Back to the table

109 Intermediate Tf Try-Fail Cycle Plot & Beats

Try-Fail Cycle

The hero tries, fails, learns, and tries again.

What it is

A try-fail cycle is the smallest complete unit of forward motion in a story. A character wants something. She makes one concrete attempt to get it. The result is something short of a plain success. Afterward she is standing somewhere new, and only that new position pushes her into the next attempt. Chain these units together and you have the middle of a novel.

Each cycle has four parts. First comes a goal, clear enough that a reader could check whether she reached it or missed it. Second comes an attempt she picks herself, with a method she could explain to someone else. Third comes an outcome other than plain delivery of the goal. Fourth comes a changed situation, and the next goal grows out of that change.

Take away any one part and the cycle breaks in a way you can spot. Without a goal, you get activity. Without an attempt she chose, you get events happening to someone. Without a complicated outcome, you get an errand that went fine. Without a changed situation, you get stasis: a chapter you could cut without the book noticing.

Here is the quick test for a real cycle. Look for a loss. Pain alone doesn't count. After the attempt, she has to be missing something she had before: a resource, an option, or a belief about herself. If she comes out of the attempt with the same money, the same choices and the same opinion of herself, the attempt didn't happen, however dramatic it was to write.

The term comes from writing workshops, not from scholars, and nobody is on record as inventing it. The clearest named version belongs to Algis Budrys, in the 1990s. He asked for roughly three attempts, each one bigger than the one before. The later shorthand, "yes, but" and "no, and", is credited to Mary Robinette Kowal, on the Writing Excuses podcast. For how the idea grew up, and how cycles are ordered across a whole book, see rising action. What follows is how a single unit works: the outcomes it can have, the losses that count, how many cycles a book can carry, and how to spot the ones doing nothing in a manuscript.

Most writers are taught two outcomes. There are four, and most drafts use only the first two. The fourth is the strongest tool you have here.

The plain "no" is the weakest outcome, and it should be rare. Use it once, early, to show that the opposition is real and that refusal can happen. That gives the book a floor. Two of them in a row read as the author marking time, because the reader decides the attempts have no price.

"No, and" is the workhorse of a middle. When most writers say try-fail, they mean this one. Its value lives in the "and" part. That part turns a dead scene into the cause of the next one.

The "yes, but" is the most neglected outcome. Writers avoid it because they fear a win will let the pressure out. The effect runs the other way. After four clean failures, the reader can predict the fifth. Once the shape of a book can be predicted, the quality of the prose stops mattering. A partial win breaks the prediction. It keeps the pressure on, and it stops her from backing out, because now she has something to protect.

The fourth outcome is rarely planned, and it does something no other outcome can. She brings about her own ruin and stays competent the whole time. Nobody beat her. The success itself is the trouble.

The craft of the fourth outcome is that the damage has to come from the success. Damage that merely shows up at the same moment doesn't count. Say she wins her court case, and then a car runs her over outside the courthouse. The crash is a coincidence with bad timing. Now say she wins, and the ruling makes her pay a co-defendant's costs. Here the damage grows out of the win, and that is the kind you want. To check, ask one question: if the attempt had failed, would this damage still have happened? If the answer is yes, you've written a "no, and" with a prize attached.

Timing matters too. The reader has to see the win and the bill at different moments. Put them in the same paragraph and the win rings false. Leave a gap of a page, or a chapter, or two chapters in a long book. Then send the bill. The reader's dread fills the gap.

A book doesn't need all four outcomes, and it doesn't need them in equal shares. Aim for a pattern instead. Don't put two of the same type next to each other. Use the plain "no" at most once in a book. Make the cycle just before the collapse a yes, either plain or costly. A protagonist who reaches all is lost without ever having won anything has nothing there to lose.

An attempt only counts if, afterward, she has lost something in at least one of three categories. Feeling bad doesn't count. The three are resources, options and beliefs.

Resources are money, time, favors, allies, objects, places she can go, and a surprise she was holding back, such as a piece of information. They're the easiest losses to write and the easiest to check, because you can count them. They're also the shallowest. A book where she only ever loses resources reads like logistics.

An option is a route that is now closed for good. A route that "would be harder now" doesn't count. Think of the friend who refused and then told someone, the window that's been boarded up, the lawyer who took the other side's retainer. Closed options are what make a climax feel unavoidable. How to order them across a middle is part of rising action.

