The clash between systems and narrative comes to a head in The Hit

Dan Vávra, creative director of the large-scale role-playing game Kingdom Come: Deliverance, was recently asked why the player character can’t be an innkeeper in-game, despite the fact that other characters can be. His response:

Well, maybe because the innkeeper has precisely defined places to stand when he puts the beers on the tables, while the player can come from anywhere and it would look ugly if we didn’t have extra animation (extra work) for it, so we’d have to deal with all sorts of new situations. Like for example if the player ignored an order, which the innkeeper never does (extra work), how the player pulls the beers at all (GUI, animation and extra work) and lots of other things, so we immediately dropped the subject—to the great chagrin of the designers and scriptures who came up with the idea.

Perhaps surprisingly, this matter of scale, and of chairs, plagues nearly all game developers. A chair can be willed into the imagination of a reader, or placed in a film via CGI (or an actual chair), but in a game, if the chair isn’t both visible and able to be sat in, it isn’t really a chair. There are no “implied” chairs in the space of the game. When George R. R. Martin describes the Iron Throne with its “barbs along the back, the ribbons of twisted steel, the jagged ends of swords and knives all tangled up and melted,” he doesn’t need to design how the Throne will be sat in. Computers, despite their complex intricacies, are still dumb—they will only do what you tell them to do, so unless you teach a computer everything there is to know about chairs, it won’t know how to fully work with them.

This doesn’t stop players from still trying to use in-game chairs though. A friend who doesn’t play games watched me play The Witcher 2 because I had explained to him how incredible it was, and the first thing he told me to do was to stand on a chair.

“You can’t stand on the chair,” I said.

“Why not?”

“Well you just … can’t.”

It’s the uncanny valley of game design. The more you tell players they can do the more they realize they can’t, but they only realize they can’t because your claims led them to believe they could. Nobody tried to marry the shopkeeper in The Legend of Zelda because there was no suggestion that that system was even present, but in games like Tomodachi Life or Skyrim, ones full or marriage options, limitations pop up all over the place.


I’m speaking with Dan Stubbs, developer of the upcoming game The Hit, and he elaborates a bit more on what this problem has done to game development:

“We’ve kind of skipped ahead and figured out how to make very intricate chair models, whereas we actually forgot to make the things for players to use, which is why we have to constrain players down narrow paths of interaction […] And I’m saying we’ve got it backwards. You need to figure out how the player interacts with the chair and everything else first. It’s the Dwarf Fortress approach rather than the AAA games approach.”

Stubbs bills his current project The Hit as a “multiplayer stealth-action game set in a city.” You’re dropped in with only a pistol, a cell phone, and a target—take them out, win the game. It’s relatively standard fare (though the inclusion of ten thousand non-player characters looks to spice things up), but after talking with him it’s clear he’s using The Hit as a trojan horse to deliver something much more revolutionary.

“The city [in The Hit] is basically going to be a brain that thinks in stories and the stories will be constantly happening around you and you can trip on them at any time, depending on how the Director seeds stories,” he says. He has written at length about this “director” system, but the gist of what he proposes is a system in which the game itself dynamically reacts to the player, as opposed to the player only interacting with the game. Think Solaris (of the Lem, Tarkovsky, or Soderbergh flavor), where the planet places people in the space station based on what it understands of those occupying it, but instead of a space station, the planet places characters in a city based on how a player interacts with it.

“[Dan Stubbs] doesn’t want The Hit to tell good stories, he wants the narrative engine underlying it to be able to produce them.”  

“It’s like the Director is playing a card game with you, with each game or round is a story or an event. It has a bunch of rules for what needs to be present, what cards you need to have in your hand, for the game to begin and to end, which can change based on how you tackle the situation.” Both the players interactions and that of the omniscient Director determine what “cards” you may be working with, your own actions altering your hand and the Director playing off them.

These “cards” can take the form of characters you’ve interacted with, their assigned traits, story beats you’ve already experienced, actual objects, information, and so on. Your interactions with the game are constantly being parsed by the Director, and the Director, based on what Stubbs (or the player) tells it it can and can’t do, decides what happens during your time playing. It’s a constant feedback loop between you and the game.

