Choice, Complicity, Consequence

Hello? Hello, welcome! Uh, I have two things, many which you’ve heard before, to say. One is, please turn off your mobile phones, I’m very nervous and if I hear a ringtone I’ll leap in the air like a startled gazelle, and probably spill something. The second thing is please submit the evaluation form at the end of this talk, this will be your only opportunity today to exercise your agency as a player character, so take advantage of it.

I’m Alexis Kennedy, I run a little indie studio in London called Failbetter Games, I’m here to talk about choice, complicity, and consequence as they relate to choice-based, non-linear, interactive narrative, and other things. Who here does not hate cocktails? [pause] Thank god, okay.

So we do these things. I hope you’ve heard of at least one of them, if you haven’t I don’t know why you’re here but I’ll do my best to make your time worthwhile.

Everything we do, apart from cannibalism, the common factor is non-linear narrative with hand-crafted prose. By that I mean, if you think of interactive narrative, choice-based narrative specifically as a continuum, with really procedural stuff like Dwarf Fortress or Ice-Bound or something at one end, really scripted stuff like Twine or traditional choose-your-own-adventure, very branching stuff at the other end, we’re about 75% of the way along. Many of you won’t be, many of you will do slightly different things or very different things, and I don’t want to tell you what the best way to do interactive narrative is, because I’m not a maniac. What I hope to do is present you with some tools and techniques that may be useful in building choice and consequence and complicity which are emotionally-resonant and interesting.

So some of this stuff will be more relevant than others. I’m going to try my damnedest to avoid saying this is the right way to do things, I am going to say these are techniques that we found useful or ways of thinking about things that we found useful over the last seven years of making interactive games.

When I talk about choice, complicity, and consequence, what I mean, choice and consequence are really familiar, but just to confirm terms, choice is the experience of making a decision, consequence is what happens after, and complicity, coincidentally, is what happens in-between. It’s the emotional experience you have as you’re making a choice. And all of these are places where you can insert emotions into your player. And that’s ultimately what you want to do.

Here’s a choice from our game Sunless Sea. Sunless Sea is a game about exploring a vast, underground ocean, and the themes of it are all about light and darkness, loneliness, exploration, coming home. So that’s the context of this choice. You decide whether or not to get it on with one of your officers, which has consequences.

If you do get it on with this officer, there’s a narrative consequence, so you have a thing with them, that’s going to stay relevant, there’s a gameplay consequence, which is that the stat which determines one of the end conditions of the game, Terror, drops because you found some comfort in the darkness. And there is a deferred consequence later. We tell you something else is going to happen, we don’t tell you exactly what, but when you get home your spouse might have something to say about it. We like to tell people about deferred consequences, a lot of other narrative designers don’t. I’ll talk about why we do later, and you can decide if I’m right or not.

Complicity, I think it’s fair to say, that players enjoy sex and romance, so much of the fun of making this decision, to go with the Tireless Mechanic, is that you get a tiny bit of titillating writing.

So that’s choice, complicity, and consequence, but, I only have 27 minutes left, and I have a preposterous number of slides to get through, so I’m going to employ the power of metaphor, and I’m going to do this by talking about two cocktails.

As many of you will know, the ‘Blood and Sand’ is a classic whiskey-based cocktail, it’s quite sweet. It was created to commemorate the 1922 film with Valentino and about bull-fighting.

The ‘Smoke and Mirrors’, I’ve seen a couple of varying recipes for, but it’s a variation on Blood and Sand. The scotch is replaced with mezcal, and the orange juice is replaced with grapefruit juice. The name commemorates Tezcatlipoca, who is the Aztec god of, I believe, night, jaguars, sorcery, storms, obsidian, and being a badass. I would like to ask you to raise your hand if the idea of the Blood and Sand appeals to you more.

[pause]

I would like you to raise your hand if the idea of the Smoke and Mirrors appeals to you more.

[pause]

That is a narrow win for the Smoke and Mirrors, and coincidentally I have here the ingredients for a Smoke and Mirrors cocktail.

