Hi I’m back.
Supports
I think the writing question I’ve been asked the most often is “how do I write supports?” It’s a difficult topic to try to answer, because the process is exponentially harder the less adept you are at the writing process in general - by which I mean, if you’re already good at creating unique and distinct characters, and you’re already good at tying those characters in to the narrative in emotional or thematic ways, and you’re already good at tying both of the above into the setting and its history, then a lot of supports will genuinely just write themselves.
However, most people aren’t already good at all of these things, and so they can struggle to find interesting things to talk about in their supports. So, I’m going to try to help, as well as clearing up some misconceptions about “best practices” that I see people throw around a lot.
Part 1: Groundwork
Before writing any supports, you need to make sure your characters are actually worth hearing extra dialogue from. Worth noting that this doesn’t mean that the non-support script needs to make them worth it, just that your notes have to make them worth it - in other words, even if the supports are where the bulk of learning about a character comes from, if their personality is Knight and all of their supports are them Training, then there is no reason for me to read these supports, because I could have learned all that by looking at their portrait and their class.
If you haven’t done this groundwork yet, fleshing out your characters into distinct and meaningful characters, you aren’t going to benefit from writing a full suite of supports. Make your characters worth writing about before you start writing supports - it will make all subsequent steps easier. The scope of that task is outside the purview of this tutorial, so I’ll leave that to you until I deign to write a tutorial for writing minor characters.
The other thing you need to do is make sure these characters have interesting things to talk about. They all live in the world, and have opinions on what goes on in the world, so write those things down. If the world doesn’t have anything interesting for them to latch onto, then make interesting things for them.
An important note is “interesting.” Let me give an example real quick:
I have two characters here, John Interesting and Mary Boring.
John is a street urchin from a major city in an ally nation supporting the “good guy” nation. He’s grown up seeing the seedy underbelly of his nation, and has very strong opinions on his home country, his home city, on wealth and power. He has a pragmatic streak that breaks when he’s dealing with the downtrodden, a personal philosophy borne of an encounter he had in his youth - for bonus points, that encounter was with another character in the party. He has firmly defined reasons for joining the army on an ideological and emotional level, and his opinions often clash with other party members.
Mary is a street urchin from the capital city of the “bad guy” nation. She’s grown up seeing how evil the bad guy nation is, and really hates the bad guys for it. She’s unerringly nice due to growing up poor and virtuous, because she lived a hard life and wants everyone else to live a better life than she did. She joined the party because the protagonist saved her from bandits, and she largely agrees with the general consensus on why the “bad guy” nation is bad.
Both of these characters are tied into the setting and have fleshed out backstories that tie into their personal motivations, but Mary uses that backstory to justify uninteresting conclusions. Neither character is particularly unique by any means, but one of them has a ton of hooks you can use to bounce them off other characters in the script, and the other mostly conforms to the general shape of the party and exploring the topics in her backstory isn’t likely to yield information more valuable than what you could already glean from looking at her - being poor sucks, the bad guys are evil, she wants to be nice and do good.
When creating characters to be interesting, please keep in mind that simply drawing connections to the world around them is only one part of the process. Drawing interesting conclusions that lead to a variety of viewpoints, personalities, and “vibes” in the party is an important part of creating a cast, and is equally important for making supports interesting to read.
Part 2: Who should support who???
Oh my god this question.
So I’m going to go into more detail but I’m going to start with the conclusion: if you did part 1 right, literally anyone can support with anybody. That is it. Full stop.
Now of course, there are characters who might be easier or more obvious to write supports for - two characters with a pre-established connection, or who focus on a similar “topic” (like two deeply pious characters) - but there’s a lot of pitfalls in picking your support partners in this way.
For the former, some characters with a connection like this are main characters - think Eirika and Ephraim, or Byleth and one of the house heads - and having important relationship details develop between these characters in purely optional dialogue instead of in the main story can leave their connection in that main story feeling limp without it, or it can lead to a support chain that’s inconsequential by necessity - if any insight from that support chain were important enough to be interesting, it would have to be main plot in the first place.
For the latter, choosing support partners based on “topic” overlap can lead to supports that feel very samey or uninteresting. Two knights talk about knightliness, then one of those knights has two other supports about knightliness. Or, worse, the support about knightliness is boring, because the two topics overlap in the same way, leading to them just agreeing with each other. If you’re not learning new, valuable insights about a character by reading a support of theirs, why are you reading it? Why are you writing it?
Really, the only criteria for if you should write a support is if the two characters involved would have interesting things to say to each other, and if your characters are good enough, then they’ll also be interesting enough, and multi-faceted enough, that the answer to this question will always be “yes.” Of course it’d be interesting to hear this character talk - they are interesting, and anything that explores what makes them interesting will be interesting by extension. There is no secret trick to picking “interesting” support partners - if your character can’t have an interesting conversation with any random character you throw at them, they might just not be interesting enough in the first place to give you consistently interesting supports.
