Classic adventure games, for all their archaisms, are about doing, not watching. In a medium frequently obsessed with the ability to replicate the tight focus and action climaxes of cinema, there’s something radical, even subversive about a genre that gives you so much to work with, so many ways to interact. By Joe Skrebels
So many ways to interact, in fact, that the thrust of the adventure genre – that videogame spaces can be toyed with as well as dashed through, that conversation can be a pleasure and reward in itself, that there are moments to miss – has become commonly overlooked, jettisoned in pursuit of more immediate pleasures as studios eschew increasingly expensive redundancy. And so modern games reductively trade story agency for commands to “Hold X to pay respects” at a funeral. Narratives have but one path. You read from the script, with the voice recorded by an actor, and express yourself in how the corpses pile up. This is not all gaming, not by any means – RPGs such as Dragon Age: Inquisition, The Witcher and Wasteland 2 provide one type of counterweight, while Thimbleweed Park, Broken Age and the remake of Grim Fandango are seeking to offer another – but it is a dominant trend.
Quietly, almost imperceptibly, however, the past few years have seen the most archaic end of the adventure genre, the text-based adventure game, slide towards relevancy once more. Appropriately enough, developers have been using an indispensable part of the modern inventory to make this so: the smartphone.
Jon Ingold, co-founder of Inkle, has found a position at the intersection of one of gaming’s earliest genres and one of its newest platforms. Inkle’s work on games such as Sorcery and 80 Days has set a high-water mark for textcentric mobile games, and the Inklewriter engine, which helps simplify the creation of branching stories, has been used as the narrative backbone in Stoic’s The Banner Saga (ported to iOS from PC) and Mi Clos’ Out There. Why return to reading, in these days of prerendered spectacle? “We want games with real characters who you can relate to and interact with,” Ingold says. “We want interfaces that give the player a breadth of interesting verbs and actions. We want players to alter and shape their protagonist in the story, and have the game listen and respond to that intelligently. We want the narrative aspect of a game to turn from an apology into a marvel.”
But text adventures have had to change to suit devices where typing can become tiresome, and an audience that wants the screen to take on a little of the world-building work once left to the imagination. Today’s text-heavy adventure games come in the shape of digital Choose Your Own Adventure books (Sorcery); expressive, if more linear, prose mysteries (Device 6, Blackbar); and hybrids of interactive fiction with other genres (Out There, for example, intersperses its panes of text with resource management). They aren’t text adventures in the classic sense, but share a common ancestor, and not simply because they all put paragraphs of prose on the screen.
“The text adventure is 30 years old, and was very much a product of its time and technology”, Ingold says. “While there are people making excellent text adventures now, they’ll never be able to reach a wider market, because these games just don’t make sense in a modern context: they’re hard to learn, slow, confusing, unfamiliar, and hugely disempowering to players. We’ve been experimenting with ways to change the interface and introduce other elements to maintain all that’s compelling about interactive stories while removing the obstacles that stop people from playing.”
That experimentation has led to packing aspects of more familiar games around a core text structure. Take Inkle’s 80 Days: while almost a hypertext novel at times, it introduces trading, finance management and carryingcapacity limitations to offer each round-the-world trip a sense of peril and consequence. Swedish studio Simogo’s recent oeuvre, meanwhile, is such a peerless aggregation of point-and-click, prose-driven storytelling and presentation that it almost defies classification. But while each of its most recent games, Year Walk, Device 6 and The Sailor’s Dream, use text as a defining element of their storytelling (Year Walk through its metafictional companion app), they are light on player choice.
Simogo co-founder Simon Flesser certainly sees text as a facet of his design, rather than the cornerstone. It results in players chasing his text, as opposed to choosing which parts they want to see, as much a reward as it is a means to earning one. “I don’t see our games as much as text-based as they are multimedia,” he explains, “mixing visual art, text, sounds and music in very deliberate ways.” Simogo’s games use reading to accompany a wider goal, and Device 6 in particular tinkers with the physical form of text to create puzzles, even a kind of level design, as multiple sentences stretch in different directions.
Simogo’s goal is to see “how text can be interacted with, what it can or can’t do by being digital as opposed to being on a paper, how text can be merged with other types of media digitally to create something that is unique to it being digital”. There’s a sense that these games are positioned almost to re-engineer how we see reading as a whole.
