Intermittent Mechanism

Ian here—

Here is the first real episode of my new Shape Up! series of video essays about form, structure, and pacing in games. My central case study is Skeleton Business’ 2019 game Vignettes, although before I get there I spend a whole lot of time placing it in the context of what Joel Goodwin has termed “secret box games” or “himitsu-bako games.” This has been a topic I’ve written quite a few blog posts on. For years, I wanted to coalesce some of those disparate ideas and game appreciations into a video essay, but I never found the proper way to do that until I inaugurated this series.

Script below the jump.

Hello everyone, and welcome to the first episode of Shape Up!, an ongoing series about form, structure, and pacing in games.

In the introduction to their book The Game Narrative Toolbox, authors Tobais Heussner, Toiya Kristen Finley, Jennifer Brandes Hepler, and Ann Lemay write:

“While screenwriters only have to equip their audience to answer a single question after every scene—“What happened?”—game writers must make sure players also know “What do I do next?”….

[N]ot knowing what to do next is the single biggest sin in interactive writing. Yet, striving too hard to be clear can easily shoot you over the line into boring, repetitive, expository, and pedantic.”

This is an interesting two-way conundrum. As a player, I have to agree: some of the biggest frustrations I’ve ever had playing games come down to not knowing what to do next. It’s a frustration that seems to be largely unique to the medium, a little medium-specific hurdle for artists working in this particular space. But also, some of the biggest moments of joy I’ve experienced in my years of playing games have been moments of surprise and epiphany, moments where free-form exploration with mechanics bore fruit in unexpected ways—or free-form exploration of a digital landscape revealed unexpected secrets. And when games are too expository and pedantic, that quashes the possibility of those joyous moments of discovery. That’s another danger that seems unique to the medium: the danger of signposting all of the wonder away.

This difficult balance affects more than just game narrative design. It permeates multiple elements of game development, from level design to UI design. The case studies I’ll be using for this inaugural episode of my Shape Up series are actually games that have very little in the way of traditional writing. I wanted to start here for this specific reason: since I’m not writing a textbook on game narrative design, I have a lot of freedom in how I define game “shape.” It’s more than just narrative shape. This series will be an all-encompassing look at how games pace and modulate the general experience of play. That might be through writing and narrative construction, but it doesn’t have to be.

In this video, I’ll be looking at a particular philosophy in game design that avoids overly-pedantic exposition at all costs. It’s a design philosophy that’s previously been mapped out by Joel Goodwin, a prolific critic in the indie games space known for his old-school blog Electron Dance and, these days, his accompanying Twitch stream.

In a post from 2014, Goodwin observed that the central activity of a lot of indie games he was seeing at the time was gradually revealing a secret. Challenge wasn’t really the primary pleasure these games offered. Instead, these games were all about about exploring a possibility space, and gradually discovering whatever unfolding experience the developer had curated for you.

Several of Goodwin’s examples are commonly referred to as “walking simulators.” That’s a term that I use from time to time, because it’s convenient, and people know what you’re talking about. But I have a little pang of regret every time I do, and it’s because of Goodwin’s critical writing on the subject. Rather than “walking simulator,” Goodwin prefers the term himitsu-bako, or “secret box” games. 

As a term, it’s … really a nonstarter, I’m afraid. It hasn’t gained really any traction in the decade since Goodwin first proposed it. Which is unfortunate, because I like Goodwin’s conceptual category. It includes 3D games like The Path and Proteus and Slave of God and NaissanceE, to name a few personal favorites of mine often described as “walking simulators.” But Goodwin points out that the secret box design philosophy is also present in things that aren’t 3D exploration games. Think of games like Happy Game, or GNOG, or, to pick an example from this year, Birth. No one would call these “walking simulators.” On the surface, they resemble point-and-click adventure games—because, after all, you have to point-and-click on things to progress. But they’re overall less dependent on puzzles than traditional adventure games. Many of the activities you undertake while playing them are much more akin to the digital equivalent of a pop-up book, or an advent calendar. These games don’t frequently trying to stump you. You don’t have to use much ingenuity or lateral thinking to get through them, and they aren’t aiming for that same feeling of accomplishment upon solving something that you get from a traditional point-and-click adventure game. The focus instead is on creating interactions that are pleasurable in and of themselves, through expressive use of animation, sound design, music, or just the basic components of UX.

I like Joel Goodwin’s concept of himitsu-bako because its broadness. So much of 2010s games discourse was wasted on arguments about what was, or wasn’t, a walking simulator. The secret box game category transcends these petty arguments, allows us to see larger aesthetic congruences in a wide range of games that traded the pleasures of challenge for the pleasures of free-form discovery.