This Director, or “brain,” exists independently of The Hit—Stubbs sees it as a system that presides over the happenings of The Hit, but defines The Hit itself as the tracking down and killing targets part. The Hit is just Stubbs’ own personal combination of the city and the Director; the system, he hopes, will be flexible enough for others to make their own versions as well. He refers to his developed template as “a narrative typewriter.” He doesn’t want The Hit to tell good stories, he wants the narrative engine underlying it to be able to produce them.

“If you put a chair in the game it will look and function like a chair in-game, and you won’t have to think about it,” he says. “I want to take the onus off the writer to actually have to make all this stuff. That’s what I mean by making a narrative typewriter. I want one person to be able to sit down and write a story in-game and walk away with something playable that they can show someone else.”

The existence of a narrative engine to drive a game is not standard practice—many games instead take a “systems-based approach,” one where the game itself is a collection of systems, and how those systems play off each other becomes the “narrative.” This is how games like Dwarf Fortress or Crusader Kings 2 work, and rely on the player to read into the game’s happenings and construct her own story; there is no separate force playing with the player that works towards its own narrative goals or beats. These games are wholly unsympathetic to the player’s presence inside of them.

Stories from these games become famous, such as Dwarf Fortress’ Bronzemurder, which describes the tale of the downfall and subsequent rise of a fortress in-game after being attacked by the killer winged iguana Oggez Rashas. What the story doesn’t tell you about is the hundreds of other things going on inside Dwarf Fortress that are presented with the same gravity as the events of Bronzemurder. When a dwarf completes the building of a chair, the elicited notification is very similar to the one you are presented with when a dwarf dies. There is no real sense of importance here as far as the game itself is concerned, nor is the game deciding what to do after your chair is built or your dwarf is killed.

“I could go to the supermarket and buy a sandwich, and if I was a great writer I could describe that in a way that would get you totally involved and take you along with me,” Stubbs says. “But that’s a result of the author, not the systems.” Bronzemurder doesn’t prove that Dwarf Fortress is full of good stories, it proves that Tim Denee (Bronzemurder’s illustrator) is good at telling them.


A recent article by Jill Lepore in The New Yorker decries the “disruptive innovation” mentality of Those With Big Ideas. The article describes circular logic of the tech world: “If an established company doesn’t disrupt, it will fail, and if it fails it must be because it didn’t disrupt. When a startup fails, that’s a success, since epidemic failure is a hallmark of disruptive innovation. […] When an established company succeeds, that’s only because it hasn’t yet failed.”

“The Last of Us doesn’t need to die for all of us, in other words, but The Last of Us needs to die for Dan Stubbs.”  

Stubbs, too, has Big Ideas. He tells me that “The Last of Us is a masterpiece, but it’s a masterpiece of an art form that needs to die. We have to stop making games this way. The reason we make games this way is because its the only way we’ve figured out how to make games, and it works.”

Stubbs’ proposed alternative solution, his “narrative typewriter,” is exciting, but the idea that it must exist instead of versus along with feels too ironclad. Surely we all dream of the best of both worlds here—we would gladly welcome a union of both beautiful chairs and fully interactive ones that we can sit on, stand on, and sell to other characters that come by our in-game shop. Stubbs wants to take the some of the first steps towards the latter direction, though, and for that we need somebody who believes in what they are doing and seeks it out with an uncompromising vigor. The Last of Us doesn’t need to die for all of us, in other words, but The Last of Us needs to die for Dan Stubbs.

Beautiful chairs and interactive ones—for now the two will remain separate, as Stubbs explains that he is “not interested in making something complex like Grand Theft Auto 5 or 6 or 7,” so he has to “make everything ridiculously simple,” but, “in doing so I’m making systems that are far more powerful than I realized they could be.” The look of the game speaks to this simplicity as well: characters and objects look almost painted, their look echoing an approximate of a person rather than one with many discernable features.

He says as well that he’s already discovering emergent, complex interactions as a result of his systems, ones that have potential to vary in ways both unexpected and satisfying. I ask him about his plan with both the The Hit and the narrative system, and whether one will exist without the other, and he tells me “I just want to get the game made so I can start testing that and figuring out if it’s right. Whatever I end up with, it’s going to be interesting. ‘Dwarf Fortress as an FPS in a city with thousands of pedestrians’ is an interesting thing in itself, and if that’s all I end up with, I’ll be very, very happy.”


You can find the Kickstarter for The Hit here.