Development Hell: An Analysis of Pony Island

Note: This article contains major spoilers for Pony Island, Spec Ops: The Line, and Undertale and mild spoilers for Doki Doki Literature Club and Frog Fractions.

You can only see and play so much horror before you become desensitised to it. For devotees of the macabre medium, what was once nail-biting becomes predictable and what was once terrifying becomes tame. Bait-and-switch horror is one response to this desensitisation. When we start games like Doki Doki Literature Club or Undertale which open with a plush, inviting atmosphere instead of trying to be chilling from the outset, we let our guard down and become more receptive to their horror. By establishing a lighter tone earlier on these games also have something to contrast their later darker tones against. Engage with a narrative that keeps dropping quarters into the grim atmosphere machine the whole time, and when the moment comes for physical and psychological pain to be afflicted on the characters, it doesn't seem that much worse than where you started, but if the characters began in a world that felt sanitised and protective and then lost that safety net, there is the sense of a nosedive. This unseats the audience because it plants the idea in their heads that even if their existence feels perfectly safe, they may, in fact, be as vulnerable as these characters. Games with hidden depth and hidden instability function through similar methods. They start stable or shallow and then become broken or elaborate, and the shift accentuates the brokenness or elaborateness. But Pony Island, I'm sorry to say, is an example of a title that's so, so confused about how to pull off this kind of bait-and-switch.

In Pony Island, we go hand-to-joystick with a fictional arcade game called, appropriately enough, "Pony Island" which is home to Lucifer, an entity who's not quite a demon and not quite an AI. He has stolen the souls of many players before you, including that of the character "Hopeless Soul" who acts your Virgil through the cabinet's infernal OS. Some patches of the Pony Island within Pony Island are unfinished, and others are underwhelming, so you manipulate the underlying code of the cabinet to hack your way through the game, even when Lucifer protests that this is cheating. At the end of your ordeal, you defeat this coin-operated Satan and free some spirits, but Hopeless Soul asks you to uninstall the game so that they don't spend eternity rotting away in the machine.

The premise of Pony Island is shaky: Why would anyone install a machine in an arcade if it didn't have something close to a finished game on it and why is an arcade cabinet running the complete OS for a PC? Is the body you see in front of the cabinet a player surrogate body or is there a character distinct from us who is playing the game? If that body is meant to be us, it somewhat takes us out of the moment to be playing an arcade game inside the Steam game, and the developers can't account for you potentially having a different physique or skin colour from the player model. If the body is meant to represent a protagonist independent from us then what's going on when the game tells us to delete it or in the scene where it produces fake Steam notifications? Neither of these things would be possible within an arcade game. The cabinet feels like an unnecessary barrier between the player and the intended spookiness of the game within the game. There is a pane of glass which keeps us safe from Lucifer, as though we were viewing his exhibit at some zoo of the damned.

But the main problem is that Pony Island is flipped the wrong way up. The name and the title screen are the cutesy bait of this bait-and-switch, but as soon as you hit the "Start" button you're swallowed into its demonic cyberspace, and you don't see that innocent, deceptive, more stable version of Pony Island resurface until roughly halfway through the game. So when you hit the "creepy" sections of Pony Island, there's been nothing for them to subvert or contrast against, and when the "normal" sections come back around, the purpose they would have served is moot. That's not to say that they're not pleasant, but they're also sometimes pleasant for the wrong reason.

The visuals of the "innocent" section of the game are better realised than those of the "corrupt" sections of it. The vibrant joy of the "finished" Pony Island feels like its approaching that of old-school treats like Kirby's Adventure, but the parts of the game supposedly under construction don't have the seams of an actual computer game prototype. Compare the way that Pony Island depicts an incomplete video game to how The Beginner's Guide does it. Not only does The Beginner's Guide have all the scaffolding of a work in progress, something that Pony Island doesn't, but you also get to see the fictional artist behind it become more proficient at 3D graphics as it goes on. Pony Island has brief stints in which its visual style changes like the 3D section or the "finished" section, but they're seen as side chambers which the player wanders into and then back out of to return to the singular look chosen to bind the game together. It makes it feel like the visuals are stuck in the mud; the style never goes anywhere, and despite Lucifer's work on it, the graphics never seem to get any more polished. The sound design and at least half of the gameplay take this same form. The game contains puzzle sections which introduce new and persistent design parts, but the endless runner play which is the game's bread and butter undergoes relatively little modification from act to act. Even the story doesn't veer and sway as much as you'd think it would: The stakes and general goals remain the same throughout; you're going after one of three daemons so you can delete their "Core File" and get one step closer to breaking the OS.