Meanwhile, it’s easy to put the success of more traditional interactive fiction, such as 80 Days (which retools Jules Verne’s Victorian travel fantasy and characters to allow players to live up to Phileas Fogg’s wager), down to the growing interest in digital reading. While literary games make sense on devices used as eReaders, Ingold sees the truth as being more nuanced than merely the power of association. “I don’t think eBooks are particularly what’s important here,” he says. “People have been reading the whole time – on websites, Twitter, email, Tumblr, blogs. Reading never went away, and generally when people say they don’t like reading, what they really mean is that they don’t like reading books. So our idea was to create a reading loop in which you’re always making progress, always having to act, always putting some information in, and always being asked your opinion and listened to. That’s the same compulsive reading loop that Twitter uses – refresh, read, refresh – except in our games there’s always something new and interesting to read.”
These new textual adventure games offer a space for unexpected storytelling, too. Meg Jayanth, the writer behind 80 Days, sees in this as much a sound business model as an artistic opportunity. “There’s space to write something that feels genuinely new to the player. Text is cheap, which means that people can make games that are niche, weird and strange, and find an audience. It’s indie cinema versus the summer blockbuster; without the pressure to sell millions of copies, it’s easier for people to take risks and experiment.”
And that might be key to why mobile formats have become a breeding ground for these experiments: textual game creators have found a space to tinker, and to be paid. The community around PC writing/design tool Twine has spawned hundreds of games, showing the demand for interactive fiction on non-mobile platforms, but only a minuscule percentage of that effort has resulted in paidfor content able to support its creators.
Design and art style plays a vital role in making these games attractive for app store window shoppers and allows them to slot into the expectations of a generation that grew up after the advent of the graphics card. Both Ingold and Flesser, despite their differing approaches to design, are united on the importance of the visuals that surround their text. “Certainly, a massive proportion of our success is due to our art and design,” Ingold says, “partly because it got Apple on board, and partly because it sends a message to players that they should trust us and that the game is going to be worth getting into.”
“I think with any media you consume, it’s always a nice sensation if it has a visual design that sits well with, and empowers, the narrative or feeling,” Flesser says. Even a genre built on simple text needs to look good. Sorcery is set across a magnificent, subtly 3D map interface, while The Sailor’s Dream presents song lyrics gently washing out of bottles.
But just as mobile consumers can be convinced to pay for textual adventures where PC users often can’t, so too do these games struggle to find footholds on home consoles. Handhelds, however, have long been a bastion of the visual novel format (the Phoenix Wright series, Corpse Party, and Nine Hours, Nine Persons, Nine Doors), propped up by ongoing Japanese interest in these forms of storytelling. So why not PS4 or Xbox One? As Flesser points out, this is a form factor problem as much as anything: “The physical design of handhelds makes them more natural for reading. Playing a text adventure on a home console connected to a TV would feel a bit like reading a book on a billboard.”
But that isn’t to say that modern advancements of text adventure themes are wholly limited to smaller screens: the heritage of the genre is also evident in the RPGs and retro projects that do continue to hold a place in console libraries. BioWare’s Mass Effect and Dragon Age games have become increasingly adept at offering player expression through sexual preference and ideological standpoints while still presenting enough directed content to make development manageable. Middle-earth: Shadow Of Mordor’s Nemesis system generates characters and evolves them based on your actions in the game intelligently enough that Ken Levine called it the world’s first “open narrative”, and “chess meets Hamlet”. Text adventures’ reactive decisions and consequences are shot through these apparently unconnected games, and Ingold sees that as the future of the genre: “We think the idea of evolving the text adventure is a decent starting point but is ultimately upside down. We should be making games that beg and steal all the good bits of text adventures – the player freedom, the adaptability, the tight action/reaction pacing, the beauty, ambiguity and expressiveness of language – and place them in a context where they can really shine.”
But there’s a cost associated with any deviation from the strictly linear path. The sheer variety of actions and reactions required by the archetypal console game already costs thousands of man hours and millions of dollars, so introducing baroquely complex storylines to the mix would only serve to make for an industry even more precariously balanced between success and bankruptcy.