[Alexis proceeds to arrange a full cocktail set on the lecturn]

So the reason I’m doing this, apart from to sneak a drink onstage — actually I’ll leave it there so that you can photograph it, but I may drink from it, I am quite jetlagged, there’s a possibility I’ll fall off the stage, please catch me if I do — is because I think good, well-crafted, interactive narrative has many things in common with a well-mixed cocktail.

[pause while Alexis mixes a cocktail]

Ladies and gentlemen, and persons of indistinct and mysterious gender, this is the Smoke and Mirrors, thank you.

[applause]

So, choice, complicity, consequence. I want to talk about what makes a good choice, first.

Theme. Important in cocktails, important in choices, and the reason is, is because when your player has the experience you’re giving them, the theme is what they’re going to be thinking about. What they’re going to be thinking about is only partly within your control, but immediately what you’ve told them in terms of the story is the most important thing.

You can only fit so much story in their brain when you’re actually giving them the choice, so what they’re also going to be thinking about is all the other things you’ve done in the story up to that point. And the most common things in that story, the things that keep coming back, the leitmotifs, the key notes — light and darkness in Sunless Sea, for example — are the theme. So you want to find choices that comment on, elaborate on, annotate, align with, the theme. Like this, light and darkness, loneliness, survival, coming home. Terror.

Second thing, you want cocktails, you don’t want draught beverages. How many of you have played The Witcher 2? [pause] How many of you liked The Witcher 2? [pause] Ooh. How many of you followed every nuance of the politics and backstory of The Witcher 2? [pause, audience laughter] I salute you.

So this guy is an elven terrorist who’s name I believe is Iorweth. I was aware that he was no-good in a number of ways, but the precise nature in which he is no good, and whether I should ally myself to him or not, was very mysterious to me when I was playing The Witcher 2, and then what The Witcher 2 did was, it put me in a circumstance where he was unarmed and facing an armed enemy, and I had a few seconds to decide whether or not to throw a sword to him, and that immediately rendered down all the backstory to a basic human choice. If you can express something in a sentence that’s immediately emotionally engaging, that’s a good choice.

Mechanical significance. One of the themes of the narrative summit so far, I’m really glad to say, has been that non-linear narrative is more than just branching. Branching is a perfectly good model for how to approach a story responding to choices, but is by no means the only one. And, plugging your story, in some respect, into your game design ... ludonarrative consonance is one of the goals we talked about for a long time, and it looks like we’re moving more towards it.

That’s choice again. This is a really bad choice, this is the archetypical dreadful choice, and it has been in more than half the games I’ve ever played. Sometimes it’s a harmless choice, if you’re exploring a space it’s a largely harmless choice. The reason it’s been in every game since the very beginning, whether you call Colossal Cave or The Warlock of Firetop Mountain the beginning is it’s a really obvious thing to think of. But this is what I think of, when I present you with that choice.

Even if you say “would you like to go up or down?”, you’ve got something more to hang a response on than left or right or straight-on, because up is probably towards daylight and safety, and down is probably towards darkness and challenge. I want to take this choice and look at ways you could hang things on it to make it slightly more interesting.

Here’s one: just theme it up a little bit. Straight away, to look at those, you probably have an opinion on which door you’d prefer to go through, and what you might find behind them. Even if you don’t, they’re pretty to look at, and if there’s anything in the backstory or history so far that you might respond to, that makes the choice more meaningful.

This is a much more famous choice, and this is actually a left or right or straight-on choice. But, it has all the three things that we just talked about, and the mixological resonances. It’s themed, because of course the whole book is about trying to get to Mordor and back again with the ring. It is mechanically significant, because it’s a resource choice, whether you’ve leveled up enough to go through Moria safely for example. And it is easily understandable, you can get through it in a sentence.

There are three choices there, which is also important. I think it’s very easy to overload a context of choices, on the basis that more choices are better. They’re often not. Two is a nice number, it’s “to be or not to be”, it’s “sleep with the mechanic or don’t”, and it’s a really nice stark, dramatic way to provoke emotion. Three choices, like this, they’re asymmetrical. So you’ll notice the third choice actually comments on the first two choices.