A belief is something she thought about herself, about someone else, or about how the world works, and can't think anymore. The procedure works if you follow it. Her brother would never have agreed to this. She can do this without lying to anyone. Belief losses drive character, and they're the ones drafts usually leave out.

Belief matters most, because only a lost belief changes what she's willing to do. Less money means a cheaper plan. Fewer options means whichever plan is left. A lost belief means an act she had ruled out. A satisfying ending rests on her finally doing what she swore in chapter two she never would. Every act she has ruled out needs a belief knocked down first, and the failed attempts do the knocking. In a book of purely outside setbacks, the climax is arbitrary. Nothing makes the last act possible except the author's need for it.

You can check this with the forfeit test. For each attempt, finish the sentence this attempt cost her ______, then label the loss R, O or B, for resource, option or belief. The test shows what kind of bill each attempt paid. That tells you whether your middle will produce an arc or only a plot. Say a middle comes out as twelve Rs, two Os and no Bs. You have competent action, and it will feel emotionally flat however spectacular it is. If more than three attempts in a row go by with no B, the inner line has stalled, and it won't start again on its own. Rising action has a simpler audit, the receipt test, which only asks whether any price was paid at all.

The table's one-line summary of this element says "The hero tries, fails, learns, and tries again." "Learns" is the word hiding the most. Learning is where the cycle stops being mechanics and becomes character.

Picture a heroine who learns efficiently. Every failure teaches her the right thing, and she acts on it at once. She's competent, but she isn't dramatic. Each attempt is a better-informed copy of the one before. She never resists the evidence. She's a professional optimizing a process, and nothing about anyone gets revealed. The drama lives in four patterns that break from that.

She learns the wrong thing. The failure was caused by A, and she blames B, because blaming B protects a belief she needs. A woman loses a hearing because her evidence was weak, and she blames a corrupt panel. Her next attempt aims at the wrong target and fails even harder. You get escalation and characterization at once, since the mistake shows what she can't afford to know. Of the four patterns, this one is the richest and the most neglected.

She learns the right thing and can't act on it. She sees that the only road left runs through her sister, and she refuses that road, for now. Her next plan is worse, and she picks it for an inner reason. Of the four, this one feels the most familiar. It needs a reaction beat on the page, and that beat is the territory of scene and sequel.

She learns the right thing one attempt too late. The information that would have saved the last attempt arrives as the next one begins, and the cost of the gap is permanent. The shape is the classic tragic form. Repeated, it builds dread into the book's rhythm.

She learns something true about the world and false about herself. The two lines can split apart. She correctly works out how the fraud was done, and wrongly decides that nobody but her can be trusted with what she found. The plot moves forward while the character moves backward, which makes a negative arc. For more on that shape, see character arc and the lie.

To track learning, add a conclusion line. For each attempt, write after this, she believed ______. Under it, write a second line: was she right? If the answer is "yes" every time, you have a protagonist who is never wrong. A heroine who is never wrong is the leading hidden cause of sag in a book whose events are sound. The conclusion also has to show up in her next action, not in her thoughts. A character who states her lesson in words is summarizing, not learning. The same principle runs through show, don't tell.

How many cycles should a book have? Everyone gives the same answer: three. That number comes from oral storytelling. No law of narrative requires it.

In 1909 the Danish folklorist Axel Olrik set out his "epic laws" of folk narrative, which had first appeared in Danish in 1908. One of them is the law of three: characters, objects and repeated actions in traditional tales tend to come in threes. Three is the fewest items you need to set up a pattern and then break it. Three is also as many as an audience listening aloud can easily hold in mind. In 1928 the folklorist Vladimir Propp found the same tendency in the tales he studied for Morphology of the Folktale, and he called it trebling of functions. Two older brothers fail, and the youngest wins. Two houses fall, and the third stands. Two tasks get done, and the third nearly kills her.

Craft teaching carried the number over unchanged. Budrys asked for roughly three attempts, each bigger than the last. David Farland, a writing teacher and author who worked beside Budrys in the Writers of the Future contest from 1991 on, built the method in Million Dollar Outlines (listed below) around a minimum of three. The family tree of the idea is laid out under rising action.