Although, that’s not quite right. There are secret box games that do have at least some challenge. I think the category Goodwin has gifted us with is broad and malleable enough to allow for as spectrum of challenge—and also, relatedly, a spectrum of objectives.

To take three examples:

Bernband is a game with no challenge. You explore an alien city. The only activity in the game is exploration. You go through a series of doors, and these doors lead to a series of rooms. Some rooms have multiple exits, and you make a decision about which exit to take. There are six dead ends in the game, so if you reach one of those, you turn around, keep exploring what you haven’t seen yet. Other times, an exit to one room will unexpectedly loop you back to a room you’ve been before. There’s no real “end” to the experience. Eventually, if you go through every door, you’ll have seen every room—but the game doesn’t then say, “you’ve won!,” and play some end credits. There are no end credits, and there’s no main menu to return to—the game only stops running if you quit it. So not only is there no challenge to the experience of playing Bernband, there’s also no objective. If you want something to do, you could spend a few extra minutes mapping the space, figuring out which door leads to where, making sure you’ve found every secret little corner, make a final determination as to whether the layout is Euclidean or non-Euclidean. You could do that, but if you do, it’s only because it was a self-imposed objective, a little challenge you gave yourself, not something the game asks of you. Bernband does not ask anything of you, any more than a museum asks something of you, or a park asks something of you, or a city block asks something of you.

Example #2: THAT NIGHT, STEEPED BY BLOOD RIVER is also basically a series of rooms. You start the game in a hotel lobby of sorts, with three rooms branching off. If you interact with the beds in each of these rooms, you’re whisked away to fantastical spaces. Each of these spaces has two possible exits, which warp you back to the elevator in the main lobby. In many ways, THAT NIGHT, STEEPED BY BLOOD RIVER is quite similar to Bernband. But there’s a crucial difference: the game tracks which exits you’ve used to each room, and there are two final areas—a basement boiler room and a tree-filled atrium—that require you to have used certain exits to unlock. So now we have an implicit objective: exit each of the three areas in the “down arrow” way to unlock one ending; exit them all in the “up arrow” way to unlock the other ending. And there’s a bit more challenge involved, too. The exits designated as “down arrow” exits are more obvious. All they require you to do is sign a series of papers, which is a mechanic the game introduces early on by forcing you to sign into the hotel’s guestbook before you proceed. The “up arrow” exits require you to solve musical puzzles based on the music playing in each area, which is much less obvious, especially to those who aren’t musically inclined. They also tend to be just a little more tucked away than their “down arrow” counterparts. So we have at least two definitive endings to the game—and I’m actually not willing to swear there’s not a third one, but if there is it’s well-hidden—with some degree of puzzle-solving to unlock the harder one. We’ve moved over one tick on the objective gradient, and also moved over on the challenge gradient.

And then you get to something like my third example, Kairo. Kairo is full of genuine challenge. There are actual puzzles you need to solve. Kairo also has a definitive objective: go through all the rooms, solve all the puzzles, finish the game, and see the end credits. And you might say: okay, so, Kairo is a first-person puzzle game, then. But this doesn’t fit with the actual experience of playing the game—that’s just not Kairo’s vibe. Kairo aims for an “alien world” aesthetic, with players vaguely grasping at only partially-understood cause-and-effect relationships. And it succeeds in this! Almost every single puzzle in the game is filled with a bunch of trial-and-error, as you re-learn the rules of whatever particular space you’re in the moment. Kairo doesn’t offer players a coherent vocabulary of mechanics it then systematically builds on, a la, say, Portal 2. It expects you to constantly explore and re-explore the possibilities of interaction, keeping an open mind about what is being asked of you and not assuming that what you’ll be doing in this particular room has any continuity with the room that came before. And this kind of deliberate inscrutability to me seems a lot closer to the “secret box” design philosophy than the “first-person puzzle game” design philosophy. (It helps that Joel Goodwin was a big proponent of the game, wrote several blog posts about it, and was actually a financial backer of the project.)

Based on everything I just ran through, we can put together a Grand Two-Axis Taxonomy of Himitsu-bako Games.

In the one corner of this taxonomy, you have space for games that offer no challenge, and no sense of progression toward an ultimate objective—so something like Bernband. 