The problem isn't that Pony Island doesn't introduce new ideas over the course of its roughly two and a half hours, this is something it's exceptional at, but because the most transformative concepts it introduces aren't persistent, the game doesn't evolve. Most titles present an idea, stack another idea on top of it, stack another idea on top of that, and before you know it, you have the kind of tower of interacting elements that can give a game depth. However, a title like Pony Island frequently adds new accessories only to remove them and so never accrues enough interconnected concepts to feel like more than an easily-distracted experiment. A select few works can get away with having lots of one-off design and aesthetic components like WarioWare or What Remains of Edith Finch, but you need remarkably engrossing frameworks to pull that off the way they do. Frog Fractions did this with a gossamer thin structure to hold it all together, but trying to replicate such an eclectic design is probably going to jeopardise the consistency in your story, and Frog Fractions is a tough act to follow. Pony Island is at this midpoint between trying to create a plot arc and aesthetic that is persistent throughout the game and making frenzied jumps from concept to concept. Attempting both at the same time means it achieves neither.

There's no set amount of original ideas a game should introduce over its course, but the hollower the game is and the more derivative the place it starts from, the more need there is for it to begin giving the player experiences they haven't had before. That is to say that if you're creating a game in an archetype that audiences are generally familiar with like "tower defence" or "endless runner", you will need to bolt some transformative ideas onto their chassis unless you want your audience to get bored fast. That goes double if your systems are never going to offer the player a wide variety of options. After years of endless runner games, Pony Island's meek iterations on the formula are not enough to support a whole game, even if that game is short.

Of course, every genre has its conventions for how to evolve a game over time and different reasons to initiate that development, but evolution may be even more necessary for self-deconstructing horrors like this one than for your average title. An experience like Pony Island needs to create not just the sense that evil is encroaching on the protagonist, but that that encroachment threatens the fabric of the piece of media itself. As play or narrative progresses, the protagonist must not just face uglier foes or more gruelling tests of strength; there should be a sense that the coherence, functionality, or fourth wall of the game is disintegrating over time. For example, Undertale gets increasingly self-referential as time goes on, and Doki Doki Literature Club exhibits more glitches and violations in the rules of its fiction the further in you get. Pony Island doesn't undergo an equivalent growth, and it may be almost impossible for it to do so when you can see menu items falling off the screen in the first minute. There's not that much more broken it can get from there.

Unlike The Beginner's Guide or DDLC, this game also fails to say very much with its metatext which is another consequence of its lack of persistent ideas. Perhaps the game's smartest moment is early on when you're cheating your way past Lucifer's faulty programming and poorly-considered systems, and he throws a tantrum in response. The message is that devs who complain that users are "ruining" their game through hacks and mods are getting in the way of people fixing and even streamlining mechanics. There is also a moment where it tries to propose the kind of catch-22 Spec Ops: The Line did, and the characters tell you that if you stop playing, you'll not see the story unfold, but if you keep playing, the files on the in-game machine will be corrupted. But there's no actual dilemma here because corrupting the data carries no negative side-effects. After this, the story doesn't get interesting again until the conclusions of the game, both of them. Unless you've uncovered all of the game's secrets, you reach the ending in which the ally character, Hopeless Soul, will tell you that they want you to uninstall the game so that they're no longer electronically incarcerated. If you instead come back to them with every hidden ticket in the game, they'll tell you that they understand you're looking for a challenge to struggle against and so Hopeless Soul makes themselves the game's dominating final boss. If you beat them, all the fervour builds to a deliberate anti-climax in which they tell you that you must shut the game off and that there is nothing left to see.

There are parallels here to Undertale's "Genocide Run" ending which you can reach by killing all enemies in the game up to that point (as opposed to running from them or pacifying them). In Undertale's story, a former buddy, Sans, reacts to your completionist attitude by becoming the pointiest of difficulty spikes in the final boss fight. If you survive this incarnation of Undertale, the world ends, and you win nothing. The design punishes your voracious villainy by effectively bricking itself. Undertale beat Pony Island to the punch on this one, but there are relevant differences between their endings. Sans is trying to dissuade you from playing any more of his game, while Hopeless Soul is using their boss fight to encourage you on. These closings are also punitive in different ways. Pony Island is more deliberately disappointing; at least when you beat Undertale you feel like you're affecting the world in some way, but there's no "true ending" or easter egg to be seen at the end of Pony Island's secret boss fight. It remains resolute in ensuring that your decision to torture this character for a few more hours of gameplay is not rewarded in any way. However, Undertale is the better game at putting across the message that trying to consume every morsel of a game can wreck it. It gives you a world and characters most people are going to let into their heart and so when you see their story hit a hideous conclusion that you're responsible for, it knocks you for ten. You don't have that emotional investment in Pony Island, a game where the characters are the literary equivalent of stick figures.