Three makes a fine working default. But it comes from a spoken form told in one sitting, and a 95,000-word novel doesn't need it. Followed as a rule, it produces thin books with three attempts, and books padded out with attempts that change nothing. Three things should set the number instead: the length of the manuscript, the genre, and how much the protagonist has to lose.

Start with length. A novella has room for one or two cycles. A commercial novel of ordinary length supports four to six. An epic-sized fantasy can run to twelve or more, shared out among several viewpoint characters, and there the real question becomes how to weave them together. No length can afford an attempt that changes nothing. A shorter book just exposes it sooner. Rising action covers how to budget cycles for each length.

Then genre. Where the reader already knows the outcome, as in romance, cozy crime and most series fiction, the cycles can't be about whether she'll succeed. They're about what success will cost, so you write fewer of them and build each one with more care. Where the outcome is unknown, you can afford more cycles, and each one does suspense work too. Which side your book falls on is part of its reader promise.

The real limit is her stock of things to lose. At the start of the middle, list her resources, options and beliefs. Each cycle spends at least one. Five things to lose means five cycles. The point where she runs out is where all is lost sits. A middle that feels too short had too little to lose. A middle that sprawls had too much, spent too slowly.

Cycles also need rhythm. Cycles that share the same shape and length read as one cycle repeated, so four identical cycles feel like the same one four times. You have three dials to turn: size, outcome type, and who starts the cycle.

Size first. Let one cycle run six chapters and another four pages. Spend resources in the short ones and beliefs in the long ones, because a belief doesn't dissolve in a single scene. Outcome type next. Don't put two of the same type back to back. Late in the book, when the reader is keeping count, never put two "no, and" outcomes in a row. Then who starts it. A cycle she launches feels different from one where she's answering the opposition. Alternate them, or the middle turns passive or relentless. The opposition's schedule is a matter of conflict, and who holds the initiative is part of rising action.

The cycle works the same way at every scale, but it feels different at each one, which is why talk about it gets muddled. There are four scales.

At the scale of a scene, she wants something, makes two or three moves against opposition, and leaves with nothing or with a spoiled version of what she wanted. You feel this as tension, and the structure stays invisible. Still, the attempt-and-outcome shape is what separates a scene from a passage. The want at this size is the scene goal.

At the scale of a sequence, three to eight scenes make one attempt, such as her whole approach to the auditors. You feel it as a plan and what became of it. Most people attach the term "try-fail cycle" to this size. The sequence is also the best size to plan in, and you can write one in about a week.

At the scale of the middle, the second act is built by nesting. Four to six sequence-sized cycles add up to one large cycle: her whole approach, tried and found wanting. You feel this as the book's argument. The midpoint usually turns the large cycle. She stops forcing her first approach and tries another.

At the scale of a series, each volume is one attempt at the series-wide problem. The trap is a volume whose own cycle stays open because its failure feeds book three. A book like that reads as going nowhere. So every volume closes its own cycle. The series cycle is built from what's left over, the price paid by the volumes that closed.

Here is one ladder with all four outcomes in it and every forfeit labeled. Iris Kendrick teaches history at a comprehensive school. Dominic Bayliss, fifteen, is in her form group. He's threatened with permanent exclusion over an assault she thinks he didn't commit, and she wants to stop it. At the start of the middle she has four routes and three beliefs. She believes the procedure works if you follow it. She believes she knows better than his family what's good for the boy. And she believes she's doing this for Dominic.

Attempt one. She asks the head, Mr. Ovenden, to delay the hearing while she finds the second boy who was there. He refuses, flatly and politely. No. She loses only that route, so the label is O. Even a plain no spends the route she asked for. The attempt sits early to show that refusal is real here.

Attempt two. She finds the second boy, and he agrees to give a statement and gives it. Yes, but his mother withdraws her consent for him to appear at the hearing, and without him the statement is worth little. It also shows Ovenden that Iris is working outside the official procedure. She loses a resource, the value of the statement. She also loses a belief, that the procedure protects her, since she has been penalized for something it doesn't forbid. R and B.

Attempt three. She persuades Dominic's mother to speak at the hearing. The mother does badly, and the panel learns of an incident from two years earlier that wasn't on the file. No, and. The hearing is in worse shape than when she started. She loses an option, since the mother can't be used in any later move. She also loses a belief, that she knows better than the family what's good for the boy. O and B.