And, in the other corner, you have things like Kairo, which offers both puzzle-based challenge and a clear progression through a series of levels toward a well-defined ending, and yet still retains that distinct vibe of a secret box game by virtue of remaining mechanically unpredictable throughout the entirety of its running time, so each new room feels like its own revelation.

Then you have a large area between where games can have a greater or lesser degree of challenge, and also a variable sense of whether or not you’re moving toward a defined objective.

And one of the interesting things about himitsu-bako games is that these two aspects—challenges and objectives—are often tied up inextricably together. The fact that these games don’t immediately and clearly communicate the bounds of their possibility space can itself become a source of challenge. The action you need to take to “beat” the game may be trivial, but figuring out that there is something the game wants you to do to reach a designated terminus can itself be a challenge. These are games about pursuing secrets—and sometimes the secret is a secret objective, or a secret ending.

Two contrasting examples here: neither Dear Esther nor Proteus ask their players to perform any actions that require quick reflexes. Neither game has puzzles. Neither game requires you to pick up any collectables to continue on to the next level. In terms of their control schemes and their mechanics, they’re identical. The one place where they differ is how clear their objective is. Dear Esther funnels its players through a series of levels in which they walk through a largely linear path toward some well-defined landmarks. As they walk this linear path, a story is narrated to them, until they reach the game’s narrative and geographical conclusion. Proteus, on the other hand, at first glance appears to be an exploratory free-form musical toy with no real objective. It does in fact have an ending, but it’s not clear from the outset how to trigger that ending. I would say that it’s more challenging to reach the ending of Proteus than it is to reach the end of Dear Esther. But the challenge isn’t one of dexterity. Nor is it one of mental ingenuity. It’s mainly one of patience: a willingness to explore the environment and observe how it reacts to your behavior, even in the absence of any traditional signposting or well-defined reinforcement mechanisms.

And patience is really the crux of things. Considered in hindsight, himitsu-bako games were part of a larger aesthetic trend in indie games during the 2010s: a trend that eschewed tutorials and as many UI elements as possible. The goal was to make games into perfectly elegant vehicles for pure discovery, where the physical layouts of the levels themselves suggested what to do, while also preserving a sense of mystery. It was an aesthetic trend that, contra the authors of The Game Narrative Toolbox, saw the act of pedantically dictating to players as the biggest sin in games, and was willing to run the risk of letting that question “What do I do next?” go unanswered, as players gradually found their own path forward.

This aesthetic trend has waned a bit now. There’s definitely a lower percentage of indie games in the 2020s that are trying to be mysterious art games, deliberately hiding their scope and toolset. Your average indie game today is much more likely to include a clear tutorial that actually teaches you the vocabulary of actions you’ll be expected to take within the game. You’re more likely to encounter text-based instructions, a formal quest log, a voice on the radio—maybe even a character doling out oodles of positive reinforcement for every correct thing you do. 

And it’s clear why—it gets back to that idea that above all else you want to avoid the player not knowing what to do. Not understanding a game’s scope, or even just the basic vocabulary of interactions that it provides, can be a frustrating experience. For every one player experiencing a moment of wonder and epiphany when a secret opened up, there will be one—or maybe many!—players who bounced off of the game hard because they were frustrated by the game’s obscurity or dearth of expected mechanics. Any design philosophy that takes player patience for granted is going to be difficult to sustain when it hits commercial realities.

I’m of two minds here. I can’t deny that even I myself—a real connoisseur of secret box games—have bounced hard off of several of them. Mirrormoon EP was personally recommended to me by a friend who knows my tastes in gaming well. But ultimately I found its scope to be too intimidating: it was just too much exploration and experimentation with too little sense of direction. 

At the same time, I do miss the golden age of himitsu-bako. I miss the era when everywhere you looked there were games that had the confidence to simply throw you into a world, and let your own interactions with that world guide you. I liked that sense of not knowing if the space I was exploring was going to be the staging ground for an allegorical story about the experience of immigration and assimilation …

… or if I was in fact supposed to figure out that I was inside a giant synthesizer and was expected solve music-based puzzles.

Maybe the non-Euclidean space I was exploring would just turn out to be an art gallery I’m free to roam without pressure …

… but, on the other hand, maybe that non-Euclidean space would turn out to be a series of escalating puzzles.

Maybe I was going to spend a couple of hours doing some light platforming across a strange landscape …

… or maybe that landscape was something I was going to have to learn how to keep an unusual vehicle moving across.

Maybe I was playing a horror game in which nothing was going to jump out and kill me, but both the space I was exploring and the game itself kept mutating in unpredictable ways …

… or maybe I was playing an interactive cartoon, in which the funniest option of what to do next was the correct one.

