The direct answer: where your chapter ends should land
For a 24-chapter novel on a steady rhythm, plan about eight or nine hard cliffhanger ends at the act turns, the midpoint, the late dark moment, and the run into the climax, about eight soft hooks elsewhere, and the remainder as rest ends, resting the chapter before the climax and never stacking three hard ends together. That single sentence is the whole method in miniature. Everything below explains how to apply it to your chapter count, your genre, and your tolerance for intensity.
A hard cliffhanger end interrupts something unresolved and urgent: danger closing in, a revelation that reframes the scene, a decision with a visible cost, a disappearance, a countdown starting, a betrayal surfacing. A soft hook end raises curiosity the reader can set down: a question asked but not pressed, an image that lingers, a promise of what comes next, a reaction shot that colors the chapter. A rest end closes the chapter cleanly and lets tension settle: a small resolution, a warm beat, a reflection, a breather. Most struggling manuscripts do not suffer from too few cliffhangers. They suffer from the wrong ends in the wrong positions, with hard ends scattered randomly and rest ends missing where the story needs to inhale.
This guide treats placement as rhythm. You decide the density first, then the tool marks which chapters carry hard ends, which carry soft hooks, and which rest. The positions follow structural peaks that hold across genres: the quarter turn, the midpoint, the three-quarter dark moment, and the launch of the climax run. Those peaks get the hard ends. The chapters between them alternate soft hooks and rests so the hard ends keep their force.
The scores on this page are deterministic heuristic checks against stated craft conventions. They measure your plan against widely taught structural positions. They do not read your prose, they do not predict sales, and they do not judge quality. Use them as a planning scaffold, then revise with your judgment.
How the placement math works
The tool takes three inputs: your chapter count, your genre, and your density setting. From those it computes a target number of hard ends, a target number of soft hooks, and leaves the rest as rest ends. Then it scores every chapter position, assigns the hard ends to the highest-scoring positions, enforces two anti-fatigue rules, and prints the full chapter-by-chapter map with a hook shape and a reason for each row.
Density sets the share. Sparse aims roughly one hard end in five chapters. Steady aims roughly one hard end in three chapters. Dense aims just under one hard end in two chapters. Genre adjusts the count by one: thriller and mystery add a single hard end because those readers arrive expecting pressure, while literary fiction removes one because that tradition favors accumulation over interruption. Romance and fantasy hold the base count. Positions do not move with genre. Only the count shifts.
Position scoring rewards structural peaks. Chapters within one slot of the quarter mark, the midpoint, and the three-quarter mark earn bonuses, with the midpoint and late mark weighted above the early turn. The chapter two slots before the finale earns the largest single bonus because it launches the climax run. Chapter one earns a small bonus because an opening chapter should generate forward motion. A light stagger term spreads ends apart so they do not clump: chapters divisible by four and chapters sitting two past a multiple of five each gain a point. The final two chapters are forced out of contention with a heavy penalty, and that penalty is deliberate, as explained below.
Once every chapter holds a score, the tool takes the top scorers as hard ends up to the target count. Then the anti-fatigue pass runs. Any chapter that would make three hard ends in a row is demoted to a soft hook. If demotions drop the hard count below target, the tool promotes the best-scoring remaining chapters that do not create a new triple. The final two chapters always rest. The remaining non-hard chapters fill soft hooks up to the soft target in score order, and everything left over rests.
Worked numbers: 24 chapters, steady, thriller
Walk through the default case so the arithmetic is concrete. Twenty-four chapters on steady density gives a base hard share of thirty-two percent, which is eight chapters after rounding. Thriller adds one, so the hard target is nine. The soft target is thirty-five percent rounded, which is eight. Nine hard plus eight soft leaves seven rest ends out of twenty-four.
