I've drawn cartoons from the moment I learned how to put pencil to paper, meaning I've been a writer practically since I could physically write. (How good a writer I am remains up for debate.) Growing up in an age before the Internet, before I'd even learned that fan fiction was a thing, much less that it had networks, my only point of reference for developing writing skills (beyond literacy classes) was printed books, and everything I made was done in isolation and largely for my own entertainment. I have always had a vivid imagination, and while I would borrow inspiration from the books I'd read as a child, I never actually thought to adapt the universes themselves—and in the exceedingly rare instances that I did, I kept the core canon at arm's length. In fact it wasn't until I discovered webcomics that I seemed to find a taste for fan fiction as it's commonly understood today, and even now I rarely tread outside parody mashups (à la DYOS) or tributes to authors with whom I have some sort of direct rapport.

I don't pretend to claim I have a 'purer' or somehow 'better' background in fiction as a result (and a lot of my early works would make an unflattering commentary unto such), but I do think I have a different approach to writing, and character development in particular, than what I see iterated and reiterated in most of the (fan)fiction writing guides I've come across online, here and elsewhere. This article is an attempt on my part to understand what, and why, those differences are, and provide a general reflection on my own writing process, which outside of dialogues with my long-suffering colleague DaemonDD I rarely do. This is not written as a guide itself, but if any of this can be of use to aspiring writers I certainly shan't complain.

What I've managed to glean from my journeys through the Internet is that fan fiction is often used as a tool by young (read: 18-and-under) aspiring writers to develop their narrative skills by practicing with pre-existing material—which requires, to varying degrees, adaptations of the canon cast. The term 'original character' (OC) is virtually synonymous with fanfic writing, for reasons that should be self-evident: published authors never use the term, because for them, every character is original; fanfic writers, by contrast, sometimes go out of their way to tout their characters, either because they consider these the hallmark of their adaptation, or the fanbase is so toxic and myopic that they feel they need to outline the divergences upfront, lest some inattentive bozo be caught by surprise midway through the story and grab the complimentary comment flamethrower.

Writers like OCs because they offer a way to personalize the story in what they (and no doubt their readers) hope is enduring. Everyone knows the canon cast (or at least one hopes they do), so new characters inherently offer a degree of surprise and a means of shuffling up the status quo. Whether they actually do this, or whether they degenerate into the dreaded Mary Sue cliché, is another question, and will be discussed further on. Fledgling writers can be especially protective of OCs because they represent not only their first tentative steps into written fiction, but also aspects of the writer's own personality—criticizing the character amounts to criticizing the author (and this sentiment isn't exclusive to fanfic).

But characters are only half the proverbial story, and OCs typically make or break themselves based on how well they integrate into the actual story. By far the most common complaint levelled against an OC is that it unrealistically (or at least unconvincingly) shoehorns itself into the setting for no clear reason—and in the case of the archetypal Mary Sue, re-wires everything to make the OC the star of the show, canon and continuity be damned. Once again, this is hardly a fanfic exclusive: recall (pre-beard) Wesley Crusher from Star Trek, Scrappy-Doo, Poof & Sparky from The Fairly OddParents...

For those unfamiliar, 'Mary Sue' refers to a character that's basically engineered (or just so sloppily-developed) such that it has free rein to do literally whatever it wants in a story with no negative consequence. The name stems from a character in a 1970s Star Trek fanfic parodying trends in author-avatar OCs, that has since been adopted by the fanfic community at large. A Mary Sue is literally too good to be true, to the point that contrary to the author's intentions, the audience hates it. Such a character can be recognized virtually by instinct, yet even knowing all the warning signs, veteran writers can still fall victim to its tropes.

Enter the OC creation guide.

