Is Spelunky perfect? This is the question running through much pro-Spelunky literature. Certainly something so idiosyncratic, so random couldn’t be “perfect.” And yet the game delivers on a core promise so far-reaching in its implication as to tease perfection. That is, to be a perpetual adventure machine. An infinitely playable game. The ur-game, foretold by the prophets, a hero’s journey to be repeated ad infinitum for eternity. Grab your whip and steel your loins and get ready to die.
It’s not much to look at, initially, a sort of throwback (think Super Nintendo) Indiana Jones knock-off, with a little adventurer descending downward, level after level. It is also impossible, pretty much. You get four hearts to traverse poisonous snakes and booby-trapped pits and then a jungle full of one-hit-kill plants and then ice caves full of spacemen (don’t ask) and then a temple full of mummies who vomit flies (seriously, do not ask). Everything kills you, quickly. Sometimes it’s pitch black, and you must get through with a torch. Other times half the level will be a gigantic beehive full of bees that are also pretty much guaranteed death. You will be lucky if you even make it out of the four mine levels in your first, say, 30 tries, and that is only the first of four separate worlds. Death comes swiftly, comically in Spelunky, and with each reset the levels completely shift, with new booby traps and snake pits and spider caves, new items to buy and new homicidal shopkeepers to upset and new wyrms to get sucked inside of (again, don’t ask).
“I was really working from a very intuitive and instinctual standpoint. So I made a lot of decisions based on feel.”
That all this endless, compulsively playable variety was the product of one person flew in the face of the prevailing wisdom about indie games when Spelunky was first released. Many of its most prominent contemporaries had been short, personal journeys, like Jonathan Blow’s Braid or Jason Rohrer’s Passage. These were confrontational works, unabashed art games that gracefully wedded their mechanics and art with some higher meaning. But Spelunky was all fucking game. Hard as hell, full of secrets—the type of thing you need a strategy guide for. You play Spelunky with the clenched-teeth tension of a competitive first-person shooter or a high-speed racing game.
The game’s creator, Derek Yu, wasn’t an unknown quantity when he released it. He had already been serving as editor-in-chief of TIGSource (short for The Indie Game Source), a hub for sharing games and talking about the burgeoning indie scene. Yu’s previous game, 2007’s Aquaria, was a dreamy, lyrical adventure that netted major praise in that community. After its release, he wanted to focus on something smaller, simpler, more attuned to his own tastes. “I was really working from a very intuitive and instinctual standpoint,” he told me recently. “So I made a lot of decisions based on feel.” He started prototyping some simpler games of the sort he’d played growing up, but none felt new, worth sharing. It was by trying to come up with something that he’d personally find entertaining that Yu hit upon Spelunky’s singular innovation: fusing the classic, instantly accessible platformer with the relentlessly difficult, arcane genre known as the roguelike.
We all know the platformer, at least in theory. It’s Mario and Metroid and Castlevania, running and jumping and climbing ladders from a side-on, two-dimensional perspective. It’s the three-chord blues of game design from which so much else springs. The roguelike, on the other hand, is a subgenre of almost unfathomable specificity and complexity, the prog metal of games. Its Wikipedia page, for example, stretches on for some 9,000 words, with a detailed family tree and explanations of each possible variation. At the top sits 1980’s Rogue, a seemingly conventional adventure game that nevertheless contained a combination of flavors so exceptional as to inspire an entire subgenre of games built (and named) in its image. In 2008, a group of developers penned a manifesto called the Berlin Interpretation, which managed to define the genre as having eight “high-value” factors and five “low-value” ones, among them randomized levels, permanent death (i.e., no saving), hack-and-slash-style combat, and so on.
If this seems more like the formation of some apostate cult than, you know, a video game subgenre, that’s part of the point. The Berlin Interpretation was an attempt to figure out what makes roguelikes—which have names like Tales Of Maj’Eyal and Zangband—so compulsively playable. Games have, over time, grown increasingly easy, unwilling to let players die for fear they might quit, and in the process they’ve lost one of the real joys of playing them: understanding and exploiting a set of rules. The classic roguelike is an antidote to this tendency. They kill you constantly, and you don’t pick up where you left off but start again from the very beginning; the term for each new game is a “run,” with all the devil-may-care attitude that implies. The point in something like NetHack isn’t avoiding death but in discovering wild new ways to die, like mistakenly attempting to saddle a petrifying monster or being forced to commit something called “self-genocide.” But many of the most prominent roguelikes also trafficked in a sort of willful obfuscation. (Yu, for example, says he hasn’t beaten any of the classics.) The earliest ones used simple alphanumeric symbols instead of more detailed graphics, such that playing them required being able to tell that an ampersand is your character and that a row of “Z”s is a pack of bloodthirsty zombies. Over time, this became codified; in the Berlin Interpretation, it’s one of the “low-value” factors that defined the medium.
Spelunky upended that. “Roguelikes had been around a long time, but they hadn’t seen mainstream popularity,” Yu told me, tired as hell after a week at the Game Developers Conference. “And I think it’s because they were getting more and more esoteric, and Spelunky just added that layer of familiarity and accessibility. It kind of shined a light on it and let other people who were not hardcore roguelike fans appreciate it.” By ransacking the best ideas from roguelikes and presenting them in a traditional, side-on platformer, he was able to draw more people into the idea of making hapless, experimental runs through hazardous randomized worlds. When one of these runs clicks—when the items fall in the right order to give you an edge, the hair-raising escapes actually pan out, and you find yourself in a make-or-break exploration of the game’s final caves—the tension, and the satisfaction, is unrivaled.
