What to Fix First When Your Fiction Draft Feels Broken

You've got a draft. Maybe it's fifty pages. Maybe it's three paragraphs you've rewritten nine times. Either way, something feels off. The characters are flat. The plot limps. You're not sure what to fix first. That's normal. Fiction writing is a craft of constant trade-offs: plot versus character, showing versus telling, pacing versus depth. This article maps the terrain. We'll look at where fiction actually shows up in real work—not just novels, but screenplays, games, and serialized web fiction. Then we'll untangle the stuff that trips everyone up, point out patterns that hold up, and flag the traps that make teams revert to boring safe choices. By the end, you'll know what to tackle first when the whole thing feels broken.

Where Fiction Writing Actually Shows Up in Real Work

Novels and short stories

Most people picture a lonely novelist in a coffee shop. That image is true, but it hides the real story. Fiction writing is the engine behind half the media you consume daily. I once consulted for a mobile game studio that spent six months on mechanics nobody played. They had tight combat loops, gorgeous art, and a story so flat players quit by level three. The fix wasn't more features — it was a three-page backstory for the villain. Fiction writing gave them stakes. That sounds simple until you watch a team burn budget on systems while ignoring why anyone should care.

Screenwriting and TV pilots

Screenwriting is fiction writing under a different name, just with stricter formatting and a stopwatch. Every pilot episode is a short story compressed into thirty-eight pages. The catch is that most corporate pitches skip the emotional arc entirely. They lead with market data, demographic overlap, and platform stats. Nobody asks what the protagonist wants by page ten. I have seen a $2 million pilot fail because the writer forgot that tension needs a clock — a deadline, a ticking bomb, a reason to turn the page. Fiction writing gives you that clock. Without it, your pilot is a resume.
That hurts. But it's fixable.

Interactive fiction and games

Games are the wild west of fiction writing. They demand structure that breathes — choice trees that don't collapse, dialogue that works in fragments, and pacing that survives a player clicking everywhere except your intended path. The tricky bit is that most game writers come from linear backgrounds. They write a novel's worth of lore, then shove it into a codex nobody opens. Wrong order. What works in interactive fiction is constraint: three emotional beats per level, two lines of dialogue per NPC, one decision that matters by minute five. I rebuilt a branching narrative last year by cutting seventy percent of the exposition. The playtest feedback flipped from "confusing" to "I can't stop playing." Fiction writing isn't decoration here — it's the mechanic.

Fiction writing in games is not about the words. It's about what the player does next, and why they want to do it.

— Lead narrative designer, AAA studio (off record)

Serialized web fiction

Web serials are the closest thing to a weekly TV show in text form. Chapters drop on a schedule, readers comment in real time, and the writer adapts based on which plot threads spark arguments. This is fiction writing under sprint pressure — no rewrites, no second drafts, just forward momentum. Most teams that try this revert hard after three weeks. Why? They pre-wrote twenty chapters in isolation, then discovered readers hated the middle arc. The fix is brutal but simple: publish the first five chapters, listen, then write the next five with actual feedback. Fiction writing is not a monologue. It's a conversation you start and then let escape your control. Serialized fiction proves that — or it proves you can ignore your audience until they leave. Your call.

Foundations That Trip Up New Writers

Plot vs. story: the difference matters

Most new writers jam every event into a draft like they’re stuffing a suitcase. That’s plot—a list of things that happen. Story is different: it’s the emotional logic that makes those events matter. I’ve seen twenty-page drafts where a character gets fired, finds a key, and drives to the coast, and none of it connects. The reader feels nothing. Why? Because the writer confused chronology with meaning. The fix isn’t adding more action; it’s asking why this scene, for this person, right now. Strip out any sequence where the protagonist could be swapped for a cardboard cutout and the plot still moves. That hurts—cutting pages feels like failure—but what remains is what holds weight.

The catch is that plot is easier to outline. You can bullet-point a heist. You can't bullet-point grief. So new writers lean on the heist and hope the grief leaks through. It doesn’t. Story demands tension—a character wants something and can’t have it, or has it and realizes it’s poison. Without that, you have a timeline, not a narrative. I once coached a writer who had twelve chapters of a spy thriller but zero scenes where the spy doubted her mission. She was a machine, not a person. We found the story by asking: what does she lose if she succeeds? That single question rewrote the first three chapters.