Peak centers land at chapter six for the quarter turn, chapter twelve for the midpoint, and chapter eighteen for the three-quarter mark, each with a one-chapter halo on either side. Chapter twenty-two collects the climax-launch bonus. Chapter one collects the opening bonus. The stagger points lift chapters such as four, eight, twelve, sixteen, seventeen, and twenty-two by a point each.
The top nine scorers cluster around those peaks. Because peaks are three chapters wide and the midpoint halo covers eleven through thirteen, the raw top nine can contain a triple such as eleven, twelve, thirteen. The anti-fatigue rule demotes thirteen to soft and promotes the next best eligible chapter elsewhere, which is typically five or six. A second collision often appears at seventeen, eighteen, nineteen and resolves the same way, demoting nineteen and promoting chapter one or a nearby scorer. The finished map holds nine hard ends with no triple anywhere, eight soft hooks filling the gaps, and seven rests including the forced rest at twenty-three and the resolution at twenty-four.
The hard share lands at thirty-eight percent, which sits inside the healthy band. Nothing gets flagged. If you switch the same book to dense, the hard target rises to twelve, the share crosses fifty percent, and the verdict warns about overuse while still printing the corrected map. That warning is the tool telling you the truth: relentless pace costs each individual end its power.
Reading your placement map
The table is the deliverable. Read it left to right. The chapter column is your position. The end-type column says hard, soft, or rest. The hook-shape column suggests the flavor of the ending beat: revelation, danger, decision, disappearance, countdown, or betrayal for hard ends; question, image, promise, reaction, foreshadow, or shift for soft hooks; breather, resolution, reflection, or warmth for rests. The reason column ties the position to structure: act-turn zone, midpoint pressure, dark-moment zone, climax launch, deliberate rest, or spacing slot.
Treat shapes as starting prompts, not assignments. If the map says chapter twelve ends hard with a revelation shape and your midpoint is built around a decision instead, keep the hard placement and change the shape. Placement is the load-bearing choice. Shape is decoration you can swap freely during drafting.
The reason column deserves attention during revision. When a chapter marked as a spacing slot keeps ballooning into a massive twist, that usually means an unplanned subplot is hijacking the rhythm. Either promote the chapter officially by demoting a neighboring hard end, or cut the twist back to fit the slot. The map works best when you respect its trades: every promotion somewhere requires a demotion somewhere nearby.
What each end type asks of the scene
A hard end needs three ingredients in the final beat: something unresolved, something costly, and something immediate. Unresolved means the reader cannot answer the open question from information already on the page. Costly means the outcome changes what a character can do or have. Immediate means the consequences start now rather than someday. A chapter that ends with a phone ringing has none of the three until the reader knows who calls, what it costs to answer, and why tonight matters. Add those and the ring becomes a countdown shape. Without them it is noise posing as tension.
A soft end needs exactly one ingredient: a specific forward pull that does not demand continued reading tonight. A sharp question works. A vivid image with an implication works. A promise stated plainly works. A reaction that reframes the chapter works. What fails is vagueness: a character staring out a window feeling that everything was about to change promises nothing checkable, so the reader closes the book without cost. Name the thing that pulls forward and the soft end does its job.
A rest end needs closure plus one small gift. Close the chapter question that opened it, then leave a warm detail, a settled decision, or a quiet image. Rest ends fail when they read as filler: a chapter that neither advances pressure nor delivers satisfaction. If your rest chapter bores you, the problem is rarely the rest label. The chapter likely lacks its own internal shape. Give it a small question, answer it fully, and let the reader sleep well.
The two anti-fatigue rules
Never stack three hard ends in a row. The mechanism is habituation. One hard end spikes attention. Two in a row sustain it. Three in a row teach the reader that every chapter screams, and the nervous system discounts the signal. Page-turning becomes page-skimming. The tool enforces this mechanically by demoting the third stacked hard to soft. During your own revision, apply the same test: find any run of three chapters all ending in danger or revelation, and soften the middle one into a question or reaction. The surrounding hards will feel stronger immediately.