You probably remember, from Rings, whatever your preferred flavour of Rings is, they’re saying Mirkwood, y’know, has got problems, Rohan, it has other problems, somebody says Moria, everybody hisses. You understand straight away that Moria is a different kind of choice from the other two. So that kind of asymmetry means you’ve got sort of a double balance; you’ve got a balance between the first two choices, and a balance with the third choice.

I prefer five choices to four choices, when we get into things that are more elaborate and generous. It does feel generous, and you can lock off one or two of the choices without even the player feeling restricted and railroaded by the final one. You always want, you always want to limit the number of choices, because if you want to provide real consequence — you don’t always have to provide real consequence — if you want to provide real consequence, you don’t want to write yourself too many content cheques you can’t pay. But very occasionally, you can do this.

This is a notorious choice from Fallen London, this is a choice where you decide how you justify entering a very selective members’ club, where you have something unique about you: that you are a notorious murderer, that you’re a really fine chess player, that your table manners are the worst in London, something like this. And no sane player — all of us have a number of quite insane players — but no sane player will ever expect you to provide forty-seven consequences, one for each of those. And this is a cheap, but delicious, trick to look really generous once in a while. Give people more choices than you could possibly provide consequences for, and let them roll around in it, and it looks very generous compared to this kind of frugality.

Choices that are visible, but unavailable, they’re kind of underused. People get shy about showing choices that are unavailable, because they worry it breaks immersion. Up to a point it does, and sometimes you won’t want to show choices that are unavailable because they’ll be spoilers, but if you show somebody an unavailable choice you’re making it clear that whatever they did earlier on to make that choice unavailable has consequences, giving them a reason to replay the game, and you’re making it visible to them, that whatever they’re doing is part of a larger world than they’re dealing with right at that moment. So, a useful technique.

This, uh, opinions differ on this. But, I wouldn’t go as far as to say that dictating your player’s emotional state is a crime against humanity. I would go so far as to say that it’s quite a missed opportunity. As I said, I don’t want to say what’s wrong or right, I do want to say what could be useful.

So here we’ve got an example where you’ve been invited to the palace, because your social stock has risen sufficiently high, and you have the opportunity to ask the player how they react to that. The problem with asking a player how they react to something, what the actual emotional state is, is their emotional state may be none of those. And even if it is, they already know what it is, it’s in their head right now.

If you specify a choice which instead of being about an emotional state, is about a physical enactment of that state, or a response to that state, first of all, you’ve immediately got more colour and incident in the choice, and it immediately becomes more interesting. Vivid flavours, remember. And secondly, if the player is feeling something completely different, if for example they can’t wait to get into the palace because they want to nick — if you’re American that means ‘steal’ — the contents, that’s one possible reaction, if they’re secretly a revolutionary, if they are delighted that they are expected, all these things are things they might be feeling, so give people a specific choice that they can use to express their emotional response, in favour to asking them what their emotional response is.

Complicity. Of the three — of complicity, choice, consequence — complicity is really the little brother. It’s the easiest to do, and also it’s really effective, and there are a couple of unexpected things worth bearing in mind about it. First of all, of course, you want whatever your theme in the story is, to bear strongly on a lot of the complicit emotions that your player feels.

Some of these would be more suitable than others, in different stories. If you’re writing a gritty noir, then ‘vengeance feels good’ is likely to be something that is extremely relevant. Compassing, or sympathy, or enthusiasm for small, furry things is likely less so. But the important thing, is you look at this list, ‘what have I done?’, guilt, something like that, can be a really powerful complicit experience. People seek out strong emotional experiences, they don’t always have to be positive ones.

And in fact, some of the choices in our games that people get most enthusiastic about, are very explicitly fantasies of failure. In Sunless Sea, should you wish it, you can enter a mystery cult on an island, you can kneel to be beheaded, you can have your body hollowed out and filled with candle wax, the head replaced, and then the thing wearing your skin will walk around and play the rest of the game as you. In gameplay terms, this makes no difference at all. [audience laughter] It’s purely an experience of complicity, it’s the choice you make in the moment, but people love this.