Attempt four. She takes the story to a local reporter, a move she swore in chapter three she would never make. The article runs. Yes, and it is worse than failing. With the press watching, the governors can't back down. In a town of nine thousand people, Dominic can be identified. The exclusion is confirmed the day the piece appears. She loses everything she had left, including her last belief, that she's doing this for Dominic and not for the part of herself that never loses.

Four cycles, and nothing bigger than a newspaper. She has no moves left, because she made the one she had ruled out. Each outcome type appears once, and no two of the same type sit side by side. The beliefs fall in rising order of what each one costs to lose, so the fourth attempt breaks her with nothing left to cushion the fall.

Four neighboring elements sit close to this one, and it helps to keep them apart. Rising action is about arranging cycles: where they go, how they escalate, and the difference between volume and difficulty. The try-fail cycle is about what goes inside a single one. Cause and effect is the chain of causes, with its own "therefore / but" test for linking any two scenes. The cycle is just one shape that chain can take. Agency asks whether she gets to choose at all. A cycle assumes she does, since someone has to pick the attempt. A draft where every attempt is another character's suggestion has an agency problem dressed up as a cycle. Scene and sequel covers the reaction after a failure: taking it in, weighing bad options, committing to one. The belief loss gets written there.

The cycle is a useful tool, and plenty of published fiction works without it. There are four fair exceptions, and each one has something else doing the job.

The first is the internal attempt. Literary and upmarket fiction often follows attempts to think or feel something: to forgive, to confess, to stop wanting, to remember exactly. The goal is a state of mind, and nobody else sees the failures. Force an outside cycle onto a book like that and you get errands stuck on, like visiting the house or looking for the letter. A better unit to track is the cycle inside the sentences, such as the thing she lets herself know around page forty and takes back around page sixty.

The second is the character who never tries. Some novels are about paralysis, and doing nothing is the subject. Forcing attempts on them would wreck them. What holds the reader instead is pressure from outside, other people acting on her, and a situation that gets worse without her consent, followed by one late act. Agency covers the line between refusing, which is a choice, and being carried along, which is a structural fault.

The third is the mystery. Here the attempts are lines of inquiry inside the detective's head: a hypothesis, a test, a collapse. The cycle is fully there, and its currency is mostly belief. A common mistake in drafts is adding more interviews instead of raising the detective's exposure. The book gets busy and stays still.

The fourth is the quiet novel, where failure is ambient. She fails at living, in small ways, all the time. The book renders this more than it stages it. Its units are months and inches, with perhaps two or three clear attempts in the whole book. A rigid structure of three makes it melodramatic, and a book like that can't survive melodrama.

September C. Fawkes is strong on these limits in her essay of December 22, 2025, which is free to read and listed below. She teaches the cycle and still says where it falls short. The cycle assumes one overarching goal. It struggles when the goal changes partway through. And taken literally, it gives you constant failure with no progress. So use the four outcomes and the three kinds of forfeit as a lens to look through. They aren't a quota. If you find two attempts where you expected six, you may be writing a different kind of book, and it may not be broken at all.

Why it matters

The ending is only as believable as the failures before it. When a climax feels earned, what the reader feels is the memory of everything that was tried and failed. When the last attempt works, the reader has to be able to answer "why didn't that work before?" The only good answer is a change the failures produced. She knows something new, or she's now willing to do something new, or she has nothing left to protect. Take the failures away and the ending is just the author's ruling.

The reader is checking whether you'll make her pay. All through a book, a reader is quietly testing whether this writer makes characters pay. If she gets through every attempt undamaged, that question is settled early, and attention drops. That trust has cash value. The problem never shows up under its real name, either. Nobody closes a book like that with the words "insufficient try-fail structure" in mind. The book just feels predictable, the stakes feel low, and the outcome never seems in doubt.

The cycle turns a mood into a worklist. "The middle sags" is a mood, and you can't fix a mood. "There are four attempts in chapters eleven to nineteen and three of them cost nothing" is a list of repairs. The cycle gives you the words for the second kind of sentence.

