What is a Puzzle?

This weekend, thanks to a small editorial mix-up, I wound up playing the recent android port of Framed, the indie noir comic puzzler reviewed here by my colleague Nick Walker. Things wound up working out for the best, though, as the game itself turned out to be less interesting to me than the questions it raised in my mind:

What is a puzzle, exactly? How do you define it? What is it not? What constitutes a “puzzle game” — and, crucially, what makes a puzzle game “good”?

They’re questions that have been up for debate almost as long as the genre has existed, and the range of answers seems only to expand with time. To be sure, in the long Western pilgrimage toward the puzzle-game Promised Land, entire genres (like the point-and-clicks of yesteryear) have been born, flourished, and fallen dead by the roadside.

As designer Jon Blow puts it, “If you look at other genres of games, there’s an idea of what the flow is. A racing game has a flow to it. There’s this thing that you expect the player to be doing. A first-person shooter has something like that that’s more complex. Puzzle adventure games never figured out what that thing is.”

It’s interesting to me that Blow uses the word “flow” there, which normally we find in action or rhythm games. But that’s the same word used in Jessie Schell’s The Art of Game Design—with applications for puzzle games shared and explored here by indie dev Toni Sokoban—and similar ideas even pop up in the dev commentary for AAA titles like Portal 2.

But while these designers may share terminology with Blow, the philosophies they preach and practice are much different than his. Proponents of a flow-based puzzle-design aesthetic are militantly player-first: they argue that a puzzle should provide only enough of a challenge to provoke an “a-ha moment”, being careful to neither harass the player into frustrated thrashing nor let them slide into boredom. The best way to find that balance? Rigorous playtesting and endless tweaking.

It’s an approach Blow disparages in a recent Polygon interview promoting his next game, The Witness:

“A lot of people when they sit down to design a puzzle game, they try to think up some smart things that are hard to figure out, stuff where people will be like ‘Gotcha!’ and figure it out and feel smart themselves. That’s step one. Then step two, they think of how to teach those mechanics to players. … But at some point — a lot of companies do this, they start making a game, and they put it through focus testing, and any time someone gets stuck you consider it bad and change the game. That leads to extremely conservative, mechanical games.”

On the other hand, then, is the alternative philosophy practiced by designers like Blow, Antichamber’s Alexander Bruce, Fez’s Phil Fish, and (I might argue) Dark Souls’ Hidetaka Miyazaki. These designers argue either explicitly in interviews or across the body of their works, that a puzzle should constitute a rigorous mental challenge and as such, must necessarily strive to push players beyond their own preconceived limits. If a puzzle is so difficult that most players can’t solve it…well, so what? Let them fail, let them fall, let them be culled from the herd, and let game designers confidently practice the same artistic disregard an author might pay to an obtuse reader; not every work of art has to cater to everyone. As Blow puts it: “If you’re going to design something that no one gets stuck on, it has to not really be a puzzle.”

These thoughts and others ricocheted around my mind while I tried and failed to lock down an opinion on Framed.

Released late last year on iOS by developer Loveshack, Framed took home strong reviews and more laurel-leafed indie awards than you can shake a stick at. The game’s concept is gimmicky, but straightforward in action: players are presented with a series of comic-panels, which they rearrange to change the context of the actions depicted therein, as the protagonist moves across the page from left to right. Your goal is to help the little comic man escape all the way to the end of the page.

What’s interesting about Framed is how it feels to play: hopelessly, impossibly complex. In the same way that a chess-board can be modeled as a hierarchical tree of moves, yet as a given game proceeds, expands exponentially beyond the bounds of human cognition, so, too, for Framed: on a page with 8-12 panels, there’s simply too much going on. It’s impossible to see more than one or two moves ahead.

What this means in practice is that you can’t approach Framed like you would a puzzle in a game by Blow, Bruce, Fish, or Miyazaki. There is no room for creative down-time, no bath-tub “a-ha” moments to be found.

Yet, neither can a player beat Framed through random thrashing. There is undeniably a point of friction to it, a mental work happening that feels distinctly puzzly in flavor. It reminds me of a Rubik’s cube, or one of those sliding-tile games. You can’t win through brute force, and you can’t win by thinking it out, so all you can do is shuffle vaguely in the desired direction, trying and failing and trying and failing until, suddenly, you’ve somehow succeeded. But is that enough to constitute a “puzzle”?

And if so, is it a good one?

I still have no idea. If you were to ask me, “What’s a better puzzle game, Tetris or Myst?” that would be an existentially-perilous question. It would rattle me to my core. I might lash out with drugs and rock ‘n roll music — or I might forsake all earthly possessions and climb back into the trees.

Maybe the answer is irrelevant, or impossible to establish, and attempting to do so is only a limp stab at lit-major-y categorization. Maybe the answer will be different ten years from now. Maybe someday we’ll look back at the puzzle games of years past and simply say, “No, they weren’t really puzzle games; they were [something else].”

What is a puzzle?

Frankly, I’m stumped.