As a result, I view this kind of surgical support selection process to be kind of a crutch, though it’s worth noting a crutch isn’t a bad thing, necessarily. If you feel your characters can’t quite be fleshed out enough to be interesting in any configuration, then keeping them limited to the aspects of their character you know are interesting by shrinking the scope of their possible support partners is a fine way to do so without needing to go back to the drawing board. But, even within that framework, there’s still no secret science to who people should support with. As long as the characters can have an interesting conversation, they’re worthy support partners.
Part 3: C supports, and why 3H is badly written
A C support is where any given support chain starts. Importantly, it is not where the conversation starts: your C support should be a full, complete topic in and of itself, even if it leaves threads dangling for your B support to pick up on. I recently played through Engage, and me and my partner made a lot of jokes to the tune of “oh my god she’s CBAing him!!!” whenever a character would abruptly stop a topic and go “I have to go, we will continue this discussion later.” If your idea for a support chain could be summed up in a single C support, then you probably shouldn’t be writing this support chain.
I like to chart out a support by first writing a short paragraph per C, B, and A, summing up roughly what will happen between each level. It can help you spot where you’re stretching to find meaningful ideas or where you feel like there’s missing connective tissue, and help you flesh out the support chain and solve potential problems with it before you even begin writing it.
As a result of solving for that, your CBA support chain should be capable of supporting three distinct, if linked, conversations. That might not sound like a lot, but secretly, three conversations is a ton of time. It’s so much time that you could potentially make a single support chain be about multiple topics if you really wanted to, resolving one of them mid-B support only to have it dovetail into a new topic for your A support to conclude. As a result, don’t be afraid to pursue relatively complex topics: you will be able to cover that ground.
Next problem to look out for: the Three Houses problem. In Three Houses, there’s a point about a chapter or two in where aaaall your party members hit their C support thresholds with each other. So, you start reading a bunch of C supports. However, each C support has a few problems that, when compounded, create for a worst case scenario.
Problem 1: the Three Houses cast is largely very simple, and don’t have very many character traits to draw from. As a result, most of their supports are very similar.
Problem 2: Each C support has to assume it’s the first C support you’re ever reading, and so must establish each of the characters as though you’d never spoken to them before.
Problem 3: Three Houses characters have a boatload of supports, meaning that there are a lot of C supports. Bare minimum, within your starting class, every character has a support with every other character, which means there are dozens of supports to read.
The end result is that you end up reading so. many. C supports. And they aaaaall read the exact same. They touch on the same key points, they’re all largely inconsequential, and for every one that’s genuinely interesting or unique, there’s a half dozen that are all the exact same. And you read sooooooo many of them. So many.
When you are writing supports for a character, you should ensure that every support chain is about something different. You should never be touching on something another support chain touched on. The problem is twofold: One, you run into the 3H problem, and the 3H problem sucks. Two, if one support chain ends with there having been significant character development for one of the characters, but then their C support for another chain puts them right back where they started before, it can be immersion breaking.
As a result, when planning a new support chain, think to yourself: Is this a topic I’ve already covered? Is this meaningfully different from something the player may have already seen? Am I actually expanding on an aspect of this character, or am I just writing to fill space?
Ultimately when writing a C support, your goal should be to establish some kind of conflict. This doesn’t need to be a conflict between the two characters, just some kind of central thing that must be resolved by the end of the support chain. This conflict should be, in some way, of interest to both parties in the conversation, and should not overlap with either of their existing support chains. By the end of the support chain, there should be a sense of “what’s going to happen next?” Leave a sense of tension or drama, such that players are interested in seeing how this topic resolves. This holds true for more comedic support chains as well - even if that central conflict is funny, the reader should be invested in how the conflict’s comedy will develop.
Part 4: B supports, and why 3H is smartly written
So 3H has this feature I really like, which is that its supports don’t have a uniform length.
A thing you’ll run into sometimes when writing supports is that you do have an interesting topic to get into, but it just can’t sustain itself for a full CBA support chain. 3H, smartly, limits some chains to end at B, or even extends some supports to have 4 entries. This allows supports that really have no business being particularly long to get themselves out of the way quickly, and for more intricate supports to have a steadier and more natural progression from C to A.
We (probably) don’t have that luxury, so our solution is a little more cutting-room-floor: if you’re writing a support chain, and your B support feels like it’s just reiterating the C support again, or like you resolved everything in your B support and have nothing to write for your A support, you may just not have enough content for a CBA in this topic and should probably scrap this attempt.
(There’s a thing I haven’t even talked about yet: You can get a support chain wrong. It’s okay to delete work you’ve done; it will inform you better on how to not have to delete work the next time you do it, giving you an idea of what does and doesn’t work about this character pairing.)
Anyway: B supports are the trickiest, because it’s easy to bring up a problem, and it’s easy to resolve a problem, but it can be difficult to develop a problem, to explore it and give it depth, to work towards a conclusion without actually reaching it yet. This also makes B supports really interesting when you do them right - they’re where you might end up learning the most about a character, now that the stakes are set by the C support and you don’t need to spend any of the word count on “wrapping up” the support chain like the A support will need to.