Michael Cook, creator of game-making AI Angelina, believes that procedural generation could one day be a solution. To improve the application of text adventures’ ideas to mainstream games, we need to make text more efficient to make. “AI can improve almost every aspect of a text adventure,” Cook says. “One of the nice things about text adventures is that an AI system doesn’t need to show the player what’s going on; just give enough details for the player to fill in the blanks. With a few words, an AI character could play a piano mournfully, or chuckle wistfully, or slip on a banana peel. The challenges are still there in understanding and generating language, but there are fewer places to make mistakes, which lets us focus on the language problem more closely.”
Generators of a sort are already common within modern interactive fiction. Emily Short’s Blood & Laurels uses Versu, an AI text engine, to power dynamic NPCs within a familiar Choose Your Own Adventure template, and even Meg Jayanth used a generator to dynamically fit her prose together for the exquisitely authored 80 Days. But this is as far as Ingold can see it going: “Actual storytelling ‘master AIs’ are probably going to be worthless. Stories are all about surprising and delighting your reader, and it’s really hard for people to come up with interesting stories. Hell, if even the writers on Stargate can’t turn out a really good story every week, how would an AI manage it? Generative narrative design is a problem for computer scientists. It’s a fun and rich problem, but it’ll never replace the wit and ingenuity of an author, and I will eat my fountain pen if I’m proved wrong.”
Cook – a computer scientist – disagrees: “There’s a stigma of software being impersonal, cold, and unable to understand the nuances of creativity. But there’s really no reason they can’t be just as good at collaborating on a game design as humans are. Everyone’s taste will vary, but I don’t think I would mind if I discovered that Zork was written by a computer, for example. We can write and direct AI to take over whatever tasks we want to.”
In fact, Cook’s own procedural generation game jam, #PROCJAM, spawned what could be seen as an early AI-written text adventure last year. Tom Coxon’s Dreamer Of Electric Sheep draws an entire game out of ConceptNet, a database of semantic associations designed to be readable by AI. The result is an eerie, appropriately dreamlike adventure that – barring a single humanauthored horror element – builds itself as you play and react to it. A pounding nightclub might lead into a deserted forest; people can be walked into in a literal sense, whereupon you find they contain a ticket booth for a reason only the AI can understand.
Without a human author, it’s garbled, but it operates with a strange logic. It can’t match the vital contextual knowledge or experience of any human, much less the touch of Jayanth, but it can create surprises indefinitely. And it was designed precisely because Coxon sees a need for text generation in mainstream games. “I think that there is a longterm trend towards making larger and larger games,” he says. “If this longterm trend is to continue, developers have to find ways to automatically generate a lot of the content. I find that the memory of a good videogame storyline lasts a lot longer than the memory of fun gameplay, so I think that future large-scale games are going to have procedural or emergent narrative of some kind. That is why I wanted to experiment with elements of AI in a small text adventure game.”
It’s very much a first step, but the goal is an engine that can automatically and intelligently create storylines without the constant supervision of human writers. This could embrace the tenets of the text adventure, offering personality and interactivity, without overburdening designers. It could also work in parallel with an authored main storyline, the linear goal we crave from our stories, but create emergent moments that belong only to the player that encounters them. At base level, textual adventures could capture the essence of a very different kind of game, an essence developers have been chasing since the ’70s. “Dungeons & Dragons,” Cook says, “sold the dream that players were the ones shaping the story. For decades we haven’t been able to replicate that in videogames, because it’s a really complex thing to do: improvisation, storytelling, game design all wrapped up in one big role. But we’re ready to have a go at it. Shadow Of Mordor really captured people’s imaginations, but it’s small-scale compared to what we could make games do one day. It won’t happen in the next year, or five years. I don’t know how long it could take. But each step towards that goal will be exciting in and of itself.”
They may disagree on the methods used, but Cook and Ingold are ultimately searching for the same thing, the kind of individuality precluded by asset pipelines and auteurs, the sense that we haven’t just entered a world by playing games, but that we can touch and affect those worlds in return.
Quite apart from execution, platform or philosophy, the future of the text adventure and its contribution to modern gaming still lies in the vast potential of those early verb sheets. We want to converse, combine, tinker, collect and inspect as well as fight in our games, to be transported to a reactive world that we can shape with our actions as we in turn are shaped by being there. Perhaps the future holds adventures that no one else can or will experience, spontaneously generated by AI. But even if that promise isn’t kept, there are increasing riches of well-written, well-produced titles that give us a verb set and the chance to partake in varied worlds far beyond the grasp of even the most powerful 3D engine.