Always rest the chapter before the climax run. The penultimate chapter is a deliberate exhale. It lets the reader absorb losses, register what is at stake, and arrive at the finale with full attention rather than dulled nerves. Writers often fear this rest because the pace feels slow exactly when excitement should peak. That fear mistakes noise for momentum. A quiet chapter before the storm makes the storm legible. The tool forces chapters twenty-three and twenty-four of a twenty-four-chapter book into rest and resolution regardless of scores. Shorter and longer books get the same treatment at their own final two positions.
There is a third soft rule worth stating: keep at least one rest or soft chapter between hard ends outside peak zones. Peaks earn the right to pair two hards because the structure supports escalation there. The long middle stretches do not. Space your non-peak hards with breathers and each one lands harder.
Illustrative hypothetical example: the thriller that cried wolf
Consider an illustrative hypothetical manuscript: a first-time thriller writer with thirty chapters who ends twenty of them on danger or revelation. Every chapter closes with a gunshot, a secret exposed, or a chase beginning. By chapter ten the critique group reports skimming. By chapter fifteen the writer is adding bigger twists to compensate, which accelerates the fatigue. This is the classic overuse spiral, and it is exactly what the dense verdict warns against.
Run this hypothetical book through the sparse setting instead. Thirty chapters at twenty percent gives six hard ends. Add one for thriller and the target is seven. The peaks sit near chapters eight, fifteen, twenty-three, and twenty-eight, with chapter one earning the opening slot and two spacing slots filling from stagger scores. Twenty-three of thirty chapters now end soft or rest. The writer must choose which seven moments truly deserve interruption, and that choosing is the craft work. The remaining twenty-three ends get rewritten as questions, images, promises, and small resolutions. The critique group stops skimming because the signal means something again.
Notice what changed and what did not. The plot events stayed identical. Only the endings moved. Several twists now land mid-chapter where they detonate inside scenes instead of decorating exits. Two former cliffhangers merge into a single stronger midpoint. The penultimate chapter, formerly a chase, becomes a quiet reckoning that the writer at first resists and then admits is the best chapter in the book. Rhythm did what escalation could not.
Illustrative hypothetical example: the literary novel that would not pull
Now consider a second illustrative hypothetical: a literary novel of sixteen chapters where every chapter ends in reflection. Each closing is beautiful and each closing releases all pressure. Readers admire individual pages and abandon the book at chapter five. Nothing is wrong with the sentences. The rhythm never asks for the next chapter.
Run this book through the steady setting. Sixteen chapters at thirty-two percent gives five hard ends. Literary subtracts one, leaving four. The peaks sit near chapters four, eight, twelve, and fourteen, with the final two chapters forced to rest and resolve. Four interruptions across sixteen chapters is gentle, and that gentleness suits the tradition. But those four hards change everything: a buried letter surfaces at the quarter turn, a relationship decision breaks at the midpoint, a long-held version of events collapses in the dark-moment zone, and the climax run launches at fourteen.
The writer in this hypothetical keeps all sixteen reflective closings in some form. The four marked chapters simply add one unresolved beat after the reflection: the thought completes, and then the letter slides under the door. Rest plus interruption in the same beat honors both the aesthetic and the reader need for pull. Early readers now finish. Nobody accuses the book of selling out because four well-placed hards across sixteen chapters still reads as restrained.
Edge cases and failure modes
Very short books need care. At four to seven chapters, every end carries enormous weight, and the tool returns an informational verdict rather than a normal rhythm verdict. With so few slots, the distinction between hard and soft matters less than making each chapter earn its place. A six-chapter novella with three hard ends is already at fifty percent density, which would alarm in a long novel but is unavoidable when the whole book is essentially one extended climax run. Read the map, keep the penultimate rest, and do not chase percentages below eight chapters.