The real meaning of 'show, don't tell'

This is the most misused rule in fiction. Beginners think it means never say a character is angry—always have them punch a wall. That leads to exhausting prose where every emotion is a slammed door or a clenched jaw. The real trick is knowing when to show and when to tell. Showing creates intimacy; telling creates speed. A draft that only shows is like a film with no cuts—every detail is equally important, so nothing is. A draft that only tells is a summary of someone’s life, not a story you live inside.

Here’s the difference in practice: “She was furious” tells us the fact in three words. “She set the coffee cup down. Hard. The ceramic chipped against the saucer, and she watched the crack spread without blinking” shows the feeling through action and detail. Both are valid. The problem is when new writers use telling for the big moments and showing for the small ones. Save your showing for emotional turning points—the betrayal, the confession, the decision that changes everything—and let your transitions coast on telling. Most teams revert to telling-only because showing everything takes ten times the word count. That’s fine. Pick three scenes per draft to slow down and inhabit. The rest can breathe.

‘You can’t show a character breathing every minute. You show the moment the breath stops.’

— overheard at a workshop, credited to a novelist who never wrote a manual

Point of view consistency

Nothing breaks a draft faster than head-hopping. I’ve read manuscripts where we’re inside the hero’s panic, then two paragraphs later we’re inside the villain’s calm, then back to the hero, all on the same page. It feels like the narrator has ADHD—and the reader can’t trust where to stand. The fix is brutal: pick one POV per scene and stay there until a clear break (chapter break, line break, or a hard scene shift). Even then, be stingy. Multiple POVs work in published novels because those writers have earned the reader’s trust. New writers haven’t. Not yet.

Honestly — most fiction posts skip this.

That said, consistency doesn’t mean robotic. You can slip into a character’s thoughts without italicizing every internal sentence. A phrase like “She watched him lie—the way he touched his collar, the same tell from their first argument” keeps you in her POV while offering judgment. The pitfall is explaining what other characters feel. If your hero can’t know that the villain is nervous, you can’t write “the villain’s hands trembled with hidden guilt.” You can only show what the hero observes: trembling hands. The guesswork belongs to the reader. I tell writers to color-code their drafts: highlight every sentence that reveals a thought or feeling not accessible to the POV character. You’ll be shocked how many pink sentences show up. Delete them—or switch POVs at a scene break. One or the other. Not both.

Patterns That Usually Work in Fiction

Tension escalation on every page

The trick that separates a draft from a real page-turner? Every scene should feel slightly worse for the protagonist than it did three pages ago. Not catastrophic—just a hair tighter. A character walks into a room looking for a letter; they find the envelope but it's already been opened. Small. Damaging. That tiny escalation is what keeps a reader leaning forward. I have seen entire chapters collapse because the author thought 'something happens' was enough. It isn't. Something has to worsen. The catch is that many new writers treat tension as a single explosion at the end of a chapter—they forget the slow, mean grind of smaller threats stacking up. Wrong order. You want the reader thinking, Oh no, not another thing, not Okay, what's the big set piece?

Scene sequencing for momentum

Most teams skip this: the order of scenes matters more than the content of any single scene. Put a quiet, introspective beat right after a chase and you get a reader catching their breath—but put it before the chase and you get a reader skimming. Simple. Yet I see drafts where flashbacks, dialogue-heavy setup, and action sequences are shuffled like a deck of cards. The cost is momentum. One concrete fix: end each scene with a question the next scene must answer—not a cliffhanger necessarily, but a hook that makes the transition feel inevitable. That sounds fine until you realise how many scenes currently end with a character going to sleep or looking out a window. Not yet. That kills pace dead. Readers don't need closure on every page; they need a reason to turn this one.

‘Tension is not a switch you flip. It's a dial you turn, quarter-inch by quarter-inch, until the reader's fingers hurt from gripping the book.’

— overheard at a workshop where the speaker had just watched a promising fantasy draft die in the middle of chapter twelve

Voice consistency across characters

What usually breaks first in a draft is not plot—it's voice. A grizzled detective suddenly using the same metaphor as a teenage hacker. A medieval farmer thinking in twenty-first-century idioms. Readers sense this instantly, even if they can't name it. The fix is brutal: read every line of dialogue aloud, then ask yourself if that character would ever say it that way. Honestly—half your dialogue will fail that test. I once had to rewrite an entire second act because the villain spoke with the same cadence as the hero's comic-relief friend. Embarrassing. But fixable. The trade-off here is that rigid voice rules can make characters feel like caricatures if you overcorrect. Leave room for overlap—two characters might both use short sentences under stress—but strip out the phrases that belong to you, not to them. That hurts. Do it anyway. Voice drift is a long-term cost that compounds with every page; catch it early or watch beta readers complain about 'flat characters' without ever telling you why.