Failure is hard to write, and the reasons are kind ones. Writers spare their protagonists out of affection and a fear of boring the reader. Cowardice has little to do with it. You've spent months with her. Competence is pleasant to write and pleasant to read. And you control every outcome, with nobody in the room to argue the other side. So the obstacle gives way. There's no grand wish fulfillment in it. The obstacle just gives a little, again and again. You can't see it in one scene. Across twelve, it's fatal.

The tell is an obstacle that turns out easier than promised. Two chapters set up a sealed archive. The archivist is a bureaucrat, and the request needs a magistrate's signature. Then the archivist, who knew her father, lets her in. The scene can be well made. The flaw is self-inflicted: you built the wall and then took it down yourself. The reader registers this without noticing, as a sense that the book is pleasant, and pleasant is fatal to word of mouth.

A resistance check finds the flinches. For every obstacle, write two lines: the price the reader was promised, and the price actually paid. Wherever the second is smaller, you flinched. Expect to find plenty.

Three repairs fix a flinch, in rising order of effort. First, keep the help and send a bill for it. The archivist lets her in only if she tells him why, and then he gossips. The obstacle gave way, but the attempt still cost her, and an accidental yes has become a "yes, but". Second, make the helper someone she wronged in act one, so the help becomes a debt. Third, and structural: decide the outcome and write it down before you draft the scene. Flinching happens in the prose, while a sympathetic character is right in front of you. An outline line such as attempt three: yes, but. She gets the file and Ovenden learns she has it takes the choice out of that moment. That alone makes a strong case for planning your cycles ahead in an outline.

Real-world examples

The Daughter of Time (1951): an investigation with no action at all

Josephine Tey's The Daughter of Time came out from Peter Davies in 1951, the last of her novels published in her lifetime. In 1990 the British Crime Writers' Association put it first on its list of the hundred best crime novels of all time. The book is the cleanest case there is of a try-fail cycle with no physical action.

Inspector Alan Grant of Scotland Yard is stuck in a hospital bed with a broken leg, unable to move and bored. A portrait of Richard III catches his interest, and he asks a question: did Richard have his nephews killed in the Tower? Every attempt in the book is a hypothesis, and every failure is a hypothesis knocked down. Other people fetch the evidence. Marta Hallard, his actress friend, does research for him. Brent Carradine, a young American, brings him sources from the British Museum.

The first test is motive. Richard needed the boys dead to be safe on the throne. But an earlier act of Parliament, Titulus Regius, had already declared Edward IV's children illegitimate. Dead or alive, the boys were no threat to him, and the hypothesis collapses.

The second test is behavior. Elizabeth Woodville, the boys' mother, came out of sanctuary, took a pension from Richard and let her daughters attend his court. A mother doesn't treat her sons' murderer that way.

The third test is the gaps in the record. There was no inquest and no charge. After Bosworth, no bill of attainder named the murder. The confession that supposedly settles the matter first appeared years later, under Henry VII.

Every forfeit in the book is a belief. Grant loses nothing material, and he couldn't, since he can't get out of bed. The book still grips. The lesson carries over to any mystery, and to any novel where the protagonist's strength is information.

The last cycle is the fourth outcome. Carradine succeeds completely. He has built the case, the evidence holds up, and he has found his calling. Then he discovers that the Ricardians, people defending Richard, have been making the same argument since the seventeenth century. Nothing in his case is new, and it changes nothing. The win is worse than a defeat would have been, because a defeat would have left work still to do. Grant answers with his private word, in use since the middle of the book, for a story people believe that isn't true. The word comes from the Tonypandy riots. For crime fiction, the ending is unusually bleak.

The novel's account of the history has been disputed since 1951. That doesn't matter for the lesson. What you're studying is the shape of an investigation where each attempt destroys a belief.

Project Hail Mary (2021): the success that eats the ship

Andy Weir's Project Hail Mary came out from Ballantine Books on May 4, 2021. No recent novel shows all four outcomes more clearly. The hero is a scientist, so his attempts are experiments with logged results.

Ryland Grace wakes up on a spaceship, alone, with his memory gone. A microbe called Astrophage is eating the sun. He's on a one-way trip to find out why a neighboring star is immune.