One of the most popular storylines in Fallen London — I can hear giggles from people who know what’s coming — for some of the most hardcore players, is the Mr Eaten storyline, where we explicitly tell people at the beginning of the quest, if you pursue this, you will destroy yourself and everything you’ve ever cared about. So obviously they signed up by the dozen.

This is a kitten. If you haven’t met kittens, I highly recommend them. I put a kitten in all my presentations.

[audience laughter]

Okay, uh. Every review of Skyrim you ever read — I almost guarantee it, it’s a law of nature — would say “I ignored the main quest, I went off and did all the sidequests”, because, y’know, Bethesda, main quests. I quite like that there’s the main quest, actually, but of course Bethesda — I’m going to go out on a limb here and say Bethesda is not populated by stupid people — and they know perfectly well, having been making these games for decades, that their players are going to ignore the main quest and go hunting the sidequests. So they deliberately provided a rich territory for the player to tell them to fuck off. And if you allow your player to rebel against things that you appear to be imposing, even if you are being completely devious in allowing them off-piece space to play in, they will love you for it.

Right, consequences. So, consequences are the devil to pay in any kind of choice-based interactive narrative. Even when you are very experienced in dealing with the possibilities of consequences, whether you’re branching, whether you’re doing any kind of immersion narrative whether it’s some sort of variable-based thing, you tend to get surprised by the amount of work it takes.

Here is Clive Owen, bleeding to death, in a field in southern France, because Matt Damon, on the right, has just shot him, in The Bourne Identity. This is basically what untrammelled consequence in interactive narrative does to you, you will literally bleed out in a field in southern France.

[audience laughter]

So I want to give you some rules, to minimise the chance of that happening, and to steer by when you are trying to provide a sense of meaningful consequence without blowing your entire developer budget.

So first of all, remember that what happens in your game is not what happens inside the machine, whether it’s a PC or a tablet, it’s not what happens in the character, it’s not what happens in the script, it is what happens inside your player’s head. If it doesn’t happen inside your player’s head, it doesn’t happen at all. Now of course you’re going to tremendous lengths to make the right things happen inside your players head, but if your player does not notice a consequence, it doesn’t happen.

That also means that if they don’t realise that a consequence is a consequence, it might not as well have been a consequence. This is why we like telling people about deferred conclusions up-front, because the more obvious it is that it’s a consequence, the more of a consequence it is. And this is why, those of you who played our games will know we use a particular UI approach to indicate that a consequence is happening. If you haven’t played our games, it’s way too complicated to explain in the next 90 seconds, so play our games.

[audience laughter]

But the third thing is, also, it is tremendously tempting, and this is especially true if you are in indie dev, with a lot of voluntary crunch in you, just to keep on spraying choices and consequence around the place. You know, if you want to do that, if you’re brave enough to do that, good on you, but remember that field in southern France. And, the more consequences you have in your game, the less your players will care about any given consequence. They have a limited amount of attention, there is a finite amount of stuff you can fit through the attention slot in their brain.

Smaller scale things are surprisingly effective ways of showing consequence. Top line is a trick everyone has seen, and most people in this room have probably used. You just put some sort of string token replacement in a line of prose: “an elf of your quality”, “a rubbery monstrosity of your quality”. I was down on this for a while because it seemed like such a cheap trick, but what I’ve actually realised is that players know it’s a cheap trick, people who have been playing video games for 50 years, they’ve worked this out by now, but it still feels like the game is respecting your choice and paying some sort of attention, it’s an incredibly low-cost thing to do, so do it.

The next line is sort of the inverse of this. In Sunless Sea, all ships are steam ships, because it’s set in an underground ocean, and obviously we want it to be historically accurate.

[audience laughter]

There are no wind-powered ships, so every time you write a chase scene, or an escape scene, we can have flakes of soot flying through the air, and the stokers stripped to the waist, and the firelight gleaming and all this sort of stuff, and you don’t have to write something different each time. So if you find the common factors in your story, and you write something nicely evocative around those, if you bring those out every time, that’s a really good way to get bang for your buck with prose.