Very long books face the opposite hazard. At forty-plus chapters, even sparse density produces eight or more hard ends, and the middle can sag between widely spaced peaks. The fix is to treat the book as linked movements: confirm that each movement of ten to twelve chapters contains its own quarter, midpoint, and turn pressures, even if the global map only marks the global peaks. The tool scores global position, so apply its logic recursively to each movement by eye. Your part breaks and section gaps are natural rest points that do not appear in the chapter map.
Series finales and stand-alones share the same final-two rule but for different reasons. A stand-alone must resolve, so the last chapter cannot cliffhang by definition. A series finale that continues into another volume tempts writers to end the whole book on a hard hook. Resist unless the series contract with readers explicitly promises it. The map forces resolution at the finale because a book that never pays off trains readers to wait for the next volume before feeling anything, which depresses both satisfaction and word of mouth.
Prologues and epilogues distort counts. Decide before running the tool whether yours are true chapters. A three-page prologue with its own arc behaves like a chapter and should be counted. A half-page epilogue that is pure denouement behaves like an extension of the finale and should not be counted separately. Either choice is defensible, but counting differently between draft and revision silently shifts every peak by one, so record your decision and hold it steady.
Multiple viewpoint novels need per-thread rhythm, not just global rhythm. If viewpoint characters alternate, check that each thread has its own spaced hard ends rather than letting one viewpoint carry all the interruptions while another carries all the rests. The global map cannot see viewpoint, so export its hard slots and assign viewpoints to them deliberately: never give the same viewpoint three hards while another viewpoint never interrupts.
The most common failure mode is promoting without demoting. A writer falls in love with a new twist, upgrades a soft chapter to hard, and keeps everything else. Three drafts later the book has drifted from steady to dense without anyone deciding it. Make promotion a swap: every upgrade pairs with a downgrade within two chapters. The tool models this discipline automatically through its demotion pass, and your revision process should mirror it.
The second failure mode is shape repetition. Six hard ends all shaped as danger reads as monotony even when positions are perfect. Cycle through revelation, decision, disappearance, countdown, and betrayal deliberately. If your story cannot supply a disappearance or betrayal at a marked slot, that is information: the subplot inventory may be thin there, and the fix is story development, not a louder danger beat.
The third failure mode is the fake hard end: an interruption with no cost. A chapter that ends because the viewpoint character faints, the phone dies, or the scene arbitrarily cuts is not a cliffhanger but a withheld fact. Readers distinguish interruption from intrigue within pages. Every hard end on your map should survive the question of what the character stands to lose before the next chapter opens. If nothing is lost, soften it honestly and save the hard slot for a beat with teeth.
Genre-specific guidance
Thriller readers arrive with a pace contract. They expect regular interruption and forgive intensity that would exhaust other audiences. Run steady as your baseline and dense only for short books or final acts you will later dilute. Place danger and countdown shapes at the peaks and reserve revelation for the midpoint, where reframing pays best. Keep the penultimate rest even here; thriller writers resist it most and benefit from it most, because a single quiet chapter makes the finale feel faster by contrast.
Mystery needs a different discipline: the hard ends should withhold the detective solution while advancing the complication. Revelation shapes work best when they reveal a fact that deepens the puzzle rather than one that collapses it. Ending on the unmasking at chapter twelve of twenty-four strands the second half. Ending on the discovery that the timeline is impossible pulls the reader forward without spending the solution. Give mystery one extra hard end per the tool adjustment, but spend it on complication, not disclosure.
Romance turns on emotional decisions and their timing. Hard ends shaped as decisions outperform danger shapes in this genre: a choice declared, a boundary set, a confession interrupted before it lands. The midpoint hard end often belongs at the false resolution or the deepening commitment, and the dark-moment hard end belongs at the rupture. Because romance readers track the relationship arc closely, avoid ending two consecutive chapters on external danger; it reads as stalling the emotional plot. Let at least one peak hard end be purely relational.