Anti-Patterns and Why Teams Revert

Info-dumping in the first chapter

Nothing kills momentum faster than a wall of backstory masquerading as prose. The writer knows everything about the world—the treaty signed in 1847, the political factions, the magic system's fourteen rules. So they cram it into chapter one, convinced the reader needs context. They don't. Not yet. I have watched perfectly readable manuscripts sink under the weight of three paragraphs describing a fictional currency. The trade-off feels cruel: withhold information, risk confusion. Dump it all, guarantee boredom. Most writers revert here because clarity feels safer than mystery. It isn't. The reader will infer more from a single character eating stale bread than from four pages of economic history. A prologue isn't a license to lecture—it's a promise that the story matters more than the lore.

Passive protagonists who wait for plot

The protagonist stands in a room. Things happen to them. The villain monologues, the sidekick explains, the mysterious letter arrives. This is not a character—it's furniture. The catch is that passive protagonists feel realistic. Real people do wait for answers, hesitate, avoid conflict. But fiction isn't real life. Fiction demands agency. A protagonist who refuses to act forces the plot to drag them, and drags the reader along painfully. "But my character is an observer," writers argue. Fine—make them an observer who intervenes at the wrong moment. Make their passivity a deliberate choice that backfires. We fixed this once by giving the protagonist a single actionable goal in the first scene: escape a room, find a name, open a door. Suddenly every hesitation became meaningful instead of empty. The anti-pattern here is treating hesitation as a personality trait instead of a narrative obstacle.

'The reader doesn't care why your character is sad. They care what your character does about being sad.'

— overheard at a workshop, dubious source, but the line stuck

Overusing internal monologue

Internal monologue is a crutch. It whispers that showing is hard, so telling what the character thinks is easier. Wrong order. A character who narrates every emotion in real-time robs the reader of discovery. "I felt a chill of dread" lands softer than a description of sweating palms and a locked door that shouldn't be locked. The temptation is understandable—you know what the protagonist feels. Translating that into action takes effort. The pitfall deepens when teams revert: they cut exposition, panic about lost clarity, then flood the manuscript with thought-tags. I have seen drafts where three consecutive paragraphs begin with 'He wondered if…' and 'She realized that…' and 'He knew…'—each one a tiny dam against the story's flow. Break these by forcing the character to interact with an object, another person, or a physical barrier. Let the reader infer the emotion. Trust them. That trust is what separates competent fiction from prose that lectures the audience into submission.

Maintenance, Drift, and Long-Term Costs

Continuity errors across a series

The first time a beta reader emails you about a character's eye color changing between chapters, you laugh. The fifth time—when your protagonist mentions a dead uncle who was alive in book two—you stop laughing. That's the moment maintenance becomes real. I have seen writers lose three weeks re-editing a single manuscript because a timeline spreadsheet existed only in their head. The catch is that readers remember. They remember the scar on the left cheek. The dog that died in chapter twelve. And when you contradict yourself across a series, you burn trust faster than a bad plot twist.

Most teams skip this until it hurts. Then they patch. Wrong order.

The hidden cost is not the fix itself—it's the context-switching. You stop writing new pages to chase down a date discrepancy from four hundred pages ago. That kills momentum. Worse, it kills the part of your brain that holds the story's emotional arc. One continuity error can snowball into a full rewrite of act two if you let it sit too long. I keep a single text file now. Just one. Every character name, hair color, dead relative, and invented holiday goes in there. It's ugly. It works.

Honestly — most fiction posts skip this.

Character voice drift over time

You draft chapter one with a sharp, cynical narrator. By chapter fifteen, that same narrator sounds like a philosophy major on sedatives. This is drift. It happens slowly—three words changed here, a more formal sentence there—until the voice you loved becomes unrecognizable. The tricky bit is that drift feels natural in the moment. You're in flow. The prose is smooth. But compare page one to page three hundred and the seam blows out.