At the small scale, he forms a hypothesis, builds a rig from what the ship has on hand, and gets a bad result. Usually the result is a "no, and": he has less gear and a fresh problem. At the middle scale, first contact with an alien ship at Tau Ceti brings him a partner, Rocky, along with new limits. Their atmospheres clash, their temperatures clash, and they share no language.

The centerpiece is the Taumoeba sequence, the fourth outcome at its best. A predator of Astrophage turns up and seems to solve everything. Then the cultures die in the nitrogen of the target atmosphere. Grace and Rocky breed a strain that tolerates nitrogen. Yes, it works. The sun's problem is solved.

Then the tolerant strain turns out to chew through xenonite as well. Rocky's whole ship is built of xenonite, and the strain is in his fuel. The success is the disaster. Without the breeding success, nothing would be eating his fuel.

Weir adds one more turn. Autonomous probes have already carried the data home. Earth is saved, and the two of them are stranded. The mission's success has come apart from the survival of the people who made it.

Run the check from earlier: would this harm exist if the attempt had failed? No. The harm exists because of the success, through the very property that made it a success. If you're about to write a fourth-outcome cycle, study this passage first.

A Sorceress Comes to Call (2024): three plans of three different kinds

T. Kingfisher's A Sorceress Comes to Call came out from Tor Books on August 6, 2024. It won the 2025 Locus Award for Best Fantasy Novel and was a finalist for both the Nebula and the Hugo. The book is a dark retelling of "The Goose Girl" from the Grimms. It uses the fairy-tale triple of attempts on purpose, in a novel long enough for you to feel the convention at work.

Cordelia is fourteen. Her mother, Evangeline, is a sorceress who makes her literally "obedient". The girl sees and feels everything and controls nothing of her own body. Evangeline sets out to marry Samuel, a wealthy squire. Samuel's sister, Hester, works out what's happening and moves to stop it. All three plans are Hester's.

Plan one: fill the house with guests, among them Samuel's former mistress, Penelope, to give Evangeline rivals. Samuel is distracted. Evangeline answers by killing Penelope's maid, and very probably Penelope too. The outcome is a "yes, but" far worse than failure. The goal was met, and a woman is dead. What Hester loses is a belief, that she can manage this socially, and no mere resource.

Plan two: Richard, Hester's own former love, proposes to Cordelia, to free the girl from her mother before the wedding. The plan fails, and Samuel and Evangeline marry anyway. A clean "no, and" follows: the attempt is spent, Evangeline is legally in residence, and the opposition is stronger.

Plan three: a rite to remove the source of Evangeline's power, which lives in Falada, a white horse. The first try at the ritual fails over a rule about who may perform which part. The second try, with the horse beheaded and with help from the dead Penelope, succeeds.

The lesson is in how the plans differ. Each is a different kind: social, then legal, then magical. Each one was chosen because the one before had shut that whole category. Tradition fixes the number of plans. Escalation explains why there are three and not seven. And the people around Cordelia, not Cordelia herself, pay for plan three, which is the hardest kind of cost to write without sentiment.

Moby-Dick (1851): two captains, one whale, opposite lessons

Herman Melville's novel came out first in London as The Whale, from Richard Bentley, on October 18, 1851. The New York edition, from Harper & Brothers, followed on November 14 as Moby-Dick; or, The Whale. You can read it free at Project Gutenberg, listed below. The book is long, but two stretches of it show what "learns" means better than anything else in English.

The final chase runs three days, and it works as a sequence of three cycles. On the first day, the whale bites Ahab's boat in two. On the second, more boats are wrecked, Ahab's ivory leg breaks off, and Fedallah disappears. On the third, the whale rams the Pequod and sinks her, and Ahab is dragged under by the rope of his own harpoon. Each attempt costs more than the last, each loss is permanent, and the last one gives him the opposite of his goal. For escalation inside one sequence, you won't find a better model.

The quieter, earlier lesson comes in chapter 100. The Samuel Enderby, out of London, crosses the Pequod's path. Her captain, Boomer, has an ivory arm. He lost the real one when his line tangled in the whale's teeth. He grabbed for a second harpoon, the whale wrecked the boat, and the iron tore him open from shoulder to wrist. He has seen Moby Dick twice since, and both times he kept his boats on deck. Once was enough.