Fantasy carries worldbuilding weight that slows endings. The temptation is to close chapters on lore: a history revealed, a map unfolded, a prophecy quoted. Lore closings are soft ends at best, because information without cost does not pull. Spend your hard slots on what the lore costs a viewpoint character now: the prophecy names someone, the map route crosses enemy ground, the history proves an ally lied. Fantasy also runs long, so apply the per-movement check from the edge cases and confirm each ten-chapter span has its own pressure.
Literary fiction earns pull through accumulation, voice, and moral pressure rather than event. Four hard ends in sixteen chapters is plenty. Favor disappearance and betrayal shapes scaled to the domestic or interior register: a person absent from a room where they should be, a confidence broken quietly, a self-accounting that can no longer hold. The rest ends are load-bearing in this tradition, not filler; write them with full attention, because the contrast between beautiful rest chapters and rare interruptions is the entire engine.
Short books, long books, and odd counts
Books under twelve chapters compress the peaks until halos overlap. A ten-chapter book has its quarter at three, midpoint at five, and three-quarter at eight, so halos at four through six and seven through nine nearly touch. Expect the map to mark consecutive hards near the middle and then soften one by rule. Read that softening as advice about your specific middle: the two surviving hards are your load-bearing interruptions, so make them your largest structural events rather than your loudest scenes.
Books over thirty chapters stretch halos apart and create long quiet corridors. Corridors are fine if each contains at least one soft hook with a specific pull. Audit every corridor chapter by asking what precise question would make a reader open the next chapter. If the answer is general atmosphere, rewrite the soft end to name a person, an object, or a deadline. Specificity is the difference between a corridor and a sag.
Odd chapter counts shift peaks by rounding, which some writers find suspicious. It is only arithmetic: the midpoint of twenty-three chapters rounds to twelve, and halos cover eleven through thirteen. Do not hand-tune peak centers by a chapter to chase symmetry. The stagger term already spreads ends, and one-chapter shifts in peak position have no detectable effect on readers. Save tuning energy for shapes and costs.
How this pairs with the chapter-hook-planner
This guide answers which chapters end hard, soft, or rest. The chapter-hook-planner answers what kind of hook each marked chapter uses. Run placement first and hook assignment second, because the number and position of hard slots constrains which hook types fit. A book with nine hard slots needs nine distinct hook assignments with varied shapes; a book with four needs only four, and each can be developed at greater length.
The practical handoff works like this. Export the hard chapters from your map. For each, assign one hook shape from the planner, checking that no shape repeats within any run of three hard ends. Then assign soft shapes to the soft chapters, favoring questions early in the book and promises late, since early readers need orientation pulls and late readers need payoff pulls. Finally, review rest chapters and confirm each delivers its small satisfaction; a rest chapter that neither satisfies nor pulls is the only truly wasted slot in the system.
When the two tools seem to disagree, trust placement over shape. If the planner suggests a brilliant revelation hook for a chapter the map marks rest, move the revelation to the nearest hard slot rather than upgrading the rest chapter. Great material in the wrong rhythmic position underperforms good material in the right one. Rhythm is the container; hooks are the contents.
Drafting the actual chapter endings
Scores and maps plan the work, but sentences do it. Once your map marks a chapter hard, draft its final beat as a standalone unit of three to five sentences: re-establish the immediate situation, introduce the interrupting fact or turn, and cut before anyone processes it. Then stop. The most frequent drafting error is explaining the cliffhanger in the same beat, which converts interruption into summary. Let the next chapter do the processing.
For soft ends, draft two sentences: the chapter close and the forward pull. The close should satisfy the chapter question. The pull should name one specific thing. Read the pair aloud and confirm the second sentence could not be deleted without losing a reader. If it can be deleted painlessly, it is decoration; replace it with a sharper pull.
For rest ends, draft the close and one sensory or emotional gift, then resist adding a teaser. Rest chapters fail when writers apologize for them with tacked-on hooks. Trust the map: the surrounding hard ends will do the pulling, and this chapter job is satisfaction.