And by virtue of creating a brand-new fusion of games, Yu was able to slot in ideas quickly: making the terrain destructible because blowing it up felt good, introducing extremely agitated shopkeepers as an agent of chaos, and rendering everything with a remarkably consistent set of rules, such that strange interactions could arise unexpectedly between them. In one famous anecdote, the game designer Tom Francis writes of unwittingly witnessing a “shopstorm,” in which a boomerang-throwing enemy walked into a marketplace and stole a boomerang, pissing off the shopkeepers and setting off an explosion of chaos that left everyone on screen dead. Everyone, that is, except for Spelunky Guy, who grabbed what he’d come to buy and tiptoed through the wreckage toward the exit.
Yu released the game on the TIGSource forums on December 21, 2008, and it immediately began enrapturing and infuriating players in equal measure. A month later, he received an email from Jonathan Blow, then experiencing a frankly preposterous wave of scrutiny and success because of his provocative puzzler Braid. He was a fan of Yu’s devilish contraption, he said, and offered to put him in touch with his contacts at Microsoft, whose Xbox Live Arcade had become the forum for wildly improbable indie-game success stories. Yu pinpoints this as one of the moments when he realized the game had a much wider reach than he had imagined. “The fact that he thought it was good enough that he wanted to stick his neck out to vouch for it to Microsoft was a really big deal for me,” he said. “His opinion meant a lot.”
Yu teamed up with an old friend, Andy Hull, to completely remake the game for Xbox, utilizing new art and a suite of funky, atmospheric music from Eirik Suhrke, and adding in a wealth of new enemies and items and interactions. Yu refers to the result as a “fan remake” of his own game, one that took him some four years, a period of time during which the indie ecosystem drifted away from Xbox and toward PC and PlayStation. So Yu brought the game to those platforms, too. With each release it netted a new round of acclaim—perfect scores from Polygon and Eurogamer, even a Game Of The Year nod from the august PC Gamer in 2013, some five years after its original unveiling. Spelunky eventually sold over a million copies in its various guises, its legend growing with each release.
And that’s the thing about Spelunky: Play it long enough and the randomness disappears. By the time you’ve unlocked the shortcuts and can make it from the top of the mines to the bottom of the temple, what you’re doing is no longer cutting a path but interacting with the very algorithms that determine the game’s beauty. It’s a perpetual adventure machine, as if something like Zelda could be drawn anew with each hit of the reset button. It wasn’t the first game to heavily feature randomness, but it may have been the first to make randomness this important and this accessible. Compare Spelunky’s legacy, for example, with the algorithmically generated universe of 2016’s No Man’s Sky, which all but the most ardent players tired of almost instantly. Spelunky, on the other hand, keeps slotting things together in fascinating new ways, a decade later. In his book about the game, Yu splays out Spelunky’s algorithm, detailing the four-by-four grid that each level is built on; the way each line of code, each possibility, ladders into the next. The game designer Darius Kazemi has even replicated the code for curious parties to tinker with online. Even with its secrets laid bare, there remains something about Spelunky, some umami that brings people back again and again, that no amount of monk-like study can explain.
I asked Yu what he thought this might be, and he deferred to the core appeal of the roguelike—that is, maneuvering through a big, complex system and relishing your failures, one run after the next. If “losing is fun,” as the slogan goes for another famous roguelike, Spelunky’s algorithmic guts conspires to make losing the most fun. And anyway, Yu has a point. The roguelike has exploded in Spelunky’s wake, becoming one of the most overly represented styles of games among indie developers, with gothic ones and gun-heavy ones and abstract ones and so on. More interesting are the cases where Spelunky’s free-wheeling hybridization has resulted in stranger-still creations, like Crypt Of The NecroDancer’s rhythm-based adventure, FTL: Faster Than Light’s real-time Star Trek simulation, or The Binding Of Isaac’s putrescent top-down descent into its creator’s id. While Blow set out to start a revolution with Braid, Spelunky made one unintentionally through sheer force of playability. The roguelike family tree now has a dividing line: before Spelunky, and after. Roguelike scholars have even determined a name for those born in Spelunky’s image, “rogue-lites.” The ASCII mentality is pervasive.
The roguelike family tree now has a dividing line: before Spelunky, and after.
And now even Spelunky is spawning its own progeny, threatening another new line in the sand. Spelunky 2 was announced last year as one of the big surprises of the Paris Games Week trade show, proving to headline writers that “there is a god” and “there is good in the world.” “Wave goodbye to your free time,” cried one site. This frenzy is a far cry from the end-of-year forum post that Yu used to quietly announce the first game, but then, he’s been busy, and not just helping design the ambitious games omnibus UFO 50. Spelunky 2’s announcement video suggests that Spelunky Guy’s daughter is now off in search of her dad, and Yu told me much of the inspiration for this comes from his own daughter, who was born in the years after the first game’s release. “Kids are these little balls of pure creative energy,” he says, saying she inspired him to make the game more confidently, imagining her playing it.
Mostly, though, he just feels like returning to the caves. Even after his “fan remake,” even after writing a book on the subject, even now, a decade later and amid a sea of me-too roguelikes, he’s certain there’s more adventure in those depths. “Once you do have an idea that works, it seems like kind of a waste not to use that and build on it,” he says. “I feel like we can do so much more now.”
Next time: QWOP proves games don’t have to be playable to be fun.
Join the discussion...