Basically, I liked that aesthetic experience of not knowing a game’s possibility space. I liked that  feeling of actively exploring the mystery of what a game was going to ask of me—whether it was going to be parsing out an unusual sequence of cause-and-effect, or navigating a maze, or maybe sometimes just wandering around, and having the patience to appreciate the journey, even in the absence of a clear destination.

But I also 100% understand how not knowing what to do next can be frustrating—to the point where it’s considered the biggest sin in game design.

And the question I want to ask is: can these two things be reconciled? Can you build a game of free-form discovery, based on revealing secrets in the himitsu-bako mold, that skillfully avoids the danger of players feelings lost and frustrated because they don’t understand a game’s scope, or objectives, and are left with no clear feeling of progression?

I you can—and there’s one game I think does it very well. It’s a game that came out in 2019, and I think you could make the case that it marks a perfect transition point between the free-form mystery that marked so many 2010s indie games, and the more safely structured experiences you’re more likely to find today. It’s called Vignettes, it was made by four folks operating under the delightful moniker Skeleton Business, and it’s going to be the final case study of this video.

Vignettes is a game about manipulating perspective. There’s a brightly-colored object in front of you, which you can rotate in 3D space. Your task is to find a maximally-ambiguous angle from which to view the object, at which point it will turn into a completely different object. Rinse and repeat—objects can turn into other objects, can turn into other objects, can turn into other objects. chaining together into a seemingly endless path of discovery.

This is a very toy-like experience—in fact, the developers themselves refer to Vignettes as a “toyish surprise-o-rama.” Playing with a toy is free-form and unstructured. You discover what you can do with it, and you make your own fun from there. But the thing about toys is that you eventually get bored of them. This might take minutes, or it might take years, but eventually the sense of novelty and delight wears off. In the absence of more structured goals, you lose interest. This gradual and inevitable waning of interest has been the death-knell for many a himitsu-bako game.

Vignettes, however, does something very simple that’s also very smart: It includes a menu. This menu has a map, that lets you see, in overview form, exactly what objects you have and haven’t discovered. It also includes little still-life images of all of the objects you’ve discovered in each zone. When you complete a given zone, you get a key that serves as a shortcut to start with a different object in your next session, speeding along your discovery of the game’s zones.

The inclusion of the map and the still lives may seem like a small thing, but it’s utterly transformative: it allows for the blending of two different experiential modes. If you want a free-form toy that indulges you in the pleasure of unstructured discovery, Vignettes can be that game for you. I’ve very much enjoyed playing the game in 10, 20-minute bursts where I’m just following my own sense of curiosity, falling down rabbit holes and discovering some of the weirder transition points, where the game becomes a bit more overtly puzzle-y. 

And then, eventually, I get bored, and start to feel a bit lost, like I’m going around in circles. But whenever I feel that way, I can pull up the map, and tap into a very different sort of pleasure. This is the structured pleasure of completionism, of filling out every last node on the map, digging up all undiscovered content. It’s the gotta-catch-‘em-all mentality, which is a very different mentality than the freeform toy mentality I was just enjoying. But Vignettes balances both of these, letting players switch from one to the other at will, mix and match not only different play-styles, but different overarching philosophies of fun. 

I know it sounds like I must be over-stating my praise—I’m talking about a menu, for Christ’s sake! It’s just a menu. But frankly it’s rare for a game to get this balance so right, to use the completionist itch as a general structuring principle, while also providing plenty of space and breathing room for the intrinsic motivation that is the free-form pleasure of discovery.

There are himitsu-bako games that have come out in the 2020s: things like the aforementioned THAT NIGHT, STEEPED BY BLOOD RIVER, also Éric Chahi’s VR game Paper Beast, and Studio Oleomingus keeps putting out masterful work. Games like these have made me optimistic that the himitsu-bako design philosophy, while no longer the force it once was in the mid-2010s, still has a future ahead of it. And I think that future is going to be dependent on learning the lesson of something like Vignettes: finding that perfect fulcrum point where players aren’t pedantically told exactly what to do next, but at the same time are given enough of an overall sense of the game’s scope and possibility space that they don’t succumb to impatience. Vignettes is an excellent example of a game that signposts just enough to give you a sense of progress, without signposting all of the wonder away.

In the next episode, we’ll be staying on this theme of signposting, with a special focus on the use of irony. That’s it for this episode of Shape Up!—stay tuned, and thanks for watching!