That is, of course, your goal in a B support. You should be focusing on taking the ideas brought up by the first support conversation - any conflict, themes, character traits, etc that were used to establish the support chain - and delve into those ideas in detail. Bounce the characters off one another, see how their personal insights on these ideas causes friction between the two characters, or unexpected harmony, or might spark a revelation. Avoid restating the premise of the C support or simply bringing up the same problems in the same way again - bring something new and fresh to the table.
By the end of the support, the conflict should be reinforced - if the A support is going to resolve it, then typically at the end of the B support should be when that conflict is at a breaking point, either at its most severe or at the exact moment one of the characters acquires the tools to try and solve it. You should be setting the stage for the A support with this ending, in much the same way your C support set the stage for the entire support chain. Endings are important, even for something as short as a CBA support, so making sure you have all your dominoes set up before knocking them down will do wonders for your ability to stick the landing.
Part 5: A supports and how to make supports matter
A big problem I have with a lot of supports, especially fangame supports (although vanilla is often guilty of this too) is having supports that just don’t matter in the slightest. 600 to 1000 words, spent talking about a topic where I could have pieced the details together within 5 seconds of looking at a character; spent on hijinks that are at best charming and at worst actively annoying, breaking the pace of the story for something pointless; spent on characters dwelling on their least interesting character traits and refusing to meaningfully develop them.
If you are making the audience read a conversation, it’s because that conversation matters. Your supports aren’t a space to show the mundane nothings that characters get up to in their free time - in fact, you probably want to make it clear that they have conversations outside of the context of their CBA support chain - but rather a space to show critical moments in that character’s growth from the start of the game to the end of the game.
If the focus on a conversation is on a connection they’re making, make that connection deeply meaningful, something that genuinely changes them for the better. If the focus is on a topic they’re exploring, make that topic something important to the world, to the themes of the story, make the conclusions they reach feel meaningful. If the focus is on comedy, make sure the joke is, like, actually funny and not annoying! Not even vanilla can do that!!
When writing an A support, you have a tricky job: You need to have conclusive and permanent character growth for the characters involved, but it can’t be the kind of character growth that would impact the way they talk in their other supports or in the main plot. You also have to make sure that character growth is meaningful, interesting, and earned - that last part is very tricky and very execution prone, you’re on your own for that one.
Lastly, some A supports are romantic in nature. Please, don’t do what the modern games do: if you intend for an A support to be romantic, please make sure the entire support chain is also romantic, or at the very least in the case of multiple paired endings make sure they support a romantic reading. Don’t do what 3H does, where every A support is a marriage proposal and you can read them all in the same playthrough, but also don’t do what Fateswakening does, where after a fully platonic support chain there’s a sudden swerve into romance at the last minute with “oh the entire time I was secretly talking to you about all that stuff because I was in LOVE with youuuu” it’s just bad writing. Don’t do it!!!
Writing a romantic support is all the same steps as above, except they’re also expressing romantic interest. It’s even more important, in supports like this, that your characters have strong chemistry - please don’t make people read two cardboard cutouts sloppily flirting at one another with 0 things in common and 0 reason for them to like one another. Make their connection meaningful and real, just like you’d make any CBA meaningful and real, and you’ll be in good hands.
And that’s it! You’ve made a CBA support chain! Now do that… more times. How many times were you planning on doing this, again?
Uh oh.
Part final: The workload
This is a big thing and I have to talk about it before I finish up here. Supports take a long, long, long, long, long time to write. A long time. You are, very likely, one person, doing this work for free. You might think that because a CBA support chain is short, that means writing a lot of them would be easy.
It is not.
For Dark Deity 2, a game with only 40 “bonds” (they’re supports), each averaging around 1.1k words, the bond writing process took many months. The final bond script, at about 45k words, was nearly 2/3 as long as the main game script, and all for missable dialogue that most readers won’t even see half of. 40 bonds gave us enough room to have 4 bonds per character in a 20 character cast. That’s a hefty chunk of bonds per character, but a very small cast by FE standards. Most hacks will have closer to 30, or 40, or even 50 characters. At 4 supports per character, for that many characters, you’re looking at 100 supports, 60k-120k words in supports alone depending on how wordy you want to get with them. That would make your support script alone potentially larger than the entire script of a 30 chapter long published SRPG.
When planning out your supports, do not go “I want every character to support with every other character!” or “I want every character to have 7 supports bc that’s the max in FEBuilder!” or other massive scope creep things like that. If you do, you will spend genuinely over a year doing nothing but writing supports, and you will go insane. Please, please, please, please please, please please please!!! limit the scope of your support writing endeavour. Having a full support list isn’t worth shit if your hack dies and never releases: If you reach a more reasonable goal and feel you want to write a little more, then you can consider it, but start with a smaller goal, because even a “small” support list will take ages to write for very little payoff.
Anyway if anybody in writans asks one of the usual suspect questions about supports please just link them this. If they have more specific questions after that they can ask follow-ups.
idk if I’ll write another one of these anytime soon, I’d just been thinking about writing this for while so here it is. Hope this helps.