In the wake of Spelunky, procedurally-generated levels have become a moniker of the age. Not to historicize preemptively, but can’t you just see 2010-2015 as the age of the Roguelike, just as 1980-1985 was the age of the arcade? The general consensus has been to expand Derek Yu’s concept with more stuff: more RPG elements, bigger worlds, more variations. Not many have thought to refine it. But Ojiro Fumoto (alias “Moppin”) thought that reduction was possible. He proved it with Downwell.
Though I only played the PC version, the limitations of the original iOS format show themselves with intrusive clarity. They are so intrusive in fact that Downwell, like every great game ever made, builds itself without dreams of its own: it is constructed entirely out of those limitations.
The vertical shape of mobile devices drives its atmosphere because it inspires its field of view. It’s a reverse shoot em’ up platformer (you come down from the top of the screen instead of the bottom), probably so that your thumb won’t get in the way of your character. Instead of two buttons for jumping and shooting, Downwell has only one. This means that you can only shoot after jumping. Which means that you can only shoot down. So its gameplay emerges out of the simple needs of its platform, as every micro-element of its design has been shaped by its limits.
The levels are brief but packed with the action of a thirty-second level of Defender or Joust. Bouncing on enemies as you fall is a joyous art-form, as is the conservation of your ammo (which only reloads once you land). End-of-level perks spice the formula without overpowering you with options: only skill will help you reach the bottom. The currency system adds depth unforeseen by the simpler days, when levels were called “boards.” But the heart of Downwell still belongs to that quarter-popping formative era, with gameplay made only more addictive by shrewd variety.
I say “shrewd” because the algorithm that shuffles the design is subtler in Downwell than in Spelunky. The caves were brand new each time you started, and spelunking them had a bit of luck involved if the intricate layout stacked against you one time or another. Downwell doesn’t change that much: I can’t remember the levels enough to notice what’s different about them. You sift through them so quickly that procedural generation would scarcely matter if it weren’t for muscle memory.
But by a slight variation every time, the insurance is on your attention more than the game’s difficulty: procedural generation doesn’t make Downwell difficult, but keeps it interesting. The levels are essentially the same, but they are never committed to your memory. You can’t sleep through a single playthrough. The odds are always equally in no one’s favor.
In fact, the low-bit style and simple mechanics make Downwell a game that could have existed in the 80s, and would’ve been a tremendous hit on the machines. Its two-color palette (which can be swapped with unlockables) is another element driven by necessity (in this case: of being able to see enemies clearly on a small screen). But that pure sense of function inspires the game to sparkle with clarity that a lot of modern releases just can’t match.
The balance Moppin achieves so brilliantly is in having enough material to make a full game without any extra elements. Like a great sculpture, nothing can be shaved from Downwell to make a better experience. Even the shops offer constant tradeoffs between ammo capacity and health regeneration, rewarding skill while cushioning beginners. Downwell is on your level immediately, no matter if you’ve jumped into it ten or ten thousand times. It has no afterthoughts.
As a player, I’m used to being treated as an inconvenience, in games that get bored when I don’t proceed quickly enough, or which feel the need to guide me with an overcautious pen-light to every pre-rendered objective and obvious obstacle like a cat that needs babysitting. But Downwell is a game that acknowledges agency by giving you the tools to understand everything without needing to be told anything.
The best thing about the great arcade games (and the thing that most modern games can learn from) is that they were self-evident. The function of their elements was a function of your interaction with them. There was very little that existed just to pretty the place up. Graphics technology provided an opportunity to place aesthetic decisions above functional ones. I like pretty things, and I acknowledge the atmospheres that some games have been able to achieve with modern technology.
But the same thing happened to film when it discovered sound. Suddenly, anyone could tell any story. And suddenly, films were without atmosphere or energy, like talky news broadcasts without an aesthetic purpose. Graphics tech is too often used like sound, as a shortcut to an atmosphere. Eventually, there’s so much non-functional extra stuff that what little you can interact with has to be explained in detail, sometimes by literal road-signs. Quest arrows and interactive tutorials are creations of an age that prioritizes self-congratulations and graphical glamor over interaction. No matter how pretty a world is, being led through it on a string is less enjoyable than being free in a basement. Being shown something is better than being told.
So when I say that Downwell is a return to a simpler time, I want to be clear that the game isn’t simpleminded. It complexly engages both your twitch reactions and your survival instincts, providing a sense of exploring new worlds as those old games did: in blurs of action. Its simplicity is admirable, as something can only be when it’s also perfectly functional. It doesn’t need tutorials, not because its possibilities are unlimited, but because its limitations are all part of its design.
In art there’s an enduring question: can something be both beautiful and functional? I’ve always thought, and Downwell is new evidence in my mind for it, that the real question should be: can something be either, if it isn’t both?