I won't name any specific tutorials because I don't need to: the basic guidelines are so ubiquitous that running a search for 'OC creation guide' will yield no shortage of examples. Excepting tutorials that focus on the particulars of individual fandoms rather than a general primer, just about everything that labels itself 'OC development', 'fanfic development' or the such, begins with a single premise, stated or otherwise: How NOT to write a Mary Sue. These open with an overview of what makes a character bad, then list categorically all the necessary elements to 'a character', from physical appearance to family and background to personality and talents. If this fill-in-the-blanks approach to character design sounds like something out of a role-playing game, it's probably not coincidental: there is a large degree of overlap between writing fanfiction and engaging in roleplay sessions (closed games or open forums alike), and many OCs are intended to be (re)used in multiple settings—a character with an illustrated biography, multiple full-body profiles and precise details like height and weight is almost a guaranteed candidate for such.

For novice writers and people with roleplaying backgrounds, these guides do offer a useful primer; for veteran writers, especially those that write beyond episodic sitcoms and soap operas, they can actually become an impediment. The Mary Sue is despised because it is 'unbalanced': it only has positive characteristics, nothing negative. Thus, the popular solution of counterbalance: for every good quirk, also have a bad one; bring the idol back down to the mortal realm. Once again, this shares a lot in common with roleplaying games where the Mary Sue is replaced by 'powergaming'—maximizing statistical advantage in contempt of story or character development. The problem with this check-box approach is that it reduces characters to mere objects—see all the 'OC tests' that use quantified scores—tacitly implying developing a 'good' character is merely a matter of balancing the equation to a zero-sum game.

And this brings me to where I depart from conventional character creation guides.

I should state upfront: the only time I have ever used a guide is when the roleplaying game demands it. For me, crafting a character is instinctual, and has been since I first started playing make-believe. In fact my innate eccentricity may be part of the reason I find it so easy: not only was my roster of toys (stuffed or otherwise) imbued with a range of unique personalities, it was accompanied by comical voices to boot. I also had (and indeed still have) a vivid mind's eye, even if I struggle to transcribe that vision onto paper; thus when I started writing, I already had a firm grounding in diverse (if at the time simplistic) characterization.

One thing I never did (unless mandated by school assignments) was use an in-story avatar. I did have a fictionalized self that intersected with an emerging pretend-universe, but even this was treated largely as an external character, and typically manifested as a neutral spectator or faceless Mission Control rather than a core protagonist—it wasn't until DRAW Your Own Story that I engaged with the idea of a direct self-insert, and even then I'm stretching the adjective. What I also didn't do was write for my age; all throughout school (and into university), teachers and students alike remarked I acted, or at least tried to act, older than I was, and this carried directly into my writing: I wasn't even ten and my stories revolved around action-adventure and military thrillers with an all-adult cast, and as I've stated several times, much of what my friends see today is merely the superstructure to foundations laid decades previous. Maybe I was acting precocious, maybe I was simply being introverted, maybe I felt my own self boring in comparison, but the end result was I simply didn't like to write myself.

A side consequence to all this is that rather than build up one central protagonist with a supporting minor cast, I made groups of protagonists that, if not fulfilling complementary roles, still have contrasting personalities—indeed, looking over the full course of everything and everyone I've written, they can almost always be grouped into associative triads: Thorvald, Bjørn & Harald; Rudolph, Peter & Mikhail; Toyoda, Tojikawa & Narakane; Koldanev, Czciborycz & Gedeonin all under the aegis of Brusilov; and dozens more characters none of you have ever seen. Even where there is a clear hierarchy to the group, it's all bound together by a subtle interdependency.

But enough of metaphysical influences; how do I go about developing a character? Despite a popular boast these days that I can churn out fully-rounded OCs virtually on demand, the process is actually mostly incidental to the story. For someone (in)famous for convoluted spiderweb plot lines and fractal relationships, the characters themselves emerge with almost no planning at all: when Feldmarschall Schäffer was first introduced at the start of DYOS 10.5, literally all I knew about him was what everyone else saw in the page; his ties to Krieger only emerged later as I drew inspiration from CivGeneral' story. Similarly, Krieger himself originated as a one-off visual gag, and Toyoda began as a sidekick to a pair of characters that, ironically, have not (yet) followed him into DYOS from their source work.

