Stone Soup

When thinking about storytelling in gaming, I often think of the folk story about stone soup.

The designer provides the pot, the fire, the water, and drops in the stone to get it started. But the player is also on the hook for bringing their own ingredients and making this into a soup worth eating.

The medium of boardgames (and to a lesser extent, videogames) necessitates this. A book or a movie is a curated experience; there's typically no direct input from the reader or watcher. But a game, by definition, will have inputs and interactions from the players.

A serviceable emergent narrative can happen if the events the player experiences interact and connect in realistic or at least believable ways. A bit of story self-contained in the flavor text of a card is not enough, unless that card connects in some way to some other game system or narrative beat.

That interaction needs to be carefully constructed. In a movie, events can be highly detailed and specific because they’re presented to the viewer in a considered order, and one can assume that anyone watching the 10th scene also saw the 2nd and 6th scenes. In a game, where these events might be randomly encountered, modified by player choice, or even skipped entirely, too much detail or dependency on other events will mean the puzzle pieces don’t fit together and the narrative fails.

Not too many games fail in this way, as typically game designers don’t attempt to tell a novel-style story through games. Events and encounters generally only get into that level of detail when the design mixes with other mediums, such as Bandersnatch or a Dave Morris gamebook.

But some games do limit their narrative options by being too specific. For example, the injury results in Kingdom Death: Monster list precisely what happens to your character. There are a lot of injuries listed, so many options are covered, but nothing outside of those lists can ever happen to your character. And it starts to get a little obvious (and eyeroll-y) when multiple characters end a combat with “destroyed genitals.”

Contrast that to the injury rolls in games like Hunters or B-17. These types of games typically have only three injury options: light wound, heavy wound, and KIA. Though not as evocative, these options will never result in narrative dissonance, as they are general enough to apply to any injury-causing event; the puzzle piece will never not fit. The designer threw in the stone (“heavy injury”), but the player is responsible for the flavor (“Bf-109 attack, hit in the shoulder”).

Another way to ensure the fit of randomly-selected puzzle pieces is to boil descriptions down to keywords. In Legacy of Dragonholt, for example, a character is simply a collection of traits. Your character may have “Dueling,” “Agility,” and “Reasoning,” which creates a particular idea of a character and would feel different from a character with “Military,” “Archery,” and “Endurance.” But none of these define a character so acutely that the same general narrative couldn’t apply to both.

But the best example of this, by a long shot, is Gloom of Kilforth.

In GoK, players have a personal plot that they are following, with some flavor text background and details. As a made-up example, a player may be told that the local townsfolk have hired them to kill a monster hiding in a mountain lair. To complete this card, though, the player is given a set of keywords they need to collect: person, monster, mountain. As they wander the world of Kilforth, randomly drawing encounter cards, they need to keep a lookout for these keywords.

The system gives the player a bit of direction, as the “person” keyword may show up more in Plains locations, and the “mountain” keyword is obviously in Mountain spots. But it allows for the story to play out on its own and emerge from the cards the player randomly encounters and solves or defeats. Through player-driven exploration, they may create a story where a local baker hired them to kill a werewolf hiding out in an abandoned mine. It makes narrative sense, and even fits their predefined plot, but the keyword system makes it possible in a game.

There are many approaches to creating an emergent narrative. Though Kingdom Death’s injuries fail for me a bit, it more than makes up for it in other ways. But to me, the best emergent narratives in games give me just enough structure for me to hang a story on, but keep it nonspecific enough that I can fill in the details myself to avoid narrative dissonance, while giving me enough agency to make the story my own.

I like being told a story, but the qualities of the medium of gaming put it in a unique position to allow the player to be an author. If the goal of the game is to create an emergent story for the player, it has to account for the fact that the designer does not have complete control over how this is going to go. They only get to start with the stone.