Skip to: main content or site footer.

The Problem with Branching

[This is a guest post by David Dunham of A Sharp. David has been creating digital games since he was in 7th grade, as well as a variety of other software. He's written about revising his most ambitious game at .King of Dragon Pass was an important early influence on Echo Bazaar and our other work: so we thought it would be nice for him to cross-post about it here.]

Humans seem to be hardwired to play and to appreciate stories. So why do so few digital games combine the two?

Many, if not most, games do contain stories. But the two seldom meet. The classic issue is that play grinds to a halt and you get to see a cut scene advance the story. (Sometimes it’s the other way around, and game play keeps interrupting a story.)

Nothing says that story and game play can’t be integrated. The early days of digital gaming were different, and in fact a game from the 1970s, Adventure, gave its name to a whole genre of story games. These were originally purely text, though eventually they gained a point-and-click interface. The most common game mechanic is puzzle solving. For whatever reason, this style of game hasn’t had much of a commercial market (at least in the US) for almost 20 years. There’s still a lively interactive fiction scene (I follow it at a distance via but not a lot of commercial activity.

One reason games tend not to have integrated stories is that games are essentially about choices. Story choices typically end up with narrative branches. Do you insult the daughter of the rival house, or try to woo her? Romeo and Juliet: the Game would obviously have two very different stories. Perhaps the two options would somehow combine back into a single point, but the more decisions, the more branches. In this style of game, you need to create a lot of content that a player will never see. And unlike the early days, commercial games need art as well as text, which can be a big expense. So branches cost both in assets and testing (you may need to test every possible permutation through the branches). The rise in art production cost is probably one reason for the commercial demise of the adventure genre.

But people respond to stories, and designers kept trying to figure out how to incorporate them into game play. One game that stood out to me was Castles, which was essentially a resource management game that occasionally slipped in small story situations. Your workers are terrified by rumors of a werewolf. Do you divert warriors to hunt it, send for a priest, or just bribe them to keep working? I don’t recall that these had a big impact on the game, and they may have only been interactive cut scenes which used text. But for me they were the high part of the game. And they led me to create one of only two games I’m aware of that are narrative-driven (outside the adventure genre).

King of Dragon Pass - - essentially has an underlying resource management game which serves as the skeleton that supports the stories. You play a clan of perhaps a thousand people, interacting with a fantasy world (the detailed, mythic world of Glorantha). Like typical strategy games (as they’re usually called), you have a number of resources to manage, and a technology track (in this case magic) to advance. Play consists of alternating resource-related actions, and responding to story scenes. These consist of an illustrated situation, usually with five choices. These may lead to secondary choices, but still within the same situation, until it’s resolved. Resolution may have specific later consequences, but typically influences the economic model of the resource management game. Parts of the story and setting are actually revealed through your advisors, individuals you pick to sit on the clan council. While you can lose the game through poor resource management (or bad luck), you can only win by story choices, so I think it’s reasonable to say the stories are key. (Strategy games are traditionally won by conquering your neighbors).

The second narrative game is quite different. Echo Bazaar is essentially a story game built on the chassis of what on computer is called a role-playing game. You play a single character interacting with a fantasy world (the charming alternate Victorian city Fallen London). Like typical RPGs, your character has a number of statistics which your goal is to increase. Uniquely, the way you do that is by picking a story let (either one randomly available or one which your statistics allows you to play), and making a story choice. The consequence is a change in one or more statistic. It’s essentially a pure text game (though there are small, somewhat generic illustrations). The game is played online with a browser, and by design limits how much you can play in a day. (You can buy additional turns, as well as new story options, with real money.)

Interestingly, the two games have many design elements in common, probably because both wanted to be narratively sophisticated without running into the branching pitfall. Both track a number of unusual statistics designed to support storytelling. King of Dragon Pass tracks relations with other clans in great detail, while Echo Bazaar tracks your relation with individuals and organizations. Both games make small stories the basic unit: King of Dragon Pass’s scenes, and Echo Bazaar’s storylets. In both games, a larger story emerges depending on the choices you make in dozens of smaller ones, even if they seem at first unrelated. And the smaller stories are designed to be reused. In King of Dragon Pass, the context is almost always different, and some factors are random. In Echo Bazaar, new options may be available, or you may want to experiment with different outcome. Both games also reuse art to keep costs down. And neither game shies away from having true branches for narrative effect (though they’re rare). Occasionally longer stories are simply broken into pieces which are revealed only after much game play. Both games also tend to have stories which aren’t entirely black or white.

The two games have actually influenced each other. Alexis Kennedy has said that King of Dragon Pass was an influence on Echo Bazaar. And after playing Echo Bazaar, I realized that you could have a worthy narrative without as much investment in complexity (the new scenes in King of Dragon Pass 2.0 tend to be much simpler than the original ones).

Both games are fantasy games. This may simply be a coincidence, since fantasy is a staple of digital games. But both have richly detailed fantasy settings that get revealed as you play.

Since both games use small episodes, they’re both easy to play if you have only small amounts of time (your browser is probably always open, and your iPhone is always in your pocket). And each player will construct their own overarching narrative by how they play through the episodes. At the same time, the elements are common, so players can discuss them. “What did you do with the Cheesemonger?” “Did Kallyr get to be queen?”

The games differ in a few ways that aren’t related to their underlying genres. Most obvious is that Echo Bazaar exposes all statistics, and King of Dragon Pass is opaque. Arguably the game statistics break the fantasy, and the opacity can be maddening to players trying to learn how the game should be played. King of Dragon Pass also has a lot more random factors. I think this is driven in large part by their monetization models. Echo Bazaar is a free game, but makes money by selling specific game play. You had better have a really good idea of what you’re going to get before you shell out money, so everything is explicit. King of Dragon Pass was originally sold in a box and now on the App Store, so you’ve already paid all you’re going to. So the design goal of showing only values that would be apparent to game characters in an uncertain world makes sense. (The game compensates by explaining things in a detailed manual.)

The most important similarity between these two games is that both succeed at being true narrative games. Hopefully they can inspire future games, and I’ll get to play more stories!