Same wound, same animal, opposite conclusions. Boomer decides another limb isn't worth it and goes back to ordinary whaling, cheerful. Ahab decides the whale owes him an answer, and that decision drives the whole plot. Melville puts the two men in one cabin so you can't miss the contrast. Ahab leaves almost at once. Boomer is living proof that Ahab's reading of the wound is a choice.

In chapter 128 the Pequod meets the Rachel. Her captain has lost his small son from a boat, and he begs Ahab to help search. Ahab refuses. A man built around one loss can't spare a day for someone else's. Here you see from the outside what learning the wrong lesson looks like. Ahab is the classic grievance, and the wound element draws the line between a grievance and a wound.

The technique is easy to borrow. Give a minor character the protagonist's failure and let him draw the sane conclusion. It costs you one scene. In return, the protagonist's mistake becomes a choice the reader can see, instead of a ruling by the author.

The Bookshop (1978): quiet failures, exact structure

Penelope Fitzgerald's short novel The Bookshop came out from Gerald Duckworth in 1978 and was shortlisted for the Booker Prize that year. Hand it to anyone who says the cycle only suits thrillers. The book has three cycles, two of them Florence's attempts, no explosions, and an exact structure.

In 1959, in Hardborough, a town on the Suffolk coast, Florence Green, a middle-aged widow, opens a bookshop in the Old House, a damp building said to be haunted. Mrs. Gamart, the town's grande dame, wants the building for an arts center. She is the opposition, and her voice stays calm throughout.

Florence's first attempt is a large order of Vladimir Nabokov's Lolita, the scandalous book of the moment. She places it after consulting Mr. Brundish, a recluse whose judgment outranks Mrs. Gamart's. The book sells. The outcome is a "yes, but" with a sting: the shop becomes conspicuous, Florence becomes someone with standing worth removing, and the opposition hardens. She loses an option, the chance to stay unnoticed.

Her second attempt is to win an ally. Brundish leaves his seclusion and calls on Mrs. Gamart to demand that she stop. The effort kills him, and he dies on the way home. The outcome is a yes, and the fourth outcome at its bleakest. He succeeded, in that he said what needed saying. And the success removed the only person with influence to match the opponent's, which a failed visit would never have done.

The third event isn't an attempt at all. Mrs. Gamart's nephew, a member of parliament, sponsors a law that lets local authorities buy historic buildings standing empty. The Old House is taken over Florence's head. She is evicted and leaves by train. The loss that counts is her belief, the one the whole venture rested on, that Hardborough wanted a bookshop.

Add it up: three cycles, one of them belonging to someone else, two with "yes" outcomes, in about a hundred and fifty pages. Around them, Christine, the young assistant, fails the eleven-plus, and Milo North, pleasant and hollow, quietly serves the other side. Every failure is domestic, and every one is permanent. If anyone tells you this element belongs to action fiction, the book is your answer.

An everyday scenario: the four attempts that cost nothing

Imagine you've finished a manuscript of 95,000 words. Your beta readers all say the same two things: they loved the final sixty pages, and the middle drags. You got the same note on your two earlier books.

Your lead is Glenys Quayle, a livestock hauler who believes a slaughterhouse is faking its animal welfare records. The middle runs from chapter twelve to chapter twenty-four. Listing her attempts takes twenty minutes, and there are four. She talks to the vet on duty. She reports the plant to the inspectors. She gets a driver from a competing firm to talk. She tries to get copies of the kill-sheet records.

Then you run the forfeit test. The vet is noncommittal, and she leaves annoyed. Cost: nothing. The inspectors promise to look and never do. Cost: nothing, apart from three weeks of story time. The driver talks off the record and confirms what she suspected. Cost: nothing, and the attempt fails the other way too, since he tells her nothing new. The records can't be had. Cost: nothing.

Four attempts, four blank rows. Your middle is empty. Slow and empty are different faults with different fixes, and tightening the prose would never find this one.

The conclusion line tells the same story. After every attempt, she believes the same thing: the slaughterhouse fakes its records and nobody will help. She was right at the start and right at the end. She learned nothing and changed nothing, so she reads as a person with a grievance instead of a protagonist.

The repair is four outcome conversions, with no new chapters, and three of them take a paragraph each.

Attempt one becomes a no, and. The vet, noncommittal, mentions the conversation to the plant manager, the natural thing for a man in his job to do. By chapter thirteen she has lost the element of surprise. She loses a resource.

