try to stay away from games like that; but any game with a random component is going to raise the possibility that the play experience will turn out more to your liking if the next roll of the virtual dice is different, and FreeCell’s random seeds are no exception to that.

Short play length is another thing that I find contributes to games that I find addictive. (Where I’m using “addictive” in a naive sense of “I pick them up and/or continue to play them even when part of me feels that that’s not what I should be doing right then”; I don’t pretend to know anything about any more formal notion of addiction.) If a game only takes 30 seconds or two minutes or even ten minutes to play, then it’s easy to pick up when I have a bit of free time, easy to keep on playing even after some lull in external activity has passed. (Or: easy to spend way too much time playing on the toilet.) Again, FreeCell qualifies: I think it’s a great game, but that doesn’t mean that I want it repeatedly worming its way into cracks in my day, let alone enlarging those cracks.

(And then there’s a special variant of that last one: games with multiple goals at overlapping time scales, some of which are short. Games that go all in on that are, mercifully, rare, but there’s a reason why I haven’t played a Civilization game for years, despite how good I think that series is.)


Still: a lot of the reason why I go through periodic FreeCell binges (or binges of other puzzle games) is because I do find them rewarding. And, also: there’s nothing wrong with playing through a game just for the pleasure of doing the right moves, even if I’m not learning much from doing so. A lot of this summer, I wasn’t in shape to do much of anything that required thought; I was quite glad to have the collective works of the fine folks at Conceptis to give me something to do. But the flip side is: longer form works are good, too, a lot of the time I get more out of reading a book than out of playing yet another puzzle in the same puzzle game. So I should listen to what my brain is telling me, I should try to figure out what will nourish me most at any given time.

Post Revisions:

This post has not been revised since publication.