What usually breaks first is dialogue. A character who swore never stops swearing in chapter four becomes polite by chapter twenty-two. Or the opposite. We fixed this once by printing the first ten pages and taping them above the desk. Every new chapter got a five-minute read-aloud against that original voice. Painful. Effective.

Drift costs you more than consistency. It costs you authority. When a reader notices voice drift, they stop trusting the narration entirely. Not consciously, maybe. But something feels off. They put the book down. That's the long-term cost: a slow bleed of reader engagement that you can't trace to any single scene.

You can't edit what you can't see. And you can't see drift until you compare the whole thing side by side.

— advice from a series editor who once tracked 87 character tics across six books

Reader fatigue and pacing burnout

Long projects sag. It's not a flaw—it's physics. The middle of a novel or the third book of a series naturally wants to slow down. You know more than you did. You describe more. You explain. And suddenly a scene that should take three pages takes eight. The reader checks their phone. They don't come back.

That's maintenance you can't fix with a spreadsheet. Pacing drift is structural. You have to chop. Remove entire paragraphs. Kill the beautiful description of the sunset that has nothing to do with anything. That hurts. But the alternative is worse: a book that feels like homework, followed by a sequel nobody finishes.

Most writers revert to shorter scenes after a burnout—tighter dialogue, faster cuts—but then the book feels rushed. The trade-off is brutal. Too slow loses readers. Too fast loses depth. The fix I have used: run a word count of every chapter, graph it, and look for the chapters that are 40% longer than the average. Those are the fatigue generators. Cut them ruthlessly. Then add exactly one moment of stillness per act—a quiet beat that earns the pacing rest. Not more. Not less.

When Not to Use This Approach

Writing for interactive fiction or games

The techniques in this guide assume a reader who consumes your story linearly, from page one to the end. That breaks fast in interactive spaces. When a player can choose dialogue branch C before ever reading branch A, your carefully built causal chain—this event because that event—collapses. I have seen writers spend weeks polishing a single character arc, only to discover that players never encounter the crucial betrayal scene; they clicked the wrong door. Traditional fiction's backbone (setup, rising tension, payoff) becomes a liability when the audience controls pacing. The catch is not that structure is evil, but that you must design for states, not for sequences. A branching narrative needs beat-by-beat test scripts, not a manuscript critique. Wrong tool, wrong job.

'The linear novel is a railroad track. A branching game is a subway map—multiple valid routes, and some passengers skip entire stations.'

— game writer, on a project I consulted for in 2023

Nonlinear or experimental narratives

Some stories are meant to feel broken. If you're writing a fragmented memory-recovery piece, or a novel told in reverse chronology, then the fixes I outlined above—tight causal links, clear scene purpose, rising tension—will sand off your teeth. That sounds fine until beta readers report they're bored. They're bored because you over-fixed. Disorder is the point. The drift is the music. I once watched a poet rewrite a prose-poem hybrid to fit "three-act structure" and the result read like a manual. Sad thing is, she knew it. She just trusted the template more than her instinct. Nonlinear work demands a different diagnostic: Does this confusion serve the theme? Yes? Leave it. No? Kill it. Not every draft needs a spine. Some need a spiral.

What usually breaks first in these pieces is consistency of voice, not plot. Characters diverge from their interior logic because the writer loses track of when the scene happens in the character's internal timeline. That's a problem you solve with color-coded index cards and a wall, not with the advice in section two of this article.

Strict word-count constraints

Microfiction. Twitter threads. Six-word stories. Flash pieces at exactly 100 words. Our advice to "show, don't tell" and "build scene through sensory detail" becomes a luxury you can't afford. When you have ninety-nine words to make a reader cry, you cheat. You summarize emotion in a single sharp phrase. You skip setup. You tell like a telegram. That's not a failure of craft—it's a different craft entirely. The maintenance costs I described in the previous section (drift, structural rot) don't apply here because the form is too short to decay. One bad sentence kills the whole thing, sure, but that's a line-editing problem, not a structural one. If your draft is under 500 words, stop reading this guide. Read Lorrie Moore instead. Or read the back of a matchbox. Learn how much weight one syllable can carry.

Field note: fiction plans crack at handoff.

The irony? Most strict-count constraints force better instincts. You learn to kill darlings before you learn to build them. Not bad. Just different.

Open Questions / FAQ

How much outlining is too much?