This presents a counterpoint to the OC-tutorial paradigm that everything has to be established upfront before letting the character loose. Indeed, I know almost nothing about my own characters when they first launch, save for their physical appearance and mannerisms—and in purely textual works, sometimes not even the first. For me, creating a character is akin to planting a seed (or indeed, raising an actual child): it may start out very simple (and if drawn, literally cartoonish), but over time it grows and develops into an individual based on the pressures of its surrounding environment (read: the story).

So, what's in that seed? As much as I hate reductionism in any context, if I had to narrow down my methodology to a crux, it would be personality. Tautological though it sounds, a character is interesting through the depth of one's character; that is, how much one resembles someone you might meet in real life, and how much one distinguishes oneself from everyone else. Set aside facial features, equipment, and arcane abilities, and what makes someone an individual is largely in the mind. What is a character thinking? How does one's speech sound? What sort of vocabulary is used? Is one an extrovert, introvert, or an introvert pretending to be an extrovert, and why? Figuring out a character's psyche is probably the hardest part of the design process, but once you know how someone thinks, everything else is made considerably easier. For me, personality is the anchor point to defining a character: it offers a built-in guide to forecasting what course of action one is likely to pursue in any given scenario, what sort of people one will or won't get along with, and what sort of hobbies one cultivates on the side—all without having to explicitly list them on a bio page. I don't need to have a pre-written backstory or primers for future plots, because I can infer and extrapolate relevant details from this core template as the need arises. I've remarked before that several of my core characters were first laid down all the way back in early grade school (and the oldest possibly predating it); these days they are unquestionably more sophisticated, and their histories and experiences have greatly expanded from all the antics to which I've subjected them, but their core personalities are essentially identical to their very first drafts.

My second point of divergence concerns the relationship between character and story. Here I wish to introduce two (blanket and by no means mutually exclusive) categories, character-driven stories and story-driven characters, to illustrate several concepts I'll examine. Character-driven stories, as the name implies, are narratives sparked, built, and maintained through the cast's personality as much as its action: think Shakespearean tragedy, psychological self-reflection, and long-running TV serials. Story-driven characters, by contrast, are basically tools to communicate a large, over-arching, and typically metaphysical theme in a work: think of how the cast to Nineteen Eighty-Four, The Great Gatsby, and In the Skin of a Lion exist almost purely as metaphorical exhibitions. Again, these categories are in no way exclusive, and indeed it's often hard to tell whether a work leans definitively one way or the other: Gary Trudeau has stated that the Doonesbury characters embody specific social archetypes in service of the plot, yet they are far more personable than, say, everyone in The Da Vinci Code.

The 'roleplay-style' character creation guides mentioned earlier assume the author is working to a character-driven story, if they even presume a story at all; they may include brief reminders to make sure an OC 'fits with the setting', but otherwise they seem to treat character design as separate from the work itself. There is nothing inherently wrong in this approach, but it can create incongruities when crafting a story of the second type. To illustrate, one guide I've read says not to name a character based on physical appearance or mental outlook, because you can't predict these things at birth. Not an unreasonable view—besides, aren't we all sick of temperamental redheads named Fiona?—but depending on the depth of your story, you may want to invoke metaphorical imagery for thematic effect: consider in Lord of the Flies how Piggy, the fat, myopic asthmatic, is the voice of reasoned caution, and is persecuted because of it; or in King Solomon's Mines the proper Englishman Captain Good, whose unwavering commitment to European customs makes him almost comical in the African wilderness, yet whose moral code subtly challenges the racial prejudice rampant at the time. Likewise, an aspiring author may become so hung up with these guides' emphasis on the little details that they miss the woods for the trees; to cite one of my favourite novels, Darkness at Noon, very few of the characters are described in detail at all, and when they are the focus is on their ideological beliefs. But for the purpose of the story—interrogating the rift between the Old Bolsheviks and the emergent Stalinist order—that's all the reader needs. If one applied the character-driven paradigm to each and every figure, the fluff it would take to properly flesh them out would smother the book's main plotline.