Underspecified narrative, I think underspecified may be my favourite word, and I’ll come back to this again at the end. It’s something that’s come up a couple of times in narrative summit already, Richard mentioned it in the last talk with his talk about open-ended stories. If you can leave space for your players to interpret things and interpolate their own thoughts, they will co-own the narrative with you. Consequently, they have a much greater attachment to it. Don’t bread-and-butter everything, don’t write everything you need to, write as much as you need to and get out of the way.

Talking of which, two quotes here. One is from Robert Southey, he was a nineteenth-century poet laureate, a not very good one but I like this quote. The other is from Emily Short, who has now been quoted in two talks in a row — she gets everywhere, she’s kind of like God or malaria, she’s probably in this room right now.

[audience laughter]

As a writer, many of you if not most of you in this room are writers, it’s always tempting to fix a design problem by throwing words at it, because words are what you produce, and we enjoy doing it. That’s why we’re writers, God knows it’s not the pay. Your player is not paying you by the word, and your player wants to cooperate with you in having an experience. The game mechanics are one of the best tools you can use to allow them to have that experience with you. Don’t feel you have to make explicit every single thing that occurs in your narrative, where you can defer some of those things to game design.

To give you a really simple example, kind of a trivial example but quite an elegant one ... nope, that’s the next slide, I’ll mention this slide first.

To think about, one of the good ways into enacting consequence through design is thinking about what the resources your player is dealing with are. Nearly all games have resources to spend in some way, even if the resource is only the amount of time you’re spending, and the kind of resources that are most important in any given game will depend, again, on the theme, and you want them to be resources that your player is very familiar with and comes back to again and again. You don’t want dozens of different resources — something that we’re quite bad at actually in Fallen London — you don’t want dozens of different resources that players have to manage, you want people to care about these specific resources, and for these resources to carry a symbolic and emotional frame.

The example I mentioned a moment ago, this is a game called The Consuming Shadow, by Ben Croshaw, better known as ‘yahtzee’. It’s a procedural Lovecraft-y, roguelike thing, where you’re trying to avert an apocalypse, and the particular mechanic I like is very simple, very cheap, very effective. Every so often you receive a text, sometimes the text is information from the Ministry, sometimes it’s a message of support from a member of your family — which is great, it’s a lovely thing you don’t often see in games — one of your family members, an arbitrary family member, reaches out to you. And imagine trying to write any kind of narrative consequence to a family member reaching out to you, that could take you a long time, and this is a very small, very cheap game.

Sometimes it’s something like this, and the way that he deals with the consequences here is simply, there’s two meters up the top, which obviously are health and sanity, and if you get a cheering text it increases your sanity, if you get a miserable text it decreases it. Where you can, fold consequences of story back into game design. They’re the tools your player uses to have an experience with you, so use them.

I think there’s one particular benefit to interactive narratives in general, particularly choice-based narratives, and that has to do with the last great resource squeeze in digital space. The one thing that nobody has enough of in the digital space, the one thing that everybody wants, the one thing that you are all very kindly giving me right now, is this.

What your players want is to express themselves. What you can give them to allow them to express themselves is your attention. Because you can’t be there in person, you give them your game’s attention. If your players feel that your game is paying attention to them, it is listening to what they say, that it is validating their expressive choices, then that is the most effective thing you can do to engage your player’s enthusiasm.

We — and this is back to underspecified story — like to use this image to highlight that. Imagine a desert seen from above during the day, dotted with villages, with paths between each of the villages. Your players are like the travellers you can see passing between the villages and stopping every so often. Now imaging that night falls, and imagine that all you can see is the fires in the center of the villages. Your players are invisible until they come into that firelight. How they’ve actually got there is up to them.

If you have a game design that is more fragmented and modular, rather than strict branching, then it adds a bit of fluidity. If it’s based on a world model or game design, you can allow players to choose their own points between the fires, and the stories that exist in the spaces between the fires belong to them, rather than to you, so once again, you co-own the narrative, and they will feel a greater attachment to the story, and they’ll thank you for it.

And that’s me. Thank you for the cocktail.

[applause]