Keeping it all together: Coherent narratives

(this essay has mild spoilers for 9:05)

I’ve been thinking about writing an article on this topic for a while, but I wanted to preface it with a warning: I’m writing from the perspective of a player, not an author. So I want to present some ideas and issues, but I won’t present many solutions.

Narrative coherence, to me, means ‘how much does everything work together in one big story’? A game is coherent if everything in the game builds on each other. A game is not coherent if it goes in a hundred directions at once.

==Facets of a diamond, not branches of a tree==

There are a couple of ways that a game can be incoherent. One example is the ‘Time Cave’ structure, where the game branches exponentially into wildly different endings. This is the style most popular with kids CYOA books. A similar style is an ‘early branching’ game where the game splits into one of a few tracks at the beginning and stays on it the entire time without merging.

Both of these structures suffer from perceived lack of quality compared to actual quality. An author of such a game will spends hours and hours on the game, but because each play through only goes through one branch, the player experiences only a fraction of the material prepared by the author. This can be partially solved by replays, but because different playthroughs are so different, the final ‘feel’ is more like playing an anthology of unrelated games. In reviews of such games, players generally point out one or two parts they really liked and ignore the rest.

What is more successful in turning author work into player enjoyment is having all branches of a game relate to each other. The most extreme example is Aisle. Aisle is a strongly branching one-move game, but all endings relate to each other. They paint a picture of the main character that is fairly consistent. Even when they factually disagree with each other, they seem more like facets of a diamond instead of branches on a tree. Other branching games that act like facets instead of branches include Galatea, Heroes, and Cape.

==Oil and water don’t mix==
Another way that games can be incoherent is mashing together unrelated elements. These can include big changes in tone or in pacing.

One game that I recently played started as a campy space comedy, with humorous, bungling attempts to get into space. In space, it became a gory horror game, but the campiness continued, and it was really dissonant.

Another game had pacing issues. The game was almost entirely a wandering, conversational game with puzzles solved by bringing items from one area to another. The game then came to a massive, screeching halt because the author had decided to make the largest combinatorial puzzle in the history of IF in one location. Anyone who worked through the game without a walkthrough would have to spend hours or days on this puzzle, forgetting all of the plotting and pacing before. Some have referred to this dissonance before as “soup cans in the pantry”.

Now, this isn’t to say that you can’t mix up the tone or pacing of a game. But like mayonnaise, you need something to bind them together. Without some sort of connection, the ingredients fall apart like unshaken italian dressing.

A good example of a game that manages this is Delightful Wallpaper. This game is split into two very different halves. The first is pure puzzle, while the second focuses on narrative. However, the game binds the two by having a very strong atmosphere, a protagonist for whom both challenges seem fitting, and mirrored mechanics. In fact, the first puzzle directly prepares you for the second puzzle by acquainting you with every corner of the building.

Another game that manages to blend changes in genre is 9:05. In this game, there is a major shift in the genre, but again, everything before the shift helps to understand what is happening after the shift. Instead of throwing away what came before, it is carefully blended together with what came after.


Based on my own feelings about reviews and reading the reviews of others, I can safely say that games with coherent narratives are a pleasure to play, and result in less work for the author. If every new discovery and puzzle in a game works well together, like facets of a gem or like a blended sauce, then the game is better set for success.

9:05 spoilers.


I played 9:05 in my usual exhaustive way and experienced this shift during the first handful of moves.[/spoiler]

Hah! Yes, you would have, I can imagine just how that came about! You were unlucky enough to have been lucky. :slight_smile:

I appreciate these geometric metaphors for story coherence. I’ll free-associate a bit with some thoughts on this.

Even uninteractive stories have a sort of cause-and-effect graph, describing their concurrent structure. People can complain that a later chapter of a story is disconnected from the themes of an earlier chapter, or that two concurrent plot lines never seem to cross paths with each other, just as they can complain about the disjointness of two alternative plot lines in a cave of time. Once Upon a Time is a show that usually interleaves two time periods or universes per season, with the aim of showing thematic similarity and dissimilarity between the interleaved parts. Star Wars is a story that tries to echo or rhyme. So stories like these sort of loop around and protein-fold upon themselves, suggesting vague bonds of coherence that can occasionally destabilize and find new equilibria as the full structure reveals itself. This dynamic action could potentially be a powerful form of plot twist, or an interesting mechanical structure for other media (fanfics, sequels, lit reviews) to build upon.

For the structure to be able to destabilize and find new equilibria, the story can’t be a single rigid gemstone. Parts of it have to be apart from each other, so that they have the freedom to move independently. Rigid structures are great; they’re resilient to imperfect data transfer, and they offer a sense of completeness. If for some reason we have to do all our thinking in some framework, we can appreciate a framework that’s rigid, not slipping out from under us. And even if a story is more dynamic, it can be useful for it to have resilient bones and skin. But stories can move, or marinade, or ripple…

If a story leans on brittle hinges and falls apart at close inspection, it might be difficult to find much use for it. Some stories may fall apart by design or by compromise, like fireworks and decorative candles. A lot of stories are going to have honestly flimsy construction that their authors would have approached differently if they had more experience or different tools. I really appreciate the stories whose structures are enabling or inspiring to the future authors who read them.

That’s interesting! You know, a lot of my favorite stories do have separate moving parts, often achieved by parallel stories set in the past (Anchorhead, Curses!, Theatre, Spider and Web). I think youre right about the story needing to not be rigid. These are really good thoughts!

One thing they told us in fiction class was that every element of your story should be set up in the very first paragraph. I am not sure this is an achievable goal in practical terms, I’m not even really sure what this means in practical terms, but it sounds good, right?

Easy to test: check it against the best opening paragraph in fiction.

…Check and check.