Two Problems for Branching Stories
This image isn’t really relevant, but I love it. From  here .

This image isn’t really relevant, but I love it. From here.

I wrote down two ideas recently: The 100% Itch about the contradictions involved in completing every option in a branching story and Choices Unchosen about the unavoidable impact of presenting multiple options, even when only one is chosen.

Last year I wrote Two Problems for Interactive Stories, which provided a lot of fuel for thinking about stories in games and has been pretty well read since. As it was so useful to have those ideas ready to link to, I want to do the same with these two, for future reference, so I’ll summarise them here, and rephrase them as more general “problems”.


The 100% itch

In The 100% Itch I observed a tendency to want to complete all branches of a story. Maybe that’s just an old habit from making games last as children, but it’s a compulsion that’s acknowledged — if not encouraged — by some modern conventions. For instance, it’s common for IF to number endings and tell you which you reached, or for achievements or trophies to check off which endings you’ve found.

The Uncle Who Works for Nintendo   (a beautiful game) lists all the endings when you reach one, and prompts you to go back to find another.

The Uncle Who Works for Nintendo (a beautiful game) lists all the endings when you reach one, and prompts you to go back to find another.

The “problem” this raises for creators of branching story games is not some unsolvable paradox, it’s just a question that needs an answer: do you acknowledge the fact that the audience is re-reading the story, or ignore it?

The reason that question needs answering is that there’s a contradiction involved in interpreting a branching story all at once. If we go all the way down branch A, then reverse that and go down branch B, we have two versions of the story to contemplate. Although we’ve undone the events of branch A, we can’t unread them. That’s fine — it’s a familiar bit of gamer doublethink — but that mental exercise of the audience shouldn’t be ignored: it has an impact.

For example, if branch A was a safe choice leading to a simple and bland happy ending then you might be quite unmoved by it. However, if branch B represented a risk that ultimately wound up in some horrific tragedy, we might respond very differently — with a greater satisfaction, perhaps — to ending A after having played B. A boring happy ending becomes a safe relieve from terrible (and in some sense present) danger. Even the act of prioritising branches of the story in your mind in order to treat one as the actual one you chose requires you to consider all branches, even those that you deprioritise.

The point is that there is no way to offer story branches as if they are the only isolated route through the game if the player is expected to replay and travel multiple routes. Maybe an author could decide to write for an audience that only plays once, leaving replayers to deal with the contradictory branches themselves. However, there’s certainly also room to write for an audience of replayers and create endings that draw on each other in some way.

This is another way to phrase the question, then: How many endings do you expect your audience to play? The answer to that will in turn raise UX design questions about signposting that expectation. (Hence ending lists and replay links in IF.)

So when I say “problem”, I mean that this is a question, an exercise, a point for consideration. It’s not a flaw with the concept of branching stories: it’s another tool for artistic expression in games.


Choices unchosen

In the post before that one, I wrote about dialogue options and the impossibility of completely disregarding the options you don’t choose, pointing out that you can’t unread things.

The point about unchosen choices is that they carry just as much weight when they’re brief options presented on the screen but never selected as when they’re whole story branches that you didn’t play this time. Apart from the 100% itch to see every optional detail, there may be details given just in presenting the options which already raise little contradictions. Again, this is no bad thing — it’s an opportunity.

The difference here is that whereas the 100% itch raises a question of which audience you’re writing for — completionist replayers or once-through fatalists — unchosen choices leave no room for that distinction. The options are on the screen and must be read if they’re to be chosen, so your audience must be assumed to be reading them all.

In this case the “problem” is not so much a question as a restriction. Player options must be presented with awareness that they’ll all be read. For example, if there’s a heartbreaking emotional scene in a game and you’re given three empathic and sensitive things to say and one mean but hilarious one, that last one’s going to change the mood just by being on the screen. If you’ve broken a smile at it then the game has a different tone now, regardless of the option you choose.

My main example was Monkey Island, where every dialogue branch was a chance “to tell four jokes at once” (notes Ron Gilbert). It’s not as though you only laugh at the one you click — the unchosen options still do some work and can’t just be ignored. I also considered examples from Gunpoint and Creatures Such As We, which I think are rather subtler.


Solving the problems

I needed to get these two points clear because I’m playing something now that has approached both of them really creatively so I want to write about it. I’ll dig into that soon and hopefully more case studies will follow.



Like this? Subscribe by RSS.


Photographer and writer covering Tokyo arcade life – the videogames, the metropolis and the people