Cracking the Paradox of Branching Narratives

What can you do when too many choices is unsustainable?

Surprisingly enough, the concept of interactive fiction is less than a hundred years old, which is relatively recent in the history of humanity.

Indeed, the idea of a story which hands over the key to its characters’ fate to the reader can be traced back to the 1930’s novel Consider the Consequences! with its 43 possible endings. The Choose Your Own Adventure series massively popularized the genre in the 80s, with more than 250 million books sold. Around the same time, computer games were exploding as well; it was only natural for them to let the player decide the fate of the avatars they play.

However, a problem arises from letting the users shape their experience: the developer has to adjust to the consequences of all possible choices, and if there are many options, it translates into an absurd amount of work. Yet if you only offer a few choices, there isn’t that much interactivity anymore, and it falls flat.

Fortunately, in the past forty years, clever designers have worked on that issue, so we have many examples of tricks & strategies to leverage branching narratives without exploding costs. No need to be a writer to learn something from them: choices are everywhere in games, from quests to progression systems, and we can all use a bit of inspiration.

Humans vs Exponential

A tree of possibilities, where each fork leads to several branches and, again, is exponential growth, something we humans have a hard time intuitively grasping because it’s unlike any law of the observable nature (including for actual trees).

Let me take a quick example: suppose a new virus comes, and an infected person transmits the virus to one another each day. On the first day, there’s one, 2nd there are two, then 4, 8, etc. How many days until everybody in a population of 10 million is infected? Only 24 days. Exponentials grow fast; that’s their very definition.

But here’s where people are usually tricked by their intuition: how many days until just 5% are infected among the population? The answer is 20. While 80% of the time is elapsed, the problem is inexistent for 95% of people. Then suddenly, it becomes a massive issue for everyone.

Why does it matter for game development? The work required to keep up with exponential story possibilities isn’t linear. While it feels utterly manageable initially, developers might be a few choices away from complete overload.

You may say that every option doesn’t have to be unique and that making adaptations might not be a vast deal of work, but it adds up quickly. Writing becomes more complex, debugging & playtesting as well, not to mention the costs of recording & localization.

Even if you can make it work production-wise, offering too many choices can be counterproductive since it will be harder for players to remember what they decided and understand how it influences their narrative. Despite having more choices, they feel less involved. That’s the paradox.

So, to address those issues, you need to be smarter about the quality of your choices rather than focus on the quantity: it takes a great deal of creativity to do so, and you have essentially two options (omg, is this a branching article?): sheep herding & embracing the tree.

Sheep Herding

The first solution to trim the exponential tree is to consider your player as a sheep and yourself (the developer) as the herding dog: the sheep might wander off, but eventually, you ensure they don’t deviate too much and keep going towards the destination you have in mind.

The king of such games (before its complete demise, anyway) was Telltale Games, which reached massive critical & commercial acclaim with their adaptation of The Walking Dead into an episodic choose-your-adventure story. Their recipe has two secret ingredients.

First, they figured they could lie about the importance of choices. One of their key “mechanic” was to display “This person will remember this” in the HUD after you made a decision, even if that wasn’t going to influence anything. Since the interface is the privileged source of info from developers to players, there was no reason to doubt it. This idea was brilliant since it helped highlight crucial moments in the scenario and trigger players' imaginations who’d make connections that didn’t exist (our brain loves to do this).

This leads us to their second “breakthrough”: how they mastered the art of hard course correction. There are countless examples in their games of decisions that feel major yet eventually lead to the same outcome. There’s one example in the graph below: at the end of the fourth episode of The Walking Dead, you can choose to save a character, which seems important. You ignore that the character barely intervenes in the following scene scenes, and he’s dying a bit later anyway.

With this approach, all players will see a maximum number of scenes, optimizing production costs for a lengthy enough playtime—no need to create tons of additional scenes.

However, these corrections were sometimes noticeable (even before replaying). A gripping narrative and surprising twists can help to disguise the smoke and mirrors, but when you start figuring out the trick, it becomes less and less engaging (which explains how quickly sales dropped for Telltale Games despite the massive franchises).

Embracing the Tree

But if you want the player to see all the content, how about doing the opposite and acknowledging the tree in order to encourage the player to explore all possibilities? This design approach made the success of the other big episodic game hit, Life is Strange, and the visual novel Virtue’s Last Reward, among others, games which use time-travelling as a narrative justification to explore the consequences of choices.

Recently I also came across Cartaventura, a series of narrative card games where you explore the life of lesser-known historical figures. It functions like a choose-your-adventure novel, and due to the limited space to write on the card, the designer had to be clever to maximize the number of scenes & dialogues: let’s dig into all the best practices I observed.

Have separate areas

Each Cartaventura makes you visit several places, which fits the exploration theme and is convenient for writing. When they move, the player gets a sense of an evolving story, which easily justifies the frequent “soft reset”. In each new area, the designer introduces new characters and therefore doesn’t have to make them react according to past players’ choices.

Structure the world like a Metroidvania

To complement the point above, the games typically include several ways to reach the same place, just like shortcuts & hidden areas in Metroidvania. It’s a way of sheep herding for sure, but one that feels more rewarding for the player. One way to reinforce this approach is to ensure that the player gets something new or different in the mandatory area, depending on which path they took to get there.

Use items to track choices

As a physical game, Cartaventura can’t hide its logic in the software; the player must know which options are available & which previous choice influences his narrative. To do so, the player gets item cards after making crucial choices, a simple but genius design. Owning an item is more readable for the player, and he can make immediate connections (compared to remembering how a dialogue option influenced a relationship with characters).

Turn the tree into a puzzle game

Items are also heavily used in the series as “keys” to unlock paths, which ignites your desire to replay and discover how you could have obtained that item that opened this door. I wrote extensively about such a “door & key” design in this article.

With just a few items, each Cartaventura can turn its branching tree into a puzzle where your goal is to understand how to discover all the outcomes. The endings are hierarchized: the “bad ending” happens if you make a few mistakes, while the “good ending” can only be unlocked by collecting everything & knowing exactly how to navigate the tree. Once the player knows the narrative, he won’t re-read the cards anyway, so you may as well switch from a pure narrative experience to a strategic puzzle feeling.

If you want to try it out, I recreated the tree of the Oklahoma episode below with all its doors and keys.

Shake the initial conditions

Proposing choices to the player doesn’t mean you must offer all the options immediately. As a designer, you may know that some outcomes are more desirable than others and, therefore, would like to drive players into seeing some branches before others. In Cartaventura, you always unlock new possibilities when finishing the game for the first time, which is a great way to encourage replay with a twist.

Design many keys, not merging rails

It’s a lot of praise for such simple games (and I didn’t even mention the fantastic themes & funny choices), so I’ll balance it with a critique of the Hollywood episode, the worst, in my opinion. It has an interesting idea where you must explore all the branches to find 8 “scenario ideas” in order to unlock the best ending. However, in some instances, like the Chicago arch I recreated below, it tries hard to funnel you towards a single encounter which feels noticeably worse than having several options to collect a given item to move on. There’s a thin line between escaping fate & feeling in control of the puzzle.

Conclusion

So, which is better in the end? Sheep herding or embracing the tree? The designer in me leans towards the second, obviously, I admire when a big constraint is turned into a strength through creative design choices.

However, I also know we must always keep in mind the unexpressed needs & desires of players: some will enjoy choices when they’re a flavour on top of a conventional story but won’t like it if it turns a laidback experience into a head-scratching puzzle.

If you liked the topic of this article, I’m sure you’d also enjoy this one: