Or maybe your game adaptation of the Odyssey needs some Penelope scenes to reinforce that, yes–Penelope is pretty awesome and worth fighting monsters and gods to return to. You want to show how she’s handling her own troubles back home to inspire the player.

Or maybe you want to show Sidekick Guy getting murdered back at the base while the player is out on a mission. Or maybe you want to show the Great Dragon slowly awakening beneath the earth. Or maybe…

…well, you get the point. You want to switch away from the player’s point of view and use a cutscene to provide information or atmosphere you couldn’t otherwise get across. In a film, you wouldn’t think twice about it.

Here’s my advice: Think twice. That’s not to say you shouldn’t switch viewpoints for a cutscene in a video game. The advantages are obvious (and pretty much the same as they are in traditional, non-interactive media–thus, not worth discussing in great detail here) but there’s a lot to consider before you jump in.

First, be aware that a viewpoint-changing cutscene reduces the primacy of the Player Character. You are quite literally “breaking immersion” (a term I normally avoid, but a relevant one here) by shattering the tie between the player and the Player Character; what I see is no longer what the character sees, and we no longer share a mind and a common viewpoint. Generally speaking, a tight player-Player Character bond is to your benefit and the consequences of violating that bond are subtle but real–you’re signaling the player, intentionally or not, that this is not purely his or her story.

Second, consider whether the information delivered in the cutscene–a cutscene presumably not witnessed by the Player Character–would change how I as the player want to act as the Player Character. If I know that the story’s villain has placed a traitor among my Player Character’s companions, then I’m going to feel helpless and frustrated if the Player Character still treats his or her buddies like trusted allies. In a film or novel, the notion of a secret traitor would generate tension; it arouses concern for the characters on the part of the audience. In a game, it emphasizes my detachment from the Player Character and blocks my full emotional integration into his or her journey.

The funhouse mirror version of this scenario involves the Player Character emotionally reacting to situations shown in cutscenes as if he or she had seen them. This is less frustrating for the player, but it’s still a bizarre and distracting violation of character–if you show, say, the charismatic city mayor cackling evilly in a cutscene, then have my police investigator Player Character treat him with suspicion when they meet for the first time, I’m going to wonder why my character is so surly and paranoid. Sure, I know the mayor’s corrupt, but my character sure shouldn’t!

These are extreme examples, but less over-the-top versions aren’t hard to think up. If two NPCs are chatting enigmatically in a cutscene and delivering foreshadowing, and I can’t ask them about it when I’m in control of the Player Character? I’m not intrigued–I’m frustrated! On the other hand, if a cutscene contains no new information–if it’s essentially something that the Player Character could have imagined (“the captured prince is sad about being captured!” or “my loved one is worried about me!”)–you avoid this particular trap. A cutscene that contains no actionable information–a cryptic hint at something utterly unrelated to what the player is presently experiencing–is similarly “safer.”

Third, if you do think the advantages of a viewpoint-changing cutscene may outweigh the disadvantages, consider other potential mitigating factors:

• Do you have multiple Player Characters and switch between them already? Then your player-Player Character bond is already relatively weak and viewpoint changes are already integrated into your game’s style. This might reduce the issues with Player Character primacy.
• Can you make the cutscene into an interactive sequence instead? If you’re already violating point of view, you might consider going one step farther and giving the player a new way to emotionally engage with the characters.
• Is your game comedic? Is the cutscene funny? Then you can get away with pretty much anything.
• Is your Player Character something of a cipher anyway? (Maybe he’s a silent protagonist, or her motivations are intentionally unclear.) Then my choice to act on information only available in a cutscene (if such a choice is available) may be less distracting.
• Are you consistent about your use of cutscenes, and do you establish their use early on? Unveiling a single viewpoint-breaking cutscene thirty hours into a forty-hour game is very likely to jar the player. Conversely, if you cut to only one other viewpoint on a regular, predictable basis and start from the beginning (e.g., every time Odysseus leaves an island we get a Penelope cutscene), your player is more likely to become inured to the negative effects.
• The big exception to that last point is introductory cutscenes, which of course come before the player’s point of view has been firmly established! In this way, an introductory cutscene can do all sorts of things a mid-game cutscene might have trouble with.

Try to be aware of all the consequences of your narrative choices and weigh them accordingly. Changing viewpoint isn’t a casual decision in a video game, and while it can certainly be employed adeptly, well… if you’re not sure it’s worth it, you’ve got plenty of other tools available.

This post is a companion to Writing Romance in (non-Romance Games): Linear Romances. You’ll find a number of assumptions about our overall topic outlined there; they still hold true, and I encourage folks to start with that post before jumping into this one. Here, we’ll cover three overall subjects: branching romance fundamentals, linear techniques in branching games (where we’ll revisit some of the approaches discussed in the previous post), and “casting” a variety of romantic interests in a branching game.

Branching Romance Fundamentals

When I talk about “branching narratives” I’m referring to games in which the player can make significant plot choices (whether inside or outside of dialogue) and, in the context of romance, choose whether to pursue a potential partner and how to treat that person during the relationship. Games in which the player has a choice of potential romantic partners but no decisions in the relationship itself fall into a gray area between branching and linear–the posts on both topics should prove applicable. For a more thorough treatment of branching dialogue generally, see this series.

Let’s run down some high-level challenges and approaches for branching games.

Is this really a good idea? “Players love romance!” is not a great reason to implement romance subplots in your game. Consider instead questions such as:

• Do the overall themes of the game lend themselves to romance (and if so, what kind of romance)? If a subplot doesn’t support, subvert, comment upon, or otherwise relate to your game’s primary focus, then it’s just diluting your strongest element. The answers here might not be obvious–one game about the importance of building relationships and alliances in service of a larger goal might be strengthened by the inclusion of romance, while another similar game might be weakened by losing its laser focus on platonic friendships.
• Do the circumstances of the plot support the development of romance? If you’ve got a game that takes place in (for example) an exceedingly short timeframe, there’s probably not a lot of options for naturally developing romance subplots. Look at your pacing, your cast, whether your player character would reasonably get a chance to initiate anything romantic, and judge appropriately. Check, too, whether there are romantic relationships between NPCs of importance–if they’re a major part of the story, it may feel tonally strange (in some circumstances) that romance can’t play a role in the player character’s own life. (The converse is also true–if I don’t see any romances in the world, their absence in my life feels less like an omission.)
• Does the player character’s sexuality matter? While the player character in a branching narrative game may not have a clearly defined personality, his or her character traits still exist within a range defined by the writer (Mass Effect‘s space marine Commander Shepard is never a pacifist). Regardless of whether you implement romance subplots in your game, know the romantic “range” you want to permit the player based on your range of player character archetypes. A coming-of-age story with a teenage protagonist will lend itself to obvious romantic options more than a military and political story starring a seventy-year-old veteran (which isn’t to say you couldn’t push either in another direction).
• Do you have the resources to do your vision justice? You only have so much time and money, after all. Are romance subplots the best way to spend your resources?

Depending on the answers to these questions, you may find yourself with a game where eschewing romance subplots feels unnatural, even if you don’t particularly want to support such stories. (If you’re writing a James Bond-style spy in a Bond-style story, it’s hard to justify not letting him or her try to seduce an asset.) Or you may discover that, despite the sparks a romance could add, it ultimately drags down your narrative more than building it up.

Give it the weight it deserves. There are lots of different ways to inject romance into your game if you determine it’s appropriate. You could allow the player character to casually flirt with dozens of different NPCs but never develop those “romances” any further; you could present a handful of potential love interests but leave the relationships largely subtext; you could present short optional romance subplots serially, in different areas of the game world, none of them lasting beyond the scope of their side quests; you could track long-term relationships with major NPCs, chronicling developing friendships, first kisses, and going all the way to wedding bells and beyond.

Figure out the scope of your romance–whether a line here and there or something hefty and pervasive–and make sure your scope is in line with your vision. It’s okay to have a pleasant flirtation that never “goes anywhere” if that’s what fits best, so long as you’re not promising the player more than you can provide.

A romance subplot’s “weight” ought to affect its integration into the overall story structure, too. If you do have a long-running, start-to-finish NPC relationship, try not to sequester the romance content in discrete scenes–weave the romance into the main plotline at appropriate moments. Let the player character talk about his or her love life with platonic buddies. Have decisions unrelated to the romance bring out references and consequences in the romance plot. (That’s not to say romance-exclusive scenes are bad–but if you want your romance to be “important” then they shouldn’t be the entirety of the subplot.) The more separate the subplot is from the rest of the game, the more players will treat it as a minigame of sorts–a minigame whose themes and expectations may not carry over to the broader narrative.

Know what sort of relationship stories you’re telling. Based on the overall themes and sensibilities of your game, do you want operatic romances often doomed to end in tragedy or betrayal? Do you want fantasy fulfillment romances, in which all problems can be eventually overcome through mutual understanding and where blissful coupledom is the natural end? Do you want a naturalistic romance, where problems and misunderstandings accrue, couples can fall apart over time, and the future is always uncertain and relationships always evolving?

You can have some variety, of course, by utilizing different romantic interests and different branches based on player choices. But you’ll want to make sure everything fits with your core “feel.” You’ll want to teach the player what to expect as early as possible (preferably before the relationship even begins), at least in a broad sense–don’t let the player go into a romance expecting the love interest’s flaws to be treated as charming, romcom-esque fodder when you’ve written a cerebral exploration of emotional abuse.

How do you indicate to the player what to expect? If your romances are in key with your larger themes, some of the work should be done already. You can also show examples of other characters in relationships to help nail down the tone in the player’s mind and use NPCs to call out expectations directly (“Are you sure you want to get involved with him? He seems funny, but also mean.”)

I advise against attempting a bait-and-switch to surprise the player regarding the general tone of a romance–rather than being awed at the narrative twist, most players will simply become frustrated with the game (and its writers). That doesn’t mean you can’t have surprises, but they should be the kind of surprises a player might expect–if my romcom-style love interest turns out to be married and many misunderstandings and hijinks ensue before the happy ending, great; if my apparent fantasy romance involves the slow deterioration of the relationship because the love interest is wrestling with deep-set problems totally beyond my control, then not-so-great. (I discuss player expectations and theme a bit more here.)

Be especially careful with fantasy fulfillment romances, even in games that are, broadly speaking, power fantasies. A romance that comes without sacrifices, that results in what’s by most definitions a “better” end for the player character, becomes a sort of trophy–and once you start offering up trophies, it’s hard to present not having a trophy as an equally viable option. If your game has the tacit assumption that ending up coupled is always superior, then you’re putting yourself in a position where either a) you need to provide a particularly wide variety of romantic interests to accommodate every taste; or b) you end up encouraging players to romance NPCs they don’t actually like, just to get the “good” result. “A” is a massive pain for the writer. “B” is awful for the player. This isn’t to say that power fantasies are bad–they can, in fact, be empowering–but know what you’re getting into.

Allow multiple entry and exit points. One of the joys of branching narrative is that it allows player characters to change over time. In recognition of this, try to give the player multiple chances to begin or end a given relationship.

Entry points are largely a matter of convenience and a means of reassuring the player that, if a given moment doesn’t feel like the “right time” to start a romance, it doesn’t have to be. Exit points are more important–if the player feels trapped in a romance about which he or she has doubts, that resentment will very likely grow stronger. Simply by presenting an escape route, you acknowledge the player’s doubts and misgivings about the relationship and assure the player that, no, the relationship isn’t intended to seem perfect–there are real problems here, and it’s totally up to the player if he or she wants to try to make it work or to part ways.

Encourage a give-and-take. As with non-romantic relationship-building, make sure that (if your game has branching dialogue) player characters are given opportunities to talk about themselves and their feelings. It’s easy to write an NPC who talks at length about his or her background, relationship with the player character, hopes, fears, and so forth–but to help create a real bond, the player character needs to be able to express the same sorts of things and feel heard.

Allowing the player character to express his feelings about the relationship doesn’t need to be limited to the relationship itself, either. Why can’t the player character’s sidekick (or best pal, or therapist) ask how the relationship is going? Let the player complain or not complain; either way, it helps build up the relationship’s narrative within the player’s mind.

Don’t make assumptions about player reactions. The love interest you intend as charming may come across to some players as sleazy; the love interest you wrote to be shy make come across as frustratingly oblivious. That’s fine–different players should bring different interpretations of the narrative to the table! But make sure you give the player plenty of options to express dislike, disenchantment, or offense at the actions of (potential or current) romantic partners–limiting a player’s options to insist that she like a character tends to change the player’s dislike for a character into a dislike for the game in general.

(Allowing the player character to react negatively can also allow you to change course, if need be; maybe the “charming” love interest turns apologetic when called on his sleaziness and tries to mend fences with the player character. Or maybe not.)

Beware the triangles. Are there love triangles in your story? Do NPCs get jealous? Does it only happen if they’re actively in a relationship with the player character, or can they get jealous if they’re not in a relationship? Does serial monogamy have consequences? Are any of the love interests polyamorous? Are only some of the love interests polyamorous?

Don’t let the complexity of your romance subplots dwarf the rest of the game (or just accept that you’re writing a romance game, not a game with romantic subplots). Make sure that you’ve got a plan for dealing with all entanglements that fit the personalities of all characters involved and doesn’t leave you overworked. If that means finding ways to geographically separate potential love interests, do it (“My first wife lives in New York; my second lives in Hong Kong. They don’t know about each other.”); if that means establishing polyamory as universal in your game setting with some quick and awkward exposition, do that; if that means reducing the number of romantic interests so that you can have a branch-heavy, nuanced set of interactions between them, go for it! But have some sort of plan.

Build equally strong non-romance content. The more substantial and pervasive your romance subplots, the more important it is to have equally compelling exclusive content for players who choose not to engage in a romance.

Why? Not out of any sense of “fairness,” but because you’ll want to persuade players that they don’t have to engage in a romance to get the game’s full experience. Fail to do so, and you’ll get a lot of players angrily taking up romances they don’t like. This isn’t such an issue if your romance subplots are relegated to small scenes and sidequests here and there, but if they’re heavily integrated into your main plot, this is an issue you should consider.

So how do you build compelling non-romance exclusive content? The most obvious way is to find a different tack to approach the relationship with the same NPC–to develop a platonic friendship or rivalry that’s interesting and rich in its own right and opens up a side of the NPC the romance-involved player never sees. You could acknowledge the player’s choice to not engage in a romance, having other characters questioning that decision and allowing the player to express her reasons. Maybe engaging in a romance blocks off certain other paths altogether–an intense relationship causes the player character’s non-romantic relationships to suffer, say. Anything that reveals a meaningful path in opposition to the romantic choice will feel rewarding to the player.

Try to make it clear to the player that exclusively non-romantic content is exclusive, as well–that the choice not to pursue a romance led to a certain point. Otherwise, the player may still erroneously believe that romance content is an extra, not an alternate path.

Take care when attaching mechanics. Attaching any sort of game mechanics directly to a romantic relationship–anything from a friendship meter to a need to grind content to unlock new relationship “stages”–can be risky. On the one hand, doing so can nicely integrate the relationship with the fundamentals of gameplay. On the other, such mechanics can provide a clear goal and “win point,” strongly encouraging players to push forward regardless of their feelings about the relationship. (Personally, I’m a big believer in using visible mechanics to augment narrative, but be aware that players tend toward optimization!)

Signpost choices and consequences. It’s always wise to signpost branches–to make it clear to the player what sort of choice she’s making before she actually makes it–and with romance subplots, players may be especially frustrated if they inadvertently move down an unintended path. If a player has an option to start flirting with or otherwise pursuing an NPC, make that choice unmistakable; if the player can break up with a love interest, make that clear, too. (I talk more about signposting here.)

Making the intent of an option clear doesn’t mean that the ultimate outcome needs to be obvious. It’s fine for aggressive flirting to turn off a potential love interest while friendly conversation makes the NPC more receptive; but such friendly conversation shouldn’t “activate” a romance without the player making a clear and deliberate choice. (If it leads to the NPC seeing interest where none may exist, that can make for interesting and awkward drama–but then you need to give the player the option to express surprise at being misinterpreted.)

More than anything, avoid leaving the player confused about the status of a relationship, wondering “Why does the game think X#8221; If the player character is meant to be confused, allow the player to express that confusion.

Don’t try to please everyone. We’ll touch on this more below, but really–don’t. If you want tragic romances in a game where they make sense, run with them and accept that you’ll have a small number of players who simply want a different game. It’s okay not to make a game for everyone–so long as you make it clear early on what game your players are getting.

Linear Techniques in Branching Games

In our last discussion we ran through a long list of techniques for romances in linear games. Let’s review a few of them and whether and how they might be applied in a branching context.

Establish the romance as part of the story’s background. Tricky, but not necessarily a bad idea. This narrows the spectrum of potential player characters considerably, of course, but if the player is given proper exposition and context then there’s no reason he or she won’t willingly buy into the relationship. You just need to show the player that she doesn’t have full ownership of the player character, but rather a small variety of options as to how to play him.

You’ll need to contend with the nature of the relationship itself being indefinite early on, in the same way that a player character’s personality is indefinite before the player has made a good number of choices. You’ll want to use other linear techniques to try to establish the plausibility and likability of the relationship, too–you can’t expect the player to much care about her love interest in this scenario without a good bit of work, and the player may revolt if the love interest isn’t sufficiently compelling and believable. (In other words, don’t expect the romance to matter to the player just because it matters to the player character; you still need to earn it.)

Make the romance engaging and enjoyable and don’t make the romance an obstacle. You’ve got more flexibility on these points in a branching narrative, largely because your player (presumably) chose to get involved in the romance and can back out if it becomes too unpleasant. Making the romance into an obstacle sometimes can add real nuance and test the player’s dedication and willingness to sacrifice.

Nonetheless, if you want the romance to be generally a positive experience, these points should still generally hold true. And if you are going to make the romance into an obstacle, make sure it’s the right kind: Forcing me to, e.g., grind for gift objects in order to maintain my loved one’s affection doesn’t add any narrative depth to the relationship; but dealing with my loved one’s frustrations about whether and how I keep working alongside my ex could strengthen or break the relationship in a story-appropriate way.

Integrate the romance into gameplay. Just as true in branching narratives as in linear ones. If your game has branching dialogue or a similar system, that is gameplay, so you’re probably a step ahead already. But… don’t forget the rest of your gameplay, either.

Make it a one-way street. This is a difficult approach to deploy effectively in a branching narrative–if an NPC is mooning after me, why can’t I just “go for it”? Even if you believe that you’ve established all viable variations of the player character as too shy or obtuse to respond to an NPC’s interest, you’re still setting up a situation where the player gets frustrated by her own character and by the game’s lack of options. Be careful of having any NPC flirt or otherwise express interest in the player character without allowing the player to respond in kind!

(As always, I’m sure there are ways to brilliantly subvert this–maybe you want a game about self-loathing, in which the lack of player options and ensuing frustrations are intentional–but know the risks you’re taking and how horribly wrong things could go.)

Make me feel my romantic interest’s absence. This one’s harder in a branching game, since most players will likely be running through the same scenarios regardless of whether they’re in a given romance or not. But by emphasizing positive, unique traits in the romantic interest you can still make his or her absence felt by separating the player character from his loved one.

The relationship is what you show. Still totally true. Keep in mind, however, that the player is also constructing a parallel narrative of the relationship in her head based on what you show–all the player character’s complicated feelings are in the player’s hands, not yours. Treat that variable with care and respect, and avoid assuming the player is on the page you want her to be.

Remember what your game is really about (not romance). As we’ve discussed, building your romances in the context of a larger story remains vital.

Go vague and archetypal. Sure, why not? It’s your weird game!

Casting for Branching Romances

“Casting” isn’t the right word, but I don’t have a better one. Assuming you want to provide more than one romanceable NPC in your branching game (common but not necessary–there’s no reason all romantic elements couldn’t be placed on a single character), how do you determine what sort of love interests are available? Whether you’re creating NPCs with the intent of making them love interests or figuring out who in an existing cast of characters should have a romance branch, you’re going to need to make some difficult decisions.

For the moment, let’s focus just on character types. We’ll get to questions of gender and sexual orientation after.

You can approach this question from a few different angles. You can take a purely meta approach and figure out what sort of characters your players might want to romance–considering your target audience, looking at popular pop culture romance characters, and so forth. This, however, doesn’t guarantee you’ll get anything particularly relevant to your story and themes. Conversely, you can approach the problem from the story side, trying to find relationships and character types that make for particularly interesting drama. Lean on this approach too hard, however, and you may end up with romances that players aren’t interested in actually playing.

Somewhere between these two approaches is one focused on the player character. If you’re writing a branching game, it’s imperative to understand the range of player character personalities you’re permitting and to have an idea of what personalities players are most apt to gravitate toward. These player character archetypes should be appealing in their own right (there’s little point supporting a player character personality that no player will end up role-playing); therefore, if you create romances that appeal to your different player character archetypes, they should (in theory) also appeal to a variety of players and naturally be well-integrated into your themes. Look at your player character possibilities and find relationships that are both attractive and challenging–build love interests that certain player character types would gravitate toward, and that have the potential to bring new depth to your story.

But here’s where we hit a problem: In most games, your cast of characters will not be genderless blobs. If you’re allowing the player to determine the player character’s gender and sexual orientation, then your life gets a lot more difficult.

To an extent, the more your game is interested in exploring gender and gendered relationships as a theme, the easier your job gets. If you’ve got a branching modern-day military RPG that’s deeply engaged in questions of military culture and masculinity, then you’ll probably find that your love interests (of whatever gender and orientation) sort themselves without massive difficulty–their personalities, backgrounds, and subplots will be tied to your core themes and you’d be hard-pressed to easily rewrite, for example, a gay male love interest subplot into a bisexual female subplot in this kind of context. Or imagine a branching game invested in questions of women’s power in 11th-century Germany–again, your love interests are going to have relatively clear thematic roles exploring these ideas.

On the other hand, the more you make gender a nonissue in your world and your central themes, the more your romanceable characters’ subplots may become detached from gender and orientation. If “wisecracking rebel” is a prominent player character archetype for my cyberpunk game, I may decide that “by-the-book member of the establishment” makes for a strong Romeo-and-Juliet-esque romance choice; but should that love interest be male or female? Gay, straight, or bi? Trans or cis? There’s no obvious “best fit” here the way there would be in our medieval German game (where if I want a romance subplot emphasizing male-female cultural power differentials, I’m not going to put it on a gay male love interest). But we need to pick something–or else build multiple love interests who are all pretty much identical, or find a way to procedurally generate the NPC based on the player character, or find some other high-risk or expensive solution. Multiply this out by several different player character archetypes and their potential matches (I need an establishment love interest for the rebel archetype, a cynic for the idealist / reformer, and so on…) and the more you have a challenge.

And of course, the player first creating a player character won’t have any idea what romances may be available in your game–it’s never great to have a player think, “Well, if I’d known a love interest with that personality was around and only interested in women, I’d have made a female player character!”

So, yeah. Those are problems. There’s no easy solution. Here are a few options to consider:

As discussed previously, beware “trophy” romances without alternatives. If romantic subplots are treated as rewards and sufficiently rich in exclusive content to be superior to non-romantic alternatives, can you blame the player with a lesbian player character for being irritated by the lack of a variety of lesbian or bi love interests? If romances are treated as a thematic or gameplay “win,” creating a world in which straight male player characters (for example) are most likely to achieve their goals (by virtue of having more and more varied love interests) probably isn’t your intent (unless, again, that’s one of the themes you’re going for!) Finally, the more your romances are fantasy fulfillment, the more likely it is that your players will want love interests of their preferred sex (rather than their player characters’ preferred sex).

Reduce the “weight” of the romances to a degree you can support. Maybe you can’t budget 20,000 words apiece for multiple NPCs of every possible variety, but you can budget 2,000 words apiece. This means a very different role for romance in your game, but is that necessarily bad? If you’re not making a romance game and you’re spending disproportionate resources on romances, is that the best use of your time?

Make your non-romance options rich and interesting and full of exclusive content in their own right.

Accept imperfection. Earlier I wrote “Don’t try to please everyone.” This includes you. Your romance subplot casting will probably not have all the variation and diversity that you’d ideally like. On the other hand, your game is not the only game in the world. It probably won’t even be the only game you ever make. Get it right over the long term, and do what you can in the short term.

Consider your target demographics. This is a sensitive topic, and one that requires a more lengthy discussion than I’m prepared to go into here–objectionable use of (often inaccurate) demographic data has done terrible damage to the games industry. That said, it’s always worth considering your content in the light of who’s actually playing your game and what their interests may be, as well as who you want playing your games in the future. When it comes to romance, consider also whether the players most interested in romance subplots form a distinctive demographic in their own right (not just in terms of age, gender, and so on, but also in regards to why they come to your game).

Knowing your audience doesn’t require catering to your audience’s whims. But ignorance of your audience won’t help you reach anyone (old or new).

Restrict player options upfront. If you can’t actually support the variety of player characters you allow the player to make, maybe you need to reduce that variety. If your plot only has one significant NPC and you want romantic tension between that NPC and the player character to be a major source of drama, well… maybe you just don’t let me make a player character who doesn’t fit the plot’s ideal gender and orientation. (Firewatch is a good example of a non-romance branching game that restricts player options in this way.)

Even in a branching game, there’s nothing wrong with telling the player upfront, “For this story to work, you need to play a lesbian” (or a straight man, or a gay man, or whatever); just be sure both parts of that statement are actually true.

Cheat. For games and characters where certain questions of social dynamics are largely pushed aside, there are ways to “cheat” by dynamically adjusting the gender or orientation of NPCs. In a branching dialogue RPG with voiceover, a given love interest is probably going to sound (thanks to the voice actor) male or female, but could have a variety of alternate lines to support his or her identity as gay, bi, or straight, depending on the player’s own choices. In a game without voiceover, text could be dynamically adjusted to result in a whole spectrum of NPCs from one baseline personality.

Of course “cheating” in this fashion has a number of downsides–in addition to the difficulty of implementation, it does make it hard to create NPCs with personalities shaped by the intersection of their culture with their sex and sexuality. For some games and some NPCs, this may be fine; for others, it could be problematic.

Say, “Screw it, I’m making a pure romance game.” If that’s where your heart takes you, then go for it!

(Thanks to Andrew, alias Anndru, for the Star Wars: The Old Republic screenshot!)

Compelling romance subplots are tough to write at the best of times. Consider their place in linear media–how many otherwise strong films or novels suffer from an unconvincing, uncompelling, or tacked-on romantic element? Add in the complexities of interactive narrative and of course things often go badly.

But that’s no reason to eschew romantic storylines altogether. No one needs convincing that there’s a rich vein of material to be mined here. So what should we think about as game writers when we introduce romantic elements into our projects? What are the pitfalls to avoid and the game-specific challenges that need to be overcome?

I’ll be tackling these issues across at least two different posts–first analyzing non-branching games, then concentrating on romance in branching narrative (with possibly some other stops in-between). Before we get into either of those, however, let’s narrow our focus a little…

The Elephants in the Room

This is a big topic. We’re not talking about a lot of it. Specifically:

Our interest here is mainly games whose primary focus is not romance. If you’re writing a story-driven adventure focused on chronicling a specific romantic relationship, a party management game where pairing up individuals is your primary mechanic, a dating sim, or any other game in which removing the romantic aspect would leave you with little gameplay and story left, a lot of this advice may not apply. (A lot still will–just be sure to consider it in a different context.)

We’re also assuming that you want your romance to be a meaningful and well-constructed part of the narrative (if your romance is intended to have the weight of the relationship between Mario and Princess Peach, you hardly need me) and that you’re already comfortable writing romances in traditional media. If you’re still struggling with basic issues like how to pace a romance, how to make a romantic interest an engaging and compelling character in his or her own right, how to handle romantic dialogue, how to avoid gendered stereotypes, etc., you’ll find no help here.

We’re looking only at romantic relationships between the player character and non-player characters. If you’re figuring out how to deliver a strong narrative with NPCs romancing one another, you’ve still got a challenge–but it’s little different from making the player care about any relationship between two NPCs. The romance is entirely filtered through the player’s eyes, and matters mainly in how it relates to the protagonist. (This may be worth an article of its own at a later date.)

We won’t be talking much about audio, visual, or mechanical aspects of romance in narrative, though we’ll touch on these items in a few spots. There’s lots worthwhile to discuss here (you could easily fill an article or two on visual character design for romantic interests–whether to favor the gaze of players or player characters, and so forth); it’s just not our topic today.

We won’t be talking specifically about sex. Since we’re not worrying about romance-focused games and we’re assuming you can write well and tastefully already, handling sex scenes should be relatively easy compared to everything else.

We will be talking about romantic storylines of all stripes–not just “fantasy” romances. Nuanced, rocky, and even hurtful relationships are all good storytelling fodder, and we’ll highlight a few challenges specific to unhealthy or simply not-blazingly-positive romances.

The concept we’ll come back to again and again is this: Employ romance the way you would any other important theme within a larger narrative framework. The difference between integrating a romantic element and a horror element into your military shooter isn’t one of kind; yet players have special expectations of and particularly strong reactions to romantic elements that raise the stakes and the risk of derailing the narrative.

Non-Branching Games

Compared to branching narrative romances, romances in non-branching games are comparatively simple. But “simple” doesn’t mean “easy.” The challenges may be every bit as great.

(Note that I’m using the term “non-branching” instead of linear for this category so as to include, for example, open world games that lack branching storylines. One can argue whether the Assassin’s Creed or Grand Theft Auto games qualify as narratively linear, but if they don’t offer multiple story branches they belong here.)

In the AAA space, non-branching games have a fairly poor track record for handling romance. Too often, romance elements boil down to “you killed my dog” or a paper-thin “obligatory Hollywood romance subplot” that adds no real thematic or character depth to the work. These aren’t approaches that can be easily redeemed by smart implementation, but let’s assume you have loftier ambitions.

Consider the following:

Is this really a good idea? The overriding risk of implementing a romance in a non-branching game is this: Game narratives are most powerful when a player’s motivations and emotions sync with the player character’s, but love and attraction are hard to reliably reproduce in a player audience.

It’s one thing to make me (as a player) like or loathe a non-player character, but much harder to make me attracted to one (assuming I’m attracted to persons of that NPC’s gender at all). My ability to generally empathize with a protagonist in a romance or simply enjoy the relationship as engaging drama–as might apply in a book or film–is less important in a game than my ability to share the player character’s headspace.

Implement a romance poorly and you risk the player becoming frustrated and detached from the player character. “Why” (our theoretical player asks) “is my character spending so much time pursuing this awful (or merely not-to-my-taste) guy? I don’t want to do this. I want to go back to shooting aliens.”

If you think you’ve got a romance element for your narrative that’s worth the risk, there are ways to avoid or mitigate these problems. You don’t need to use all the tools below, and likely don’t want to–but individually, each is a way to smooth out the experience.

Establish a strong player character personality. So much of game writing is getting player “buy-in” by establishing expectations early on. If you intend to compel the player character into a romantic situation, make sure that the player understands from the start that she’s not free to project her own motivations and personality onto the player character–that the player character’s personality is very much predetermined, and that the player is as much accompanying as embodying him. The more control the player thinks she has over the character, the more frustrating it will be if her character does something she doesn’t want.

And make sure that the player character’s personality isn’t just well-realized on paper–convey the particulars in an unmistakable fashion! Broadly speaking, players will be a lot less resistant to a romance involving a Nathan Drake-style player character (a clearly defined personality) than a Gordon Freeman (silent and entirely player-projected). That said, for truly vague or passive protagonists, see “Make it One-Way” below.

The more the player feels a sense of ownership over the player character (and by extension, the romance), the more likely the player is going to have complaints at some point when the romance plot moves forward without player control (“but I don’t even like him–why are we kissing#8221;). The downside of reducing ownership, of course, is that ownership is a powerful and valuable tool to improve player engagement; you need to make sure this tradeoff is worth the gain.

Establish a strong personality for the romantic interest. This goes hand-in-hand with the item above. In taking this approach, your goal is to convince the player that the player character and romantic interest obviously should be attracted to one another, based on their established personalities. It only makes sense, even if the player doesn’t actually like it. If your player is ever wondering “what does my character see in him / her#8221; then you’ve got a serious problem.

You may not be able to achieve emotional synchronization between the player character and the player, but you can at least get the player on-board intellectually (which in turn may lead to the sort of audience-protagonist sympathy you’d find in traditional media). This works well for rocky or doomed relationships, too–if the attraction is obvious and convincing, the player is more likely to run with the storyline rather than simply being frustrated by it.

How do you handle this in practice? Great dialogue helps–if the banter is funnier and flows easier between player character and romantic interest than with anyone else, or if player character and romantic interest share a unique interest no one else has, it’s a place to start. Don’t rely on physical appearance or some sort of vague, charismatic “presence” or mystique–if I can’t glance at the dialogue in a scene and come away thinking, “Man, I may not personally like the love interest, but he’d obviously be appealing to the protagonist!” then you haven’t gone far enough.

Establish the romance as part of the story’s background. Declaring that the player character was already involved with his or her romantic interest before the game’s start helps with player buy-in–the player may not like it, but won’t question the characters’ pairing as would be possible with a romance kindled in-game.

That said, this isn’t an excuse not to do the groundwork–you still need to convince the player that these characters are a natural match. It’s just a little easier this way.

Make the romance engaging and enjoyable. Does it need to be said? Yes, it does. The more the romance is something I want to see–the more it’s funny, heartwarming, hot (though be careful with this one, given varied player orientations and interests), and so forth–the more I’m willing to buy into the romance and accept it as part of the story. Make me grin at two characters joking together; make me smile at two shy characters fumbling toward romance; make me laugh in shock and horror at the antics of two villains in love. If it’s fun for me, I assume it’s fun for the player character, and can give it a lot of leeway.

A lot of games fail on this particular point, presenting lovers or would-be lovers tragically separated; or relationships at their lowest ebb amid stress and tension. But such presentations don’t make the player want to rekindle the romance per se–at best, they make the player want to alleviate the characters’ pain. If you want me to want the romance, make the romance something worth wanting.

For “bad” romances, this method has its limits–if a relationship has profound flaws and is doomed from the start, you can’t make it entirely desirable–but it can also play an important role in emphasizing the good side of the equation. A bad romance without any redeeming qualities isn’t terribly plausible or terribly interesting.

Make the romantic interest impossible to dislike. This is a cheat I’m not really a fan of, but it can work: Scrub away the romantic interest’s rough edges, make sure she’s relentless supportive (but not servile or humorless), give her a sharp-but-inoffensive sense of humor, and never let her seriously disagree, lie, or put her needs above mine. Load her up with appeal and never give the player an excuse to dislike her.

The problem with this approach, of course, is that you’re not left with a particularly interesting character and you’ve exorcised most of your potential for genuine emotional drama. You’re giving the player a sugar rush, which is rarely a boon to storytelling intended to carry real weight.

I won’t lie: people love sugar rushes. But you’re aiming higher than that, right?

Don’t make the romance an obstacle. Conversely, be exceedingly careful about making your romance into an obstacle to–well, to pretty much anything the player may want. There’s little that irritates a player more than something that gets in his or her way. Does it impede progress in the main storyline (a mission in which I need to rescue my lover instead of killing the alien boss)? Does it keep me from gameplay (lengthy romantic cutscenes with no direct bearing on the plot, or a romantic interest who wants to keep me out of danger)? Does it frustrate my play experience (escort missions!) When you make your romance into an obstacle, you risk souring the player on the romance as a whole for reasons not true to the narrative.

Angst, no matter how appropriate outside of games, is another obstacle and difficult to deploy effectively–it’s enormously difficult to get players and player characters mooning over an NPC or a lost love together, and it’s difficult to give a player much to do in this regard. Games are active experiences, and while emotionally anguished player characters are fine, you’ll have an easier time giving the player something to directly work toward.

That’s not to say you can never deploy romance as an obstacle, but that tactic tends to work best when the player is thoroughly sold on the concept already. If I have to change course to rescue my romantic interest, I’d better be really invested in that character. (That’s not the player character who needs to be invested–that’s me, the player.)

Integrate the romance into gameplay. This relates closely to the previous points (see also “Remember what your game is really about,” below). If your romance has no impact on gameplay, it’s almost, by definition, an obstacle to the game’s primary experience.

That said, “integration” and “impact” can mean a wide variety of things. Is the player character’s romantic interest a helpful sidekick character, an enjoyable and consistent presence during gameplay? Is the romantic interest a voice providing advice or instructions? A fun enemy to be battled in boss encounters? Are gameplay levels representative of different levels of the romantic interest’s soul? Are “relationship mechanics” built into play, where the more the player takes certain actions (protecting the sidekick romance character in gameplay, collecting flowers, whatever) the stronger the relationship becomes and the more bonuses are stacked onto the player character?

Be very careful if you’re using a romance plot to contrast the primary plot of a game, treating it as a passive gameplay experience that the player returns to between “real” game segments. Try not to make romance a speed bump. If you want my tough-as-nails soldier player character to return to her husband during downtime between missions, don’t just make these passive cutscenes–find a way to engage the player and make the experience relevant to the rest of the game.

Make it a one-way street. Or… do things very differently. This one works well with relatively undefined player characters–silent protagonists and the like–and gets increasingly awkward the more the player character develops a true personality. A romantic interest who adores the player character can be a powerful thing–players generally respond well to flattery and adulation and a sense of being wanted, even if they don’t necessarily wish to respond in kind. I’m probably going to at least be sympathetic toward a well-realized and intriguing character who clearly is falling in (non-creepy, non-stalker-ish) love with me.

Of course, you need to justify why the player character doesn’t respond (whether positively or negatively). For silent protagonists in relatively subtle “relationships,” you may be right at home; for more active protagonists, you may need to put an actual communications barrier in place (the relationship is literally one-way, with the romantic interest able to speak to the player character but not vice-versa).

As mentioned, you really want to be careful to avoid a creepy vibe with this approach. For that reason, I’d absolutely recommend not trying this in reverse–an adoring player character and a passive / silent / dead romantic interest is very likely going to seem overbearing.

Make me feel my romantic interest’s absence. Romance is intense. If you want to convey that intensity, make sure the player feels when the player character and romantic interest are apart.

Assuming the romance is generally positive, emphasize those positive traits when the romantic interest is present and emphasize the opposite elsewhere. e.g., if I’ve got a romantic interest sidekick character and the relationship is defined by excitement and one-upsmanship, make sure the missions where we’re together feel appropriately exciting and over the top; whereas those without the romantic sidekick might feel somewhat stifling, lonely, or overly cautious. Make sure many of my best times (however you define “best”) are with the romanced character so that I always want to get back to him or her. You don’t need the player character to mope when he’s alone if the joy of the active relationship is powerful enough; the player will make those comparisons herself.

If the romance is ultimately harmful, go the opposite route: No matter what enjoyable aspects life with the romanced character brings, I should feel better when the character isn’t around.

The relationship is what you show. If all you depict of the relationship is arguments, nagging, or one partner getting into trouble and frustrating the other, then that’s what defines the relationship. You can’t convey a healthy, vibrant, warm partnership if you spend most of your time with two characters displaying something else entirely; make sure that whatever the core of your relationship is, that’s the primary aspect in a majority of scenes.

Remember that plausibility isn’t likability. Returning to a point we’ve touched on several times: Just because you’ve convinced me that a relationship between player character and NPC is plausible doesn’t mean that I’m going to like the romance or the romantic interest. You still need to do all the groundwork you’d need to do to make me like any character–establishing character, agency, and likability certainly isn’t less important with a character who’s part of a romance.

If you’re trying to build a “bad” relationship, be careful not to simply end up with an irritating one–you very likely want a relationship the player is torn over, uncertain whether he or she really wants it to end. In this case, you still probably need to emphasize the positives over the negatives (in both the relationship in general and the romantic interest)–unlike in real life, it doesn’t take much to sour a player on a relationship; the player hasn’t invested much in it, so why stick around for something bad?

Remember what your game is really about (not romance). If you’re making a first-person shooter, your game is probably about violence. If you’re making an RPG, it might be about exploration or ascending to power. Your game is about whatever your core mechanics are about, flavored and directed or subverted by your narrative. Which means, in most cases, your game is not about romance.

Therefore, make sure that your inclusion of romance in your game fits neatly within–and is subsidiary to–your larger themes. The idea that the power of love wins over all else at the end of your shooter probably isn’t going to fly–you just built 10 hours of gameplay about how the power of shooting people is what wins, and threw a detached romantic subplot off to the side and expected me to care.

Whatever reason you have for including a romance in your game, it needs to have an appropriate place within your larger narrative. Don’t put more weight on it than it can support; and remember that what it can support is defined largely by its integration into gameplay (as we discussed above).

Go vague and archetypal. Want a way to ignore almost all of the advice above and still pull off a successful romance? I’m sure there are methods. Maybe you can play on the power of archetypes–a game like Passage manages to generate emotion without falling back on crutches like “actual characters.” Or maybe you can can lean on symbolism instead. Maybe your “romance” is entirely subtext, and the text allows for an entirely platonic interpretation of a player character and NPC relationship. Maybe procedural generation and clever mechanics can create a compelling romance in the context of a roguelike!

This post spends a lot of time looking at traditional romances in games with mainstream storytelling sensibilities. But there are whole worlds left to discover. Maybe innovation is worth a shot for your project?

Allow branching after all. Finally, here’s something of a Hail Mary pass: Even in an otherwise non-branching game, there’s nothing stopping you from making a romance optional–by giving the player a way to “opt in” or “opt out” of the romance, through whatever method is appropriate to the gameplay. This doesn’t mean you don’t need to make the romance compelling, of course, but it lets your players escape a subplot that doesn’t appeal to them (so long as you aren’t tying the romance to exclusive content or mechanical benefits).

Such a romance, of course, can’t play an enormous role in your narrative–otherwise, it couldn’t be made optional. And we’ll get into all the particulars of more complex branching romances–like what variety of romance options to present, setting player expectations, and offering rewarding non-romance content–another time.

But acting as the player’s judge (and jury, and executioner) is in some respects the primary job of a game’s developers. Moreover, surely all art emerges from the artist’s own experiences and worldview to convey a particular set of ideas. How does all that square with avoiding being judgmental?

I want to dig into that question here, using the notions expressed by Spector and Jaros as a jumping-off point. This isn’t a rebuttal or a reiteration–rather, it’s a different framework for viewing many of the same challenges. We may even come to some of the same conclusions.

Know Your Themes and Your World

Let’s first dispel–briefly–the idea that any game can avoid espousing a particular worldview or moral philosophy. Say we’re developing an open world action-adventure game set in a modern-day city. The player is able to engage any non-player character in combat at any time, and now we’re forced to determine what should occur if the player kills a civilian somewhere isolated and out of sight.

Most games either:

• allow this heinous act and let the player character depart without further consequence, relying on the player’s own conscience to determine the morality of the situation.
• immediately send police officers after the player character, despite the lack of any in-world way for the police to be aware of the crime.

But of course neither of these results is in any way realistic. The problems in the latter example are obvious, but no less substantial than in the former case where one must wonder:

• Why don’t the police investigate the murder at a later date and track down the player then?
• Why doesn’t the neighborhood change, knowing there’s a vicious murderer around who’s never been caught? Why aren’t there candlelight vigils and impromptu memorials?
• Why doesn’t the victim’s son grow up to become Batman?

We construct our game worlds in a way that suits the genre and moral dimensions of the story we want to tell. There’s no right answer here, but the consequences we build into a game are inherently a judgment on the player’s actions. Attempting to simulate “reality” will always fail–we must instead build a caricature of truth that suggests a broader, more realized world. Declaring “in a modern city, murderous predators can escape any and all consequences” is as bold a statement on civilization and humanity as deciding “in the long run, vengeance and justice will always be served up by the victims of crime (metaphorically by means of a bat-costumed hero).”

Knowing that, what’s the world we want to build? What are the themes and moral compass points we use to align our game?

This is a relatively easy task when working with a licensed intellectual property. In Star Trek, we know that creativity, diplomacy, and compassion are privileged above all else, and that greed and prejudice always lead to a bad end. A Star Trek story in which the protagonist freely lies, cheats, and steals without any comeuppance probably stopped being a Star Trek story somewhere along the line. Game of Thrones, on the other hand, takes a more laissez-faire approach to personal morality while emphasizing the large-scale harm done by men and women who strive for power. (No one comes away from watching Game of Thrones believing that the titular “game” is a reasonable way to run a country.)

These core ideals should affect more than your game’s storytelling–they should dovetail with your gameplay loops and systems, as well. A Star Trek farming simulator might be a fun game, but using the franchise’s key ideals to guide narrative and mechanical choices probably won’t be useful. (“Maybe we reward the player for reaching an accord with the corn#8221;)

Know what principles drive your game world. You’re going to need that knowledge for everything that’s coming.

Teach Your Players About Your Worldview (and Fast!)

Of course, you knowing the compass points of your world doesn’t necessarily mean your players will. (Unless you are using a license, they almost certainly won’t.) For most games and most narratives, it’s best to lay out at least the broad strokes of your world’s principles early on; if one were to equate a game with a television show, you’d want your audience to have a sense of the themes by the end of the pilot.

Teaching the player the thematic basics of your world shouldn’t be overly difficult–low-stakes choices, examples of your world and character arcs in a microcosm, gentle words of wisdom, obviously bad advice, and so forth can all help guide the player’s expectations. You can introduce theme in a game the way you would in any medium, so we won’t dwell on that here.

You can, of course, spend a great deal of time exploring the nuances of the moral philosophy of your game world across the course of the whole game. You’ll probably want to. So why is it so important to give the player the right idea from the start?

Because you need the player to buy into the kind of story that you’re telling. To some degree, this is true even in traditional, linear narratives: if I walk into a theater expecting the romcom stylings of The Taming of the Shrew and get Romeo and Juliet instead, I’m not going to be delighted by having my expectations subverted; I’m just going to be irritated.

When you give a player a measure of control over the narrative, the player’s expectations for a certain type of story become even stronger. We’ll discuss this more in the next two points, but don’t allow your player to shoot first and ask questions later in the aforementioned Star Trek game while naively expecting the story to applaud her rogue-ish cowboy ways. Interactive narrative is a collaborative process, and the player needs to be able to make an informed decision when she chooses to drive the story in a given direction. This is the pact between player and developer: “You show me how your world works, and I’ll invest myself in it to the best of my understanding.”

Every Path Should Present a “Satisfying” Story

Once your world is defined and the player understands it, you’re essentially asking “What sort of story do we want to tell together#8221; when you present the player with choices. But what those choices are and how they play out is critical to shaping a satisfying narrative.

First, avoid presenting options that can’t lead to an engaging and stimulating narrative. That doesn’t mean every option needs to hand the player a trophy. What it does mean is that you should avoid wasting the player’s time–if one or more narrative-driving choices results in substantially “better” stories than others, why are the not-very-good options even available?

When you ask the player “What sort of story do we want to tell#8221; it’s okay if the player says “a bittersweet story” or “a tragic story of hubris” and the game obliges. But if the player wants an unhappy story, it still needs to be narratively satisfying. If you’re offering me a choice between The Taming of the Shrew or Romeo and Juliet, great; don’t offer me a choice between The Taming of the Shrew or a YouTube video of someone weeping.

In order to make all your paths satisfying (within reason–there will always be certain combinations of choices that are unpredictably unaffecting or fail to suggest an internally consistent player character), consider your themes closely when developing your choices and mechanics. The ability to murder people on the street might be an appropriate option for a modern-day crime game; the ability to murder any crew member probably doesn’t add anything to your Star Trek game, as it’s unlikely that’s going to lead to a satisfying story worth telling. Use those resources to broaden your relevant options instead. (A Star Trek game might not give me “violent vs. peaceful” choices, but “technical vs. diplomatic” choices, each shaping a different sort of story and outcomes. e.g., you might allow a player to explore the dilemma of a sturdy peace rigorously enforced by technological disarmament vs. a more fragile, bloody peace reliant on the will of all parties involved.)

If you must have dead ends, make them quick to resolve–don’t force your player to experience a dissatisfying story just because he tried something thematically off-key.

Judge the Player Character, not the Player

Making sure that your choices and story paths have equal narrative validity brings us to the question of judging.

In order to determine the results of any given choice, you (that is, the game you’ve designed) must judge the actor according to the dictates (intended or implicit) of the game world and story. If you’re building a game inspired by 1940s comic book Crime Does Not Pay, then in your game world, crime should probably not pay.

But if you’ve set the player’s expectations correctly and made all paths narratively satisfying, then there can be no bad choices on the part of the player–only bad choices on the part of the player character which the player has decided to explore. The player is no more complicit in the (nonexistent) crimes of the player character than an author is complicit in the crimes of her characters. Therefore, there is no reason to attempt to punish or shame the player for “bad” decisions–the player made those decisions to explore the consequences with you, the designer. (Punishing the player character is just dandy, so long as it’s an engaging experience.)

Is it possible to trick the player into thinking an option will lead to a different outcome than it does? Of course, and it’s not particular difficult. You can offer the player a Captain America-style superheroic action story and give her grim-and-gritty Platoon instead, but to what end? Any message you’re trying to send will need to overcome the massive barrier of the player’s irritation at your bait-and-switch. Better to be honest about what the player is getting into upfront (thematically–you don’t have to give away your plot points!) or design a more linear game without multiple narrative approaches (or both).

And of course–as we discussed above–“judging” needn’t be a question of a positive or negative outcome. For combat-heavy role-playing games, most developers try to provide many viable “builds” of player characters for different combat abilities and playstyles; judging a narrative “build” of various choices most likely isn’t about whether the ultimate result is good or bad, but about what form the world and characters take according to the themes and ideas you’ve chosen to prioritize.

Don’t Pull Punches

It’s okay to explore difficult themes without offering up a “correct” answer. It’s okay to let players try out deeds and consequences and decide for themselves what it all means. But don’t forget that the game is rigged.

If you create a world in which consistency and determination win out while moderation and compromise result in disaster, or where technological and diplomatic solutions can both be applied to reach largely the same outcomes, you’re already making a case to your audience. Your game is your study of a particular worldview, and every player choice is a supporting argument. (Even if you throw in choices and consequences that go against the overall thrust of your narrative, you’re only arguing for a more nuanced worldview–not stepping back from the case itself.)

You owe it to yourself and your players to present a vision that’s well-considered and coherent. Once you’ve determined the rules of your world, follow through on them. If you find that means some of the choices you want to present to your player result in uninteresting, non-engaging, or “not fun” outcomes, then you’ve got a problem–but most likely, that problem is either in the basic principles of your narrative or in the kinds of choices you’ve put on offer. If you find yourself “watering down” or skewing choice outcomes in a way that feels narratively wrong in some ineffable way, you’ve likely got the same root trouble.

Intentionally or not, a game judges and a game teaches. It shows, through a multiplicity of possibilities, what might happen if the player does X or Y, and the player learns the unseen rules that underlie your world. Embracing the didactic elements of your work doesn’t mean slapping the player’s wrist every time she’s wrong–it means building a game where the player can play and learn and experiment within the boundaries of the lesson.

There are Always Exceptions

None of this may apply to your project. If you’re building a serious game that’s meant to double as a police procedure training lesson, you’re likely going to have a lot of absolute right-or-wrong answers. Or maybe you want a game with an absolute narrative score, where more points correspond to a better ending.

That’s okay. There are no rules. But then, I’m here to defend didacticism–not ensure its absolute triumph.

This is neither a purely mechanical question (“non-player characters scattered through the open world will offer quests, highlighted by giant floating question marks over their heads”) nor a purely story-driven one (“the player is a one-man army, so he’s always getting orders and advice from military commanders”). It’s a question that needs consideration from the very beginning, and ties into everything from the most basic game mechanics to your choice of intellectual property.

Let’s talk about one particular way of thinking about this problem. Let’s talk about story engines.

What’s a Story Engine, and How Much Will It Cost Me?

The term “story engine” is commonly applied when discussing episodic television. It’s the part of a show’s premise that allows a new story and conflict to be created every week–stories and conflicts different enough to cover a wide variety of topics and situations but similar enough to feel like natural extensions of the show’s core. Comedies and dramas both use story engines–“mismatched office workers must serve under a wacky boss who’s constantly causing trouble” is as valid as “monster-hunters live over a small town’s monster spawning grounds” or “explorers in space investigate and learn about a new planet every week.”

Not actually a Story Engine (from The Legend of Grimrock)

These are, of course, simplistic summaries–the “real” engine (note: there is not a real engine) is more complex–but it gives an idea of what we’re discussing. Most good story engines can actually generate stories through a few different methods (“sometimes the space explorers go to war with aliens, or experience conflict within their ship’s crew”), and not every story needs to emerge from the story engine (“this week, in a special episode, the detectives rally around a colleague whose mother has cancer”). Nonetheless, a strong story engine is what allows an episodic narrative to run for one hundred entertaining episodes instead of ten.

With our terms defined, let’s turn to games. Starting with…

What games benefit from strong story engines? Your linear, level-based first-person shooter probably doesn’t need much of a sustainable story engine. (It might! But it probably doesn’t.) Your 40-hour role-playing game, your quest-driven MMO, and your open world action-adventure set in a city packed with tasks and objectives probably do.

Basically, if your game has a lot of content–particularly if that content is consumed in discrete components (such as designated quests or missions)–and that content is offered in a nonlinear fashion, you probably want to think about your story engine. With enough content delivered over a long period–e.g., a game with regular downloadable updates–even linear games may benefit (as we’re essentially in the television model again).

How does a game story engine work? A basic story engine should do the following:

Generate diverse player-oriented stories and conflicts. This may be a passive part of the story–the setting’s background might handle this (“in Raceworld, every citizen wants to win the daily 100-meter dash… no matter what it takes!”), or it may be integral to the overarching plot (“the player’s goal is to become the most popular kid in school by winning over her classmates one at a time”). The generation mechanism could also be active or mechanical (e.g., in a faction-driven game, a player’s progress might always anger one faction or another and create conflict).

It’s easy to end up with a story engine that’s too limited or too generic. Can you really get 50 stories out of a lone wolf’s battle against mindless alien invaders? Does your protagonist’s quest through Under-Earth for the seven mystic talismans actually set up anything useful, or have you simply given yourself a framework without ridding yourself of the burden of introducing a new story with every quest? How is your resource-scarce society constantly scrabbling for food going to generate player-focused stories?

Use a reliable mechanism to engage the player with those stories and conflicts. This mechanism must be supported both by game mechanics and narrative. It may be extremely simple (the player character stumbles upon NPCs with dialogue that open up the story), practically invisible (a pop-up tracker appears when entering a new area, revealing a set of objectives), tightly focused (the player is encouraged to regularly return to his house, where letters arrive in his mailbox pointing to new content), or incredibly intricate (the player can watch crowds of NPCs to spot individuals exhibiting unusual behavior, then follow said NPCs into story areas).

This delivery mechanism has enormous impact on the type and complexity of the stories you can tell. If every piece of content is introduced through a lengthy conversation with a new character, your central conflicts will be very different than if content is introduced via a series of subtle breadcrumbs or a small group of NPC supporting characters. Quest format dictates, to a large extent, quest content; and how you open a quest is an important part of that format.

Allow the resolution of said stories and conflicts through normal gameplay. This one is pretty obvious, but still worth noting: If the conflicts your story engine generates are all about reaching a given destination in a short time, don’t pair them with a game oriented around patient, stealth-based gameplay. If your story engine creates stories about political maneuvering, don’t use it for a game without a conversation mechanic. And if your game is oriented around combat, don’t build a story engine that generates intricate meditations on philosophy and diplomacy.

In the classic action / adventure / strategy game Star Control 2: The Ur-Quan Masters, the player explores and fights through an occupied galaxy to find alien allies against the occupiers. Naturally, different aliens have their own troubles with the occupiers that must be resolved before an alliance can be struck; these troubles can be solved by further exploration (finding more allies, discovering odds and ends on lost worlds, etc.) and combat.

This is a simple, elegant means of generating stories appropriate to the game. It works if there are three alien races to find and it works if there are fifty. It keeps the player at the center of the story (the aliens, being oppressed, can’t organize on their own) and allows for numerous variations (some of the aliens have sinister agendas!) It feels natural. It feels obvious.

Upgrading a (non-story) engine in Star Control 2

But tweak it just a little bit and the engine suddenly becomes weaker. Change all the aliens to humans from a single society and your story variety drops considerably (unless you spend a lot of upfront effort detailing that society). Replace the imperative to make allies with the more vague need to stop the occupiers and the player lacks clear motivation and drive when encountering new characters. Change the exploration-based discovery of content to distress calls that can reach the player at any time and you’ve suddenly got the player flailing to track a dozen threads at once.

(In fact, Star Control 2‘s story engine is far more elegant and far less simple than this, but we’ll save that for a different post.)

It shouldn’t be hard to test a story engine’s basic functionality. Look at your game’s setting, story, and mechanics as they stand. Can you crank out a dozen very short and very interesting quest pitches pretty much immediately? Are they all different? Do they feel true to your narrative and not overly generic? Do they not require excessive setup? Then you’re at least on the right track.

Can a procedural content generation system work as a story engine? Maybe. A good procedural content generation system probably does double as a story engine (see, for example, Shadow of Mordor‘s Nemesis system). It should create a diverse array of conflicts that all naturally emerge from the game’s premise and readily motivate the player to engage.

But story engines are only a mental model for the creative process–they’re not a formula. Identifying your story engine doesn’t mean you can build it as an actual piece of software.

What are some advanced story engine features? In addition to the modular, episodic-style content that story engines are built to create, most games also have a larger-scale throughline–a roughly linear overarching plot of some sort. A good story engine should feel like the natural consequence of this plot–conflicts arise because of the player’s overarching goals, and resolving them pushes the player down the overarching path. (If the player wants to free the Earth from alien invaders, every mission to shoot aliens is a means of progressing. If the player wants to become the Avatar of Virtue, then every encounter with townspeople in which she can demonstrate truth, love, or courage contributes to the main plot.) A poor story engine creates stories tangential to the main plot (“Your goal is to kill the king, but you’ll need money to equip yourself. Fortunately, there are hundreds of ways to make money in this city…”)

In more complex games, stories engines can also nest. Some conflicts and stories are larger-scale than others–a top-level story engine may boil down to “The player must bring peace to a land in anarchy after a war by traveling to hotspots and preventing them from erupting into total chaos; the player receives leads on each hotspot from the heir to the throne back in his home garrison”; while each hotspot might have an engine of its own focusing on local concerns and using different delivery mechanisms.

Similarly, a game might also use multiple story engines that run in parallel but open the same content. This can be a distracting mess if implemented poorly or utterly invisible if handled well. In the Star Control 2 example, a narrative-light resource gathering game emphasizes the need to explore. The process of resource-gathering naturally brings the player into contact with potential alien allies, while returning with resources to a central location triggers pointers to content the player may have missed.

Advantages of Using a Story Engine

In truth, you’re probably using a story engine already–it’s just a question of whether you think of it in those terms. But by tuning the engine’s elements, here are some of the clearest benefits:

Storytelling Shortcuts. A good engine helps cut out the setup of every “episode” of story and get right to the conflict. With many of the story elements already clearly defined and with the player character already motivated, every story can focus on what’s unique and interesting rather than getting bogged down in exposition.

Familiar Form. Expectations are powerful. An audience who understands that a victory taking place too early in a story will soon turn sour feels a dread and tension that adds to the experience. Knowing an audience’s expectations also allows a writer to subvert those expectations for occasional additional effect.

When a player understands the form and themes of a game’s episodic stories, it allows a game writer to design to that experience. A player’s excited expectation when the traditional quest-delivering radio call comes in is a powerful tool; whereas a game that delivers content in a different way every time denies the player that particular sensation.

Episodic Structure. Content-rich nonlinear games tend to be episodic even when the designers don’t intend them that way. In practice, a player’s play session is likely to be bounded by the length of one or more “quests” (or other relevant story components)–that’s the pattern he falls into, starting at the beginning of one story and taking a break at the end of another.

In fact, for many players, the overarching storyline will be secondary to the episodic stories–it’s rare for the majority of players to actually complete a given game (there’s lots of data on this, but see, for example, here). By focusing your game on these episodic stories and building a strong engine, you’re optimizing for the player’s true experience. You’re writing a book of short stories instead of a novel because you know your audience won’t read everything.

See also “Designing Stories for Nonlinear Game Segments” for additional discussion of this and related issues.

Modular Content. Game development is an uncertain art. Content gets cut and content gets added with impressive regularity. If you’ve got a strong story engine, you should be able to crank out additional stories without enormous difficulty when the call to add 5 more hours of gameplay comes along; and when your game needs to be trimmed for cost, you know your stories are modular enough that you can afford to cast them into the abyss.

A Few Examples, Please?

Let’s briefly consider a few examples of story engines in games and how they succeeded (or didn’t).

Mass Effect vs. Mass Effect 3. You might expect two games in a series featuring the same protagonist to use the same story engine. But while they have many similarities–including the player character’s role as a sort of militarized space cop (in the Lensmen / Jedi / Green Lantern mold)–these games spin up their content in dramatically different ways.

Still not a Story Engine (Mass Effect 2, via masseffect.wikia.com)

In Mass Effect (BioWare, 2007), the player’s primary objective is to hunt down a rogue officer involved in a mysterious galactic conspiracy involving ancient but powerful technology. You may notice that this is an extremely specific objective, which is both boon and bane. It gives the player clear direction for the main plot, but–as we’ve discussed–the “main plot” is generally not the majority of content for a game of this type.

Mass Effect integrates its side quests (the real test of a story engine in an RPG) with limited success. Most have limited relation to the main plot, instead relying on the player’s status as space cop to involve him in smuggling schemes, salvage missions, and diplomatic entanglements. Most are received by stumbling into the correct NPCs at the right time and place. For the most part, they feel appropriate to the universe–but they always feel like side stories and often a bit poorly motivated. After all, I’ve got a conspiracy to investigate. At their worst, they’re generic and impersonal.

Whereas Mass Effect 3 (BioWare, 2012), by focusing its story engine, creates a much more integrated feel. In this game, the galaxy is at war, and the player’s role is to lead a coalition to victory. The vast majority of ME3‘s side quests contribute directly to this goal–the player recruits new allies, acquires weapons, makes strikes at enemies, and so forth. The motivation for engaging in these missions is immediately clear; each story can focus on what makes it interesting rather than on giving the player a reason to even care.

Mass Effect 3 also tunes its story engine when it comes to mechanics. A universal “Galactic Readiness” rating rises whenever the player aids the war effort. More quests than ever are delivered to the player directly from her home base–delivered through calls or e-mails to her ship. The story engine creates content that is narratively appropriate, immediately accessible, and mechanically rewarding.

Sunless Sea. In Sunless Sea (Failbetter Games, 2015), the player is an explorer and merchant sailing through a mysterious world: discovering new islands and ports and meeting their strange inhabitants. The player is given very little overall motivation (a selectable “Ambition” determines certain win conditions but otherwise plays no substantial role) and there is no main storyline–all quests, in effect, are side quests. Such a game requires an effective story engine.

Sunless Sea succeeds largely by virtue of the breadth of the type of stories its engine generates. The game’s world includes steampunk technology, Lovecraftian monsters, sorcery, fire-and-brimstone-style demons, Victorian romances, pulp-style talking apes, and a hundred more oddities besides. With so many toys in the toybox, every side quest is able to offer something new–the player is unlikely to grow tired from seeing the same themes too often.

Of course, that same breadth and lack of coherence means that a player may lack the motivation and emotional stake that games like Mass Effect provide. More than once I encountered a questline in Sunless Sea that appeared intriguing, but which I had no compelling reason to involve myself in. (“Why would I risk my life to do that#8221;) Fortunately, there were plenty more questlines to choose from.

Sunless Sea also makes a strong effort at integrating its story delivery mechanisms into the natural course of gameplay. In a game so devoted to exploration and discovery, it’s natural that the most exciting moments of discovery (finding a new island, say) are most likely to deliver a large number of new stories; but even the obligatory tasks of refueling at ports and selling off cargo can result in new quests, thanks to a mechanic that “refreshes” opportunities after time away from a given island.

Look, this was a hard post to illustrate. Okay? (Sunless Sea)

If we were to make a television analogy, Sunless Sea is closer to Doctor Who (wildly careening from genre to genre, largely disintereted in ongoing storylines, and purposefully inconsistent in tone) than Mass Effect 3‘s Battlestar Galactica (tightly focused thematically but able to explore numerous aspects of its constrained setting).

Dismantling the Engine

A story engine is a thought experiment–a creative tool, useful for modeling narrative possibilities and generating inspirational templates. It is not prescriptive or a necessary step in anyone’s process. So why even bother?

Because models can be useful, and help constrain creation to the realm of the viable. The more people on a team, the more templates can lend clarity as to intent. And because it’s good to test whether you can create a thousand interesting side quests before diving in.

Let’s talk about empathy and the moral responsibility of the writer.

Suppose you’ve got an audience of a million people for your AAA video game, broadly representative of the American population (because these days, everyone loves games!) As you hash out your plot, think about your audience for a second and ask yourself: Are you comfortable with the story you’re telling to them and about them? Would you be prepared to sit in a room and justify your artistic choices, one-on-one, to your players?

• If you’re not Hispanic, are you prepared to justify to a Hispanic person how you cast (or didn’t) Hispanic characters?
• If you’ve never been a sex worker, are you prepared to justify to a sex worker your story’s handling of strippers and prostitutes?
• If you’ve never been in the military, are you prepared to justify your war story to a veteran?
• If you’re a man, are you prepared to justify to a woman the casting and portrayal of women in your story?
• If you’ve never been in a gang, are you prepared to justify your crime story to a gang member?
• If you’ve never been disabled, are you prepared to justify your handling of a paraplegic character?

These aren’t the only people and experiences you should be thinking about–this is a grab-bag of “things games have historically handled poorly.” But you see the pattern. I began writing this post to offer a rule of thumb for sensitively handling sexual violence in games, but then decided such a rule has much broader applications; “could you justify this in a one-on-one conversation” is an approach that works for so many things.

If the thought of people in the above groups (or the group of your choice) engaging with your story makes you uncomfortable… think hard about why that is. Most folks can handle opposing viewpoints, but are much less easygoing about being portrayed inaccurately, or as a stereotype, or as something less than a three-dimensional human being. Or if you can’t picture that personal conversation at all, why not? If you can’t even start to form that image, surely you don’t know enough about Group X to do it justice in your portrayal.

Of course, the fact that you are prepared to justify your handling of Topic Y doesn’t mean you’re right. Maybe you feel your over-the-top depiction of gun violence is so far in the realm of fantasy that it doesn’t convey any real-world message; maybe you did your research about life under the Khmer Rouge and you’re pretty sure you’re conveying as accurate a picture as you can. Ultimately, you’re the one who has to make the call. You could still be wrong. But be sure you’re as certain as you can reasonably be.

And if you do think you might be wrong, consider this as well: Are there other voices out there–other games, other narratives–that convey an alternative viewpoint? Or are you the only one showing characters of this particular type at all? In the former case, you can afford to be a little more slipshod; not that you want to neglect your responsibilities to truth, but if you fail, the consequences are mitigated. In the latter case, you’re potentially doing something useful but you also have a special level of responsibility–bad representation isn’t always superior to no representation.

And what if your portrayal is a common one, but not one derived from personal experience? That is, what if there are plenty of other narratives out there featuring characters of Background Z, but they all tend to treat them in a similar fashion? Again, the harm you inflict by going along with that trend is probably pretty minimal even if it’s an unfair portrayal–but it doesn’t excuse you from taking time to reflect on your writing, or from doing your research.

And that’s all this post is really about: Reflection, research, and empathy. Cover those three bases and you’ve done what you can. You can’t ever really speak for anyone else (no matter how similar to you that person may be), but that’s the mantle you’re picking up all the same, every time you write any character.

The least you can do is try to do people justice.

For four decades, the adventure game genre has been practically synonymous with “story-driven” video games. In the early 1980s, Infocom was trading in thought-provoking, experimental text-only narratives while pioneering role-playing game series like Ultima and Wizardry were still fumbling with the basics. (We won’t even talk about contemporary narrative in other genres.) By the 1990s, non-adventure games had become more competitive in the storytelling sphere, but companies like Sierra and LucasArts were still where gamers went when they wanted “pure” interactive narrative that eschewed complex combat mechanics or action sequences.

Times have changed. The adventure game genre is still around, but is no longer a dominant industry force. Now even first-person shooters and other action games are marketed with a strong emphasis on story.

What hasn’t changed is the way story and gameplay are separated in most games. We continue to see games structured as a sequence of story moments broken up by gameplay challenges (or vice-versa), whether those challenges are comprised of elaborate combat systems, exploration and discovery, or object interaction puzzles. Sometimes the line between story and gameplay is perfectly blurred or the transition made seamless; more often, it’s stark.

In traditional media, conventional wisdom says that every element of a narrative piece (whether a line of dialogue or the color palette of a theater set) should advance a story’s plot, characters, or themes in some way. It’s a high aspiration and, I think, a worthy one. It’s something that games can aspire to, as well, in how they craft their challenge-based gameplay.

Another way to look at how well gameplay integrates into narrative is to ask, “If I were to adapt this game’s story to another medium, or recount the story to a friend, would I mention this challenge? Or would I skip over that part altogether#8221;

As a genre that, broadly speaking, aims to provide extremely high density story delivery, adventure games make for interesting examples in gameplay integration. That is, adventure games often pride themselves on eschewing extraneous, repetitive gameplay in favor of narratively integrated challenges; but even within adventure games, some challenges suit the narrative far better than others.

Let’s look at a few examples of traditional adventure game puzzles and see what we can learn:

The Longest Journey – Rubber Duck. The duck puzzle from The Longest Journey (Funcom, 1999) has been examined and excoriated more than once. It’s notable not only for its artificiality but because it’s a particularly bad fit in a game that makes a real effort at maintaining a consistent tone to the larger narrative.

I won’t break down the puzzle in detail (the preceding links do so thoroughly). Suffice it to say that The Longest Journey‘s protagonist sets out retrieve a key from an electrified subway track. Her motivation for acquiring the key is poorly established, at best–it unlocks a fuse box she wants access to (in order to create a diversion), though she has no way of knowing this–and her method involves building an elaborate contraption using, in part, a rubber duck.

But forget the overcomplicated solution to the puzzle–that’s absurd on its surface. More importantly for our purposes, what does retrieving the key have to do with the larger narrative? Does it advance character or theme? Would you choose to show it in your film or novel adaptation of the story?

Does it even make sense within the context of the game world? Are there simpler or more character-appropriate ways to accomplish the same goal? This is a less obvious trap than the sheer complexity of the puzzle–how many game protagonists go to elaborate lengths to move crates, murder foes, or retrieve that one special object when buying a ladder, making a phone call, or getting a replacement special object would be at least as simple?

Loom – Dyeing Sheep. Loom (Lucasfilm Games, 1990) fits into a subtradition of adventure games in which the primary game mechanic represents some flexible “power” on the part of the protagonist, and where creative use of this power is valued above object interaction puzzles (of the “use chewing gum on leaky pipe” variety). In Loom, the power is an array of magic spells the player character learns over the course of the game; I view Loom as a successor to games like Enchanter (Infocom, 1983) (another game of spellcasting) and predecessor to adventures like Life is Strange (Dontnod Entertainment, 2015) (in which the protagonist manipulates time itself to solve troubles).

By centering both narrative and gameplay on a single concept (spellcasting, time manipulation, etc.), producing gameplay challenges that feel narratively integrated becomes much more straightforward. But that’s no guarantee that every challenge will feel perfectly appropriate. One of Loom‘s more artificial puzzles involves a shepherd who keeps losing sheep in her flock to a hungry dragon. To solve the problem, the player must cast a dyeing spell–a spell earlier learned through, essentially, random examination of a cauldron of dye on a mage’s island–to turn the sheep green and thus camouflage them in the field.

It’s not a downright unnatural solution, and it has a certain fairy tale logic, but it lacks the straightforward feel of using magic to disguise oneself or using a simple “close” spell to seal a rift. Loom is not a game about an archetypal trickster using lateral thinking to solve problems in unconventional ways, and thus, for me, the sheep puzzle falls flat from a narrative perspective. Even with a narratively integrated set of mechanics, challenges using those mechanics still need to clear certain bars for optimal synthesis.

Gabriel Knight: Sins of the Fathers – The Drums. Recently remade in a “20th Anniversary Edition,” Gabriel Knight: Sins of the Fathers (Sierra On-Line, 1993) is a game better remembered for its engaging supernatural mystery storyline and strong characterization than for its puzzles. In truth, its puzzles are something of a mixed bag, combining some highly traditional pixel-hunting, “find and use this arbitrary object in an unusual way for unconvincing reasons”-style puzzles with other puzzles that are integrated into the narrative perfectly.

Gabriel Knight‘s drum “puzzle” is one of my favorites in adventure game history. The protagonist is effectively an amateur detective, trying to unravel the mystery behind a series of murders in New Orleans. From day one of the storyline, the player encounters (in the background) drummers playing on the streets. On day four, the player spots an informant talking to one drummer and learns from him that the drummers are part of an information network, sending messages via coded drumbeats. Research turns up the particulars of the code, and by day six, the player is able to decode the messages and infiltrate a gathering of the murderers’ cult.

But there’s more to it than this. The drum puzzle intersects with another puzzle involving the cult’s written messages–publicly displayed in plain sight like the drums, encoded as graffiti. As part of this sequence of events, the player must alter one of the written messages to manipulate the scheduled gathering. The need to send an additional drum message comes up yet again later on.

What makes the drum puzzle so effective from a narrative perspective? Let’s look at three things:

First, it resonates with and further develops the character of the protagonist. It involves our amateur detective unraveling a mystery that wouldn’t feel out of place on a television procedural, honing his skills and applying them in more advanced ways. The alteration of the coded message emphasizes the protagonist’s status as a manipulator–a liar and a scoundrel–even while putting those skills to benevolent use.

Second, the drum puzzle builds on elements introduced into the setting long before it becomes game-relevant (the drummers) and continues building on those same elements after the first stage of the puzzle is complete. The existence of written coded messages is a logical extension of the musical communications network. The drums aren’t simply a component of a puzzle, but a thread in the overall story.

Third, the puzzle contributes significantly to the mood of the game. The revelation that the murderous cult is operating in plain sight, watching and communicating with its agents in daylight is effectively creepy. The fact that the player has seen non-gameplay-relevant drummers before helps build a sense of oppressiveness and paranoia.

In other words, we have here a puzzle that advances character, story, and theme. A genuine rarity!

I want to stress that a failure to perfectly integrate gameplay challenges into narrative doesn’t make a game bad. But it’s a certainly a detriment to the narrative experience and should be evaluated as such.

We could take the same approach with quests in role-playing games–does a mission to gather gemstones in a dangerous ruin to open a magically sealed door really fit the overall narrative, advancing story, character, or theme in a meaningful way? What about the first combat encounter with skeletons in the wizard’s tower? What about the next twenty combat encounters? If you were adapting your game into a novel, would you leave these sequences in?

Even an action-adventure game or a story-driven first-person shooter can be broken down in this way, with each objective, level, or gameplay sequence justifying itself (or not) on the pure narrative level.

Is there a way to maintain the challenge and depth of gameplay content while still fully justifying it? Can every puzzle be a secret drum message puzzle?

Let’s try to find out.

Or, more likely, not.

The notion of killing someone close to the protagonist to kick off a story is an old one. In traditional media, most of the time it’s not really meant to be a motivation that’s of particular interest to the audience: we’re not expected to be too distraught over Hercules’s children, or the Punisher’s family. Instead, we’re meant to accept the motivation as valid and move on–it’s just enough to give the protagonist an excuse to do what he (with occasional exceptions, it’s usually he) does while we enjoy the resulting twists, turns, and carnage. The inner pain driving the protagonist isn’t the subject of the story–what matters is what he does with that pain, and the consequences that ensue.

It’s not the most intellectual or humane way of telling a story, but you know what? Sometimes it works.

It doesn’t work in a video game. We’ve talked before about the importance of converging player and player character motivations, but to recap: The closer the player’s motivations are to the player character’s, the more engaged you can expect the player to be in the narrative. Wanting the same thing (beat the boss!) is good, but wanting the same thing for the same reason (beat the boss because he’s so, so irritating!) is better. When the player’s motivations are unlike those of the protagonist, you lose one of the unique strengths of the medium and–assuming you want a compelling story at all–you burden yourself with the need to craft a dual-track narrative that functions to motivate the player and player character separately. (A topic potentially worth tackling some day in a separate article.)

You can’t motivate a player by killing someone she’s never met before, or who she’s seen only in the briefest of cutscenes. That much should be obvious. As a consequence, if the player character is supposed to be distraught by the loss of someone dear early on, you can almost guarantee the player won’t be. Your story won’t fail on a technical level (in the sense that all characters can still act reasonably based on their internal motivations), but you’re raising a barrier to the player’s actual interest and involvement in the narrative.

This peril can arise in less contrived scenarios, as well–any situation in which a player character’s motivations are driven by something the player hasn’t experienced is a risk.

Let’s move from theory to something more concrete, and examine some different scenarios for a theoretical action-adventure game we’ll call Drug Murderer: Yes, Thou Shalt Totally Murder.

Scenario One: The player character, Karla, is a loyal member of the Red Panda gang. Within the opening half-hour of gameplay, a fellow Red Panda betrays her and kills her best friend, sending her on a one-woman rampage to defeat the gang that was once her surrogate family.

With a full half-hour in which to get to know the Best Friend, the player might actually care when he dies. But the player isn’t going to care with anywhere near as much fervor as Karla–and most likely, the player’s anger will drop off very quickly without reinforcement. At best, this can serve as an inciting incident to a larger plot in which new motivations can be introduced. At worst, it’ll be entirely ineffective.

And that’s assuming that we get a solid half-hour demonstrating that a) Karla and the Best Friend have a believable relationship, and b) the Best Friend is an appealing enough character that the player will actually like him. Fail to devote enough time and effort to either of those tasks and failure is a foregone conclusion.

Scenario Two: As Scenario One, but the game’s development team has decided that Scenario One’s perfunctory motivation isn’t going to work. Still, that opening sequence has already been designed, the cinematics have been motion captured, and so forth–the dead Best Friend stays.

As a corrective, once Karla’s rampage begins, she now quickly discovers that the Red Pandas have acquired a nuclear weapon and intend to detonate it downtown in return for payment from a terrorist group. Karla’s motivation now shifts from a simple quest for vengeance to a desire to save the city.

This is a nebulous, impersonal motivation, but it gives something for the player to experience and latch onto. Handled with delicate care (and with effort put in to make the player invested in the city), it might work; at the very least, it gets us past a severe Karla / player divide, where Karla focuses only on the Best Friend and the player has nothing driving him other than “finish the game.” A hand off in which a non-functional motive is exchanged for an adequate one is better than not making the swap at all.

Scenario Three: As Scenario One, but instead of trying to hurry past the initial motivation and replace it with something else (as in Scenario Two), the development team decides instead to make the death of Best Friend a turning point for the entire game.

The story is realigned so that for the first third of playing time, Karla and Best Friend are partners inside the Red Pandas. The death of Best Friend doesn’t happen until well after the point where the player is sold on the character and the relationship–once the player has bonded with Best Friend, killing him off has the same effect on the player and on Karla. This approach is simple and potentially very effective.

Scenario Four: After much debate, the development team has decided to get rid of the Best Friend motivation altogether and try to find something less dependent on a particular moment and character. Now Karla is an enigmatic figure with a hatred for the Red Panda gang and a depressed and detached affect. Over the course of the game, the player discovers the root of Karla’s present state–she was once a DEA agent who was held hostage by Pandas, locked away in isolation for years. Now she’s sworn vengeance on the group.

This really isn’t appropriate for most games. By intentionally obscuring Karla’s motives, the player is actively prevented from empathizing with Karla even as he works to accomplish her goals. The player is being asked to put himself in Karla’s shoes, to strive (through gameplay) to do what she’s trying to do, without understanding why.

Scenario Five: In this scenario, Karla’s past is the same as in Scenario Four, but her time as a hostage is shown in a non-interactive introductory sequence. This is an improvement–at least the player can understand what’s going on–but it’s not as effective as it could be.

Scenario Six: Instead of the non-interactive introductory sequence from Scenario Two, the player actually plays through short segments from Karla’s time as a hostage. These would need to be intense to be effective, but the job of this game is to convey a taste of the feeling of isolation and helplessness (as opposed to building a complex relationship, as would be required for the Best Friend scenarios). It’s still a challenging narrative and a tough protagonist to work with, but the approach becomes more viable this way.

Alternatively–or in addition–the player could spend time with the depressive Karla post-hostage-scenario but before she begins her murderous rampage (e.g., dragging herself through her day job at the DEA, talking to therapists, etc.) While it’s rarely as effective as creating emotion and motivation through direct experience, players can respond to a “character tutorial”–learning what sort of character they’re playing, how she reacts and how she thinks, and establishing the limits of behavior and agency–in the same way they respond to a tutorial that teaches game systems.

As with a mechanical tutorial, a character tutorial should come early and shouldn’t challenge or confuse the player. It should ease me into the character I’m playing, clearly delineate her motivations, and allow me to gradually buy into her personality, laying out the baseline from which I’ll work going forward and helping me adopt the character’s mindset.

There’s no single right answer among these scenarios. They all have strengths and they all have weaknesses. For most types of games, some will clearly be more appropriate than others, but there are always exceptions to wrestle with.

Nonetheless, it’s imperative that game writers are able to see these strengths and weaknesses instead of relying on the tricks of non-interactive media. Because if a writer (of all people!) isn’t aware of the peculiarities of interactive narrative and able to steer a game’s development appropriately, who else will?

Games–especially open-world action-adventure games and role-playing games–tend to have extremely large casts. This is partly due to length (you can cram a lot of characters into a 40-hour game), but mostly due to mechanics: If every quest requires a unique quest-giver, or if every town needs to be populated with a dozen or more conversable NPCs, or if every item seller needs dialogue attached, you’re going to end up with vastly more characters than the narrative really warrants. When it comes to speaking roles, a game like Dragon Age: Inquisition makes the entire Lord of the Rings trilogy look intimate and focused by comparison.

Let’s forget mitigating this fundamental issue for now (and there are ways to design games to reduce this problem) and focus instead on how to handle all those characters. We’re not just dealing with a cast of thousands, but a cast of a thousand walk-ons–characters who appear for one scene, serve one purpose, and then may never be seen again.

These characters still need to be interesting and engaging. They need to establish themselves fast–within a handful of words, not by the end of a full scene. They need to engage the imagination of a player who may not be fully focused on the narrative. And they need to be different. So in the space of a small number of lines, how do you make every character feel unique and memorable? If most characters can be distinguished in at least, say, two of the three following aspects, you’re probably on the right track.

Voice

When I talk about character voice, I’m not talking about actual voice acting. Instead, consider all the aspects that determine how a character speaks: Education, class, word choice, syntax, and so forth. A character who speaks in short, to-the-point phrases using precise but not obscure language has a wildly different voice than a character who speaks in lengthy, ultra-formal paragraphs. Consider these variations on one line:

• “I wouldn’t have killed her, and I didn’t. Doesn’t mean she’s worth crying over.”
• “I did not kill her. I do not mourn her.”
• “I had no involvement in her death, but I grant you this–I’d relieve myself on her grave, given the opportunity.”

They’re all (probably) awful people expressing the same sentiment, but they sound dramatically different. When characters speak in the same voice, it’s one way they can end up blending together in the audience’s mind.

It’s easy to go over the top with character voices and end up with a parody instead of a natural speaking style. It’s also easy to default to a few specific voices (e.g., working class and upper class). Not every character needs to be instantly recognizable from her speech pattern alone, but if you’re falling back on the same few styles over and over, you probably need more variety.

Presentation

I use “presentation” mostly to refer to physical aspects of the character. Age and ethnicity and accent and choice of hat aren’t personality traits, and they don’t excuse a bland NPC–but when you’re dealing with a large cast, they’re effective signposts to differentiate characters in the player’s mind. I’m going to be a lot more likely to mix up two middle-aged white male security officers than I am to mix up a young black female security officer and an elderly Asian male security officer.

This is true even if (especially if!) they have the same basic voice. Physical traits are big, clumsy, and effective differentiating features. (As a side benefit, differentiating characters in this way also gives you a more diverse cast.)

Attitude and Context

Just because characters have similar voices doesn’t mean they’ll actually end up sounding the same. Two characters with similar backgrounds and speaking styles may sound drastically different depending on their respective reactions to what’s going on, to the Player Character, and so on.

An educated, absent-minded shopkeep may be thrilled to see the player if he’s had no customers all day; or he might be irritated to have his day interrupted, if he only wants a few moments of peace. A soldier under fire may be gung-ho or terrified. If all of your characters default to emotionally neutral, they not only all feel the same–they also end up feeling as if they have no lives or history beyond the purpose of their one scene.

When writing a game with large numbers of minor roles, it can be well worth the time to select a few lines at random from a variety of characters and put them all side-by-side. Without focusing on content (that is, specific people or events or quests those lines mention), can you tell who’s speaking just based on the tone and style? If you can’t, you’ve got some work on your hands. Consider how you can adjust NPCs’ voice, attitude, and context to make each one unique–and, if not the star of the game, at least someone who’ll catch a player’s interest for a little while.

Additional Reading: The great Patricia C. Wrede has articles on similar topics (from the perspective of a prose writer) here and here.

That’s not a statement that should need much explanation, but in brief: Storytelling is communication. It’s about reaching out to strangers and presenting ideas to them in a compelling, resonant fashion. You wouldn’t write a television show for 10-year-old American children the same way you would write for Japanese senior citizens. You likely wouldn’t refuse to revise your Broadway-style musical if most of the audience wandered out midway through your dry run.

Game metrics are nothing more or less than a way of increasing your understanding of your audience. They’re not a magical means of producing an engaging story–no one’s suggesting that if players respond positively to orcs and plot twists, every game should feature an orc Player Character and a major plot twist every 15 minutes–but they are a powerful means of understanding how players experience narrative.

Sometimes, you may want to adjust how you deliver your story to better match data–if a large percentage of users tend to play in short sessions, you may want to condense your narrative into bite-sized episodes. Sometimes, your goal will be to change the data–if one of your three sidekick characters almost never gets used, maybe you’ll decide to dial back that character’s abrasive personality. And sometimes, you’ll decide that the strength of a story element is worth less-than-ideal results–but you’ll do so knowing what it costs you in audience engagement, and be prepared to compensate elsewhere.

Below is a short list of example data points you may want to know to help craft a better narrative. Often, this data won’t be available for your game until relatively late into playtesting, limiting its usefulness–but it may well be available for earlier similar games (previous games in a series, games built on the same engine, etc.) And it’s never too early to start thinking about what you did right and what you did wrong this time, so you can put that knowledge to work for the next game…

1. Know Play Session Length

From the start of any project, it’s important for designers to have an estimate of their game’s typical and optimal play session lengths. This applies to narrative as much as any other aspect of design–no one would suggest that 5-minute YouTube micro-dramas should be written the same as a multi-season television epic, or that a multi-season television epic should be written the same as a “dump the whole series on the viewer at once” Netflix original.

If you can’t guess at how long an average player will be staring at the screen, you can’t effectively break down your narrative into individual elements (quests, scenes, rooms, whatever); you can’t pace plot and character arcs appropriately; you can’t know how often to remind the player of the story so far; and so forth.

And what’s better than estimates? Actual data. Demand it, love it, and use it.

2. Know Your Target Demographic (and Actual Demographic)

Who are you writing for now, and who are you trying to reach? Demographic data can be invaluable when helping you determine style and subject matter (if you’re writing for adults, you can use a more advanced vocabulary than with children; if you’re writing for millennials, your gritty drama about the Weather Underground will need more exposition than if you’re writing for Baby Boomers) and of course representation (if you’re trying to grow your disproportionately small audience of Latinos, maybe consider how many Latino characters actually show up in game).

That’s all obvious–albeit vitally important–stuff. But demographics also impact other metrics. Suppose your current audience is mostly college students. Your storytelling is optimized for lengthy, intense (movie-length, at least) play sessions. If you’re trying to reach an audience of adults in their thirties, you may be dealing with a lot of parents of young children–which could mean that you need to change your storytelling style to support shorter play sessions from people apt to drop the game suddenly to resolve one domestic crisis or another.

3. Know How Many People Finish Your Game

Related to but distinct from the issue of play session lengths is the question of how many people actually finish your game. Any number of factors may cause a very high rate of non-completion (and the impact on narrative–as well as narrative’s impact on that rate–is a subject in its own right), but you can’t begin to respond to this issue if you don’t know the numbers.

Consider also how these numbers impact storytelling in downloadable content and expansions: if only a third of your audience finishes the original game, perhaps story-based DLC would be better aimed at adding parallel content and “broadening” the game world rather than acting as a direct sequel. Then again, if that one-third of your audience is especially rabid, there may be reasons to simply shrug and support it. Again, there’s no easy answer, but looking at the data allows you to ask the right questions.

4. Know Typical Play Styles

Do you have a game with multiple Player Character “builds”? e.g., options for classes, weapon loadouts, special abilities, and so forth? Consider play styles when writing so that you don’t repeatedly reference the player’s pinpoint gun accuracy when half your players are putting all their skill points in martial arts or stealth.

Writing can also boost player interest in underutilized play styles. If no one’s even trying to play a healer, maybe an exciting combat medic NPC can help “sell” the audience on that particular concept. Work with your systems design team to see how you can better integrate story and gameplay by using metrics as a starting point.

5. Know Content and Aesthetic Choices

Distinct from player preference for game systems and playstyle are choices of content with minimal gameplay impact: e.g., choice of player sex / race / species, outfits or skins, companion or sidekick characters, and so forth. If very few players are creating, say, elves, is that a problem with the presentation of elves? Could it be changed with more interesting elf-exclusive content? Or is it not a problem at all? If players tend to dress their sidekick characters in sexually revealing clothes, is that something to push back against–emphatically rejecting the idea of those sidekicks as romantic or sexual objects in the storyline–or lean into, accepting that players want romance / sex and building those opportunities into the characters?

Straddling the line between content choices, play styles, and our next category is nonlinear content. Do players uniformly leave one mission or map for last? If so, is it because the story is naturally inclined that way (in which case, such a result may be desirable) or because Map X is described in the least interesting way upfront? What about repeatable, “grindable” quests–if players favor one variety of side quest grind over others, it’s probably for gameplay reasons… but adding story rewards to underused quest types could encourage players to diversify.

Finally, what about cutscenes? Knowing what cutscenes the player skips and in what context (if players tend to start fast-forwarding halfway through, you probably need shorter scenes, more compelling drama, or both) can be an enormous boon.

6. Know Branching Choices

Finally, for games with branching narrative (and / or branching dialogue), knowing what branches your audience tends to follow can make a world of difference. As always, determining why players choose the way they do can be difficult–did eighty percent of your audience choose to kill instead of save the comedy sidekick because he’s deeply annoying, or because you did too good a job justifying it#8211;but such data may be the single best window into the mind of the player. When you know the player’s choices, you know a lot about how the player is reacting emotionally to the story.

Look for data that suggest a lack of narrative engagement as well: Do players choose the “default” response at an usually high rate in certain scenes? They may not care about what’s going on, they may not understand their options, or they may be impatient to get back to gameplay.

Remember that narrative branches not taken aren’t necessarily a failure–the mere existence of an alternative path adds value to the game and weight to the primary path–but it should play a role in how you write, design, and budget different branches. If very few players take the “evil” option, then perhaps it’s best to make most evil options relatively cheap to implement, pulling out all stops only for key rewarding moments.

That’s six examples–just enough to get you started. Whether you get this sort of data from QA testers, focus groups, convention demos, beta testers, online players, or through some other method entirely, study it. Learn to use it. Trust your judgment to interpret it. And make your game better.