While Pony Island may have intentions of being an aesthetically darker game than an Undertale, the rough similarity between the demonic and the cartoonish sections of Pony Island unintentionally highlights a shallowness in the lenses we have for looking at "edgier" games. Computer games are no longer as adolescently-minded as they were ten or twenty years ago; there's not this gnawing insecurity that if we don't make games look like grit-encrusted slaughterfests, that the medium won't feel "mature". But we haven't entirely detached ourselves from those roots either; maybe popular culture in general also retains some of that attitude. We tend to look at a game that's trying to be any more soft and fluffy than say, a Super Mario, and think that its innocence represents naivety. Pony Island would agree, with the smiling animals and bright colours of its innocent section intended to look laughably naive in comparison to the maelstrom of demons and garbled code writhing underneath.

The odd thing is, there's no less depth in the softer portion of the game than there is in those where it outright lets you know there are devils in the wires. Because the innocent sections are a complete reskin of everything you've seen up to them, they're just as mechanically competent as any other, and while this chapter's stylistic shift might mean the abandonment of complex characters and story in another game, Pony Island doesn't those have elements, so you're not losing anything in the switch. Put another way, Pony Island gives you an example of a screeching, satanic horror that is meant to be the "real experience" and then a candy-coated twee cousin to that game which is implied to be ridiculous by comparison. However, the latter version of the game comes off as no less legitimate because it's no less developed in its writing and design than the former. If the game cut off where the "happy" levels end, it could have closed on a prescient point, highlighting the validity of games that aren't intending to be as intense or violent as the AAA norm. However, by dismantling the "cutesy" aesthetic of the game to return to the "defective" video game look, it makes it clear that it does not think of a softer art style as anything more than an interruption to the supposedly more worthwhile edgy content.

Pony Island does occasionally show a little intricacy in the design, but it happens in the puzzles rather than the narrative or action. The puzzles consist of interactive screens of fake program code; diegetically, this is the code of Pony Island itself. A cursor scrolls its way through the lines of the program, executing the commands on each of them, and we have to find a way to get that cursor to a line at the bottom of the screen. At fixed points in the software, there are empty slots into which we can insert different tiles, each kind causing the cursor to move ahead in a different pattern when it reaches them. It's par for the course that levels contain multiple columns of code that you can shift the cursor across, and there are sometimes other conditions that you must fulfil before the cursor can pass over specific gates in the program.

Over the years, we've seen plenty of video games with puzzles in which players must manipulate programming to clear objectives. Few of these games ask the player to do any actual programming, not least of which because the real deal can quickly tie your brain in knots, and so most video games with fake programming look at what parts of the task they can remove to make it accessible to their audience. Some games let players click puzzles pieces together in place of writing actual code, others use highly simplified programming languages. Pony Island's tactic is not to make us construct or edit any of the code it gives us, but to present a pre-made program and ask us how we should navigate it, and because these puzzles are about navigation, they're really spacial puzzles more than they are hard logic brainteasers. They remind me a bit of Chu Chu Rocket in that you have to use your arrows to redirect moving objects to targets.

The gameplay elements that Pony Island stacks onto the puzzles work better than those it stacks onto the endless running because these elements radically alter how you act within and think about the play whereas those in the endless runner don't. And while a few games before Pony Island explored the idea of being a tourist in a cyberspace where the walls are coming down around you, Pony Island might be the first game I've seen where you're using your designer-given tools to put it back together. The software may be corrupting, but you can maintain some order through play.

It speaks volumes about Pony Island that it gets the play where you're maintaining the integrity of the game correct but that the portions of the game where it's meant to be falling apart are not that fascinating. If we look at it from a developer's standpoint, however, this is not that unusual. It takes more time and expertise to get a game that simulates brokenness with artistic flair than it does to get a game that doesn't appear to break at all. As ironic as it may be, the relative consistency and professionalism in Pony Island is a detriment to it when the premise is that it's a haunted, crumbling work of engineering.

A piece of media doesn't have to succeed at everything it tries to be interesting. It doesn't even have to be good at anything it attempts to be interesting. But it needs original ideas in at least one field, or it will be outmanned and outgunned in all areas, and Pony Island is. If you want a game that challenges how consistent games need to be, Frog Fractions is better; if you want a game that speaks more fluently through the fourth wall, there's Undertale. If you want a game that devolves into graphical chaos try the demo of Memory of a Broken Dimension and if you want a more rousing endless runner about being a horse that can fly and shoot lasers, play Robot Unicorn Attack. That said, I think game creators would be remiss if they didn't pick up and run with Pony Island's most captivating ideas: Its systemisation of the battle between the player and the developer, and its illumination of the worth of the softer, cuter game. Thanks for reading.