Make no mistake: I'm not saying one method is superior to the other. As I've outlined, my own process leans toward character-driven stories, and when actors are totally enslaved to the plot you end up with the Star Wars Prequel Trilogy. The takeaway point is that characters and story commute on a two-way street. Outside of obvious pandering/powergaming, I see bad OCs as symptomatic of weak writing in general. All the same criticisms against character shortcomings will typically mirror themselves in the story structure: a character with skills and/or history overly-convenient to the plot will end up in situations of contrived coincidence; a gratuitously pathetic background will receive equally melodramatic attention; characters acting against their established personalities forewarn the story is either a trollfic or the author doesn't know/doesn't care about the canon. Good characters are good because they contribute something to the story, and OCs are best when they contribute something beyond what is already offered by the canon cast, without stepping on its toes to do so. The reason OCs (and especially self-inserts) take so much flak is because authors that don't understand this (or are simply writing for personal wish-fulfilment) will employ them as substitutes for an existing main character, rather than expand the story to give them a unique purpose. If you've ever found yourself crying "What's the point of this?", odds are it's because there isn't one.

To conclude, I'd like to offer one last self-spotlight that draws this all together. I have one character, used in several projects, that is effectively God in everything but title—not in the sense of "in a world of gods and demons", I mean literally omnipotent such that the universe could implode and it could watch from the sidelines. Sounds like a Mary Sue, right? But here's the thing: I've run it through many of these online tests, and every one gives a different answer. The only definitively critical responses were from the RPG-style stats balancers, while some even suggested the character is underplayed. How is this possible?

Partly it depends on the metrics of the particular quiz, but again, this all rolls back into how story and character relate to each other. The Mary Sue's fatal flaw is it abolishes all narrative tension: why would anyone bother reading a story when the story itself makes clear our immaculate hero is in no real danger, won't waste time building a relationship, will never make a grievous mis-step? Eon Films' James Bond can be argued to exhibit many of the staple problem tropes: he beds all the women, bests all his foes, and escapes from perilous situations with convenient trick gadgetry—yet even though we know he'll survive the climax since we know how many sequels there are, the films still engross us because the danger is communicated believably enough that we still feel uncertain in its outcome. You can fool all of the people some of the time, and that hinges on knowing when, and how, to momentarily suspend their disbelief. On paper my God-character is undeniably overpowered, but in practice almost none of its special abilities are ever showcased, and those that are help form the thousand-piece jigsaw that is my typical narrative master plan. Personality-wise there's very little to overtly suggest it's anything more than an eccentric savant; indeed, as with so many of my characters it was born of humble origins, and accidentally evolved into an all-powerful deity when I realized it was the perfect vessel to interrogate the great moral and existential quandaries. It's not there to show off; it does a job in the story. Whether a character is overpowered or underpowered depends on what's being factored and under what context; it's no coincidence that my example's worst rankings fell under tests that aggregated its qualities at face value, while the best rankings examined it in relation to the story as a whole.

If I can condense this all into a pithy quoteworthy, it is: Don't write a character like you'd solve a math problem. The whole is more than the sum of its parts, and this is especially true in biology. OC creation guides would have you believe it's impossible to write a 'flawless' character that audiences will still find engaging, and yet my example above is one of the most beloved even amongst my most incisive critics. There is no perfect formula to ensuring a character 'measures up' because each character, and each story, is different, and holds itself to different standards. Regardless of whether the characters drive the story or vice versa, it's the cohesion between the two that ultimately determines whether the project sinks or soars. I have submitted psyche as the cornerstone to my character creation process; climbing one rung higher, I submit that my writing overall balances on purpose: purpose of story, purpose of individual plots, purposes of characters in regard to said plots, each other, and their own life outlook. Character design guides will help you through the basics—who, what, where, when, and how. The author's eternal challenge is to figure out why.