The real question isn't word count on your outline—it's whether your outline is serving you or shackling you. I have seen writers spend three weeks perfecting a chapter-by-chapter beat sheet, only to abandon the whole project on page forty because the story felt lifeless. That's too much. On the flip side, writers who plunge in with only a vague notion of an ending often stall hard around the sixty-thousand-word mark. The sweet spot? Outline until you can see the major turning points—the inciting incident, the midpoint crisis, the dark moment—but leave the connective tissue loose. You're not building a blueprint for a bridge. You're mapping a hike. Wrong turn? Cut through the brush.

Some writers swear by the one-page pass: a single sheet listing act breaks and emotional arcs. That's enough for a solid first draft. If your outline starts resembling a tax form—color-coded, cross-referenced, timestamped to the minute—you have likely crossed into procrastination territory. The work isn't in the plan. The work is the plan failing and you fixing it.

An outline is a hypothesis, not a contract. The draft will break it. Let it.

— overheard at a Pacific Northwest writers' retreat, name withheld

Should you write in order?

Conventional wisdom says yes. Momentum says no. I have coached two novelists who wrote their endings first—both said it unblocked everything. Knowing where the characters land let them write toward that destination with actual confidence. That said, writing out of order creates a specific trap: you end up with a stack of brilliant, isolated scenes that refuse to stitch together. The seams show. The timeline blurs. You lose a day—sometimes a week—just figuring out what order those fragments actually belong in.

Here is what usually works: write the scenes that burn first. If a confrontation in chapter twelve keeps you awake at night, chase it. But don't skip the transitional scenes entirely—those are where most character work actually happens. The fix is simple: after each burst of out-of-order writing, force yourself to write one bridge paragraph that connects your last finished scene to the next one you plan to tackle. Not a full scene. A paragraph. That single habit saves months of later restructuring.

How do you know when a scene is done?

Honestly—when it stops lying to you. A scene is not done when it reads beautifully. It's done when the emotional payload is clear and the reader can't be confused about what changed. The catch is that most writers over-polish. We fix line-level prose to avoid facing a deeper structural problem: the scene doesn't matter. The scene is fine. The scene does nothing.

Try this test: summarize the scene in one sentence. If that sentence includes the phrase "they learn that" or "they realize," you're describing knowledge transfer, not drama. A working scene should change the character's situation or deepen a conflict—preferably both. Otherwise, kill it. Recycle the good lines. Move on. Next experiment: take your longest scene and cut it by forty percent. If it still makes sense, it was never done before.

Summary + Next Experiments

Write with your hands, not your brain

Pick any 300-word scene from your draft—preferably one that feels flabby. Rewrite it using only action and dialogue. No interior thoughts, no narrator commentary, no stage directions about the weather. Just what a camera and microphone would capture. The constraint exposes every lazy summary you hid behind. I tried this with a chapter that had seven paragraphs of throat-clearing; the rewrite ran 110 words and hit harder. Your reader’s imagination fills the gaps you refuse to—if you let it. That’s the trade-off: you lose control over the character’s inner life in that pass, but you gain momentum.

Revise one page for interiority

Now do the opposite. Take a page where the protagonist does something—opens a door, pours coffee, receives bad news—and strip out every external detail except one sensory anchor. Replace the rest with what they feel, remember, or dread. Not their thoughts about the action; their body’s reaction, the flash of a past hurt, the half-sentence they won’t finish. Most drafts fail because they land in the middle: too much action to feel intimate, too much interiority to move. This exercise pushes you to one extreme so you can find the balance later. The catch is that your first revision will probably overshoot—dialogue scenes read thin, interior scenes read maudlin. Good. Now you know where the seam is.

Swap POV for one chapter

Rewrite a short scene from another character’s perspective—preferably the antagonist or a minor witness. Not to publish, just to see. What breaks first is usually your assumption that the original POV character must be the one telling this part. I once rewrote a confrontation chapter from the viewpoint of a bartender wiping glasses in the corner. Suddenly the whole power dynamic inverted: the hero looked desperate, the villain looked bored, and the real tension sat in what the bartender didn’t report to the police. That single experiment killed three rewrite passes I’d been stuck on for weeks. The pitfall: you might prefer the new voice and want to switch entire drafts. Don’t. Use the exercise as a diagnostic, not a transplant.

‘I write to find out what I think. The experiments are where I’m wrong fastest.’

— overheard at a workshop, 2019, name I forgot but the debt remains