Writers using Co-Writer can draft these closing beats where the manuscript already lives. Open the chapter in the write view, compose the final three to five sentences for a hard end or the closing pair for a soft end, and compare the result against the map position for that chapter. Keeping the planning map and the drafting surface side by side shortens the loop between deciding an end type and testing whether the sentences deliver it.
Questions writers ask at this stage
How do I know my hard end has real cost? State the loss in one clause: custody, money, safety, trust, time, or standing. If you cannot name the loss, the end is loud but empty. Either add the loss to the scene or demote the chapter to soft and relocate the hard slot to a chapter where loss is present.
Should every viewpoint character get a hard end? In books with two or three viewpoints, yes, distribute at least one hard end per viewpoint, and give the primary viewpoint the peaks. In books with five or more viewpoints, no; minor viewpoints can live entirely on soft and rest ends, which actually strengthens their charm as relief threads. Check the distribution after assigning shapes, not before.
Can a chapter contain a twist but end soft? Absolutely, and strong books do this constantly. A mid-chapter revelation detonates inside the scene, the characters deal with it for pages, and the chapter then closes on a question about consequences. This spends the twist where it has room to breathe and still pulls forward. Ending everything on the twist itself is the beginner pattern; burying twists mid-chapter and ending on their implications is the advanced one.
What if my chapters are very short? Short chapters change the feel of every density. Fifty two-page chapters on steady produce far more total interruptions than twenty-four ten-page chapters at the same share, but each interruption costs less because little accumulates between them. For very short chapters, drop one density level from your instinct: choose sparse when you feel steady, steady when you feel dense. The map percentages assume average-length chapters.
What if my chapters are very long? Long chapters invert the advice. Each end carries more weight because more story accumulates behind it, so sparse density in long chapters can feel eventful. Keep your instinctive density but audit every hard end for cost with extra skepticism; a weak hard end after thirty pages of buildup disappoints more than the same weak end after eight pages.
How do I revise an existing draft against the map? Mark your current endings hard, soft, or rest honestly, then overlay the tool map and list mismatches. Reclassify mismatches in passes: first convert fake hards to soft, then upgrade true softs sitting on peaks to hard, then install the penultimate rest, then break triples. Four passes, one rule each, beats a single agonized read-through.
When should I ignore the map? Ignore positions when a structural event moves: if your midpoint lands at chapter ten of twenty-four instead of twelve, move the midpoint hard end to ten and let the map halos go. The tool assumes conventionally centered structure. Your actual structure outranks its assumption everywhere. Keep the counts, the triples rule, and the penultimate rest; relocate the peaks to match your book.
A revision workflow that holds up
First, run the tool with your finished chapter count and record the hard chapters. Second, label your draft endings without editing and compute your actual hard share. Most writers discover they run ten to twenty points hotter than intended. Third, fix triples and install the penultimate rest before touching anything else; these two changes alone usually restore the sense of pace. Fourth, align peaks: ensure your largest events sit on hard-marked chapters or move the marks to your events. Fifth, vary shapes so no shape repeats within any three consecutive hard ends. Sixth, read only the final beat of every chapter in sequence, which takes under half an hour, and confirm the rhythm by feel.
Repeat the final-beat read after every structural revision. Chapter counts change during editing: cut chapters shift peaks, merged chapters create accidental doubles, added chapters dilute density. Re-run the map whenever the count moves by more than two. The five minutes it costs prevents the slow drift that turns a planned steady book into an accidental dense one.
Keep one page of notes per book: density chosen, hard chapters, shape assignments, and every deliberate deviation with its reason. When the next book starts, that page becomes your personal calibration. Writers who keep it converge on their own signature rhythm within two or three books, and at that point the tool becomes confirmation rather than discovery, which is exactly the right trajectory.