Attempt two becomes a yes, but. The inspectors do come, with warning. They find nothing and close the file, so any second complaint from her will look vexatious. She loses an option for good.

Attempt three becomes the belief loss, and that one needs real drafting. The competitor's driver tells her that two colleagues from her own firm have driven those same runs and know what's going on. She loses the belief that this is her side against theirs. Her next move now aims at people from her own lunch table, and you're writing a different book from the one you drafted.

Attempt four becomes yes, and worse than failing. She gets the kill-sheets. They're real and damning, and one entry is damning about the date of a run she drove herself. She loses her last belief, that she is the only clean link in the chain.

The scenes stay the same, and so does the chapter count. The middle now has a ladder of forfeits, a protagonist who is wrong twice, and a reason for the final act.

Try it yourself

Call this the attempt audit. It takes an afternoon on a finished draft, or about an hour on an outline. Work in a spreadsheet or on paper, never in the manuscript itself. The six-line template after the last step gives you one row's worth of questions.

  1. List every attempt. Write one line for each time the protagonist deliberately moves toward her goal. An attempt needs a method. If you can't fill in she tried to ______ by ______, it was worry, and it comes off the list. Expect fewer rows than you'd guess: a 90,000-word draft typically holds four to eight.
  2. Label each outcome. Mark each row N for no, N+ for no and worse, Y− for yes but, or Y+ for yes and worse than failing. If you can't decide, the page probably isn't clear either, and that's a finding in itself.
  3. Run the forfeit test. For each row, finish this cost her ______, then add R, O or B. If the true answer is nothing, leave the box blank. The blank boxes are the whole point.
  4. Add the conclusion line. For each row, write after this she believed ______, then was she right? Flag two kinds of row: the ones where she's always right, and the ones where the belief changes nothing about her next move.
  5. Read the outcome column as a string. A run like N+ N+ N+ N+ Y− can be predicted from chapter twelve. Four identical labels in a row give the game away, whatever happens in the scenes.
  6. Check that each attempt grows out of the last. Between each pair of rows, write therefore, but or and then. Every and then is a broken joint, and cause and effect shows you how to repair it.
  7. Run the resistance check. For every obstacle, write what the reader was told it would cost and what it actually took. If the second is smaller, you flinched.
  8. Count the Bs. If there are none, or they all come after the three-quarter mark, the plot escalates while the character stands still. The result is a book where nobody cares about the people.
  9. Hunt for the missing Y+. Nearly every draft lacks one. Look for an attempt that now succeeds cleanly, or a late one that now fails cleanly, and ask what would happen if she got exactly what she asked for and it cost her. Attempts to get information are the best candidates.
  10. Order the forfeits by weight. Rank her beliefs from least to most costly to lose, and check that the draft takes them in that order. Lose them out of order and the middle is full but feels arbitrary.
  11. Fix outcomes before you move scenes. Converting an outcome usually costs a paragraph. Moving a sequence costs about two weeks. Do all the paragraph fixes first, then reread, and many of your structural complaints will be gone.

Attempt __ : she tried to ______ by ______ Outcome: N / N+ / Y− / Y+ Forfeit: ______ (R / O / B) After this she believed: ______ Right? Y / N Link to next attempt: therefore / but / and then Advertised difficulty: ______ Actual: ______

With the sheet filled in, you have your revision worklist, ordered by cost instead of by chapter. For more on diagnosing a finished draft, Steven James's Troubleshooting Your Novel is listed below.

Common mistakes and misunderstandings

Failure that costs nothing. The scene is strong, and nothing moved. The forfeit test catches it. The fix is usually a sentence, not a new scene: her request reaches the wrong ears, or the attempt uses up a favor she'll need in chapter thirty.

Five failures of the same size in a row. Five equal failures are repetition. They don't escalate. If you could swap attempt four for attempt two without changing a word, the attempts have no order. Rank them by what they cost her, not by how big they look on the page. Ordering cycles is the job of rising action, which has its own tests.

Failure caused by an outsider. A stranger's car accident closes the road. The crash is luck, as random as weather. To feed the next cycle, the defeat has to trace back to the attempt: the opposition reacting, a flaw in her method, or the result of an earlier choice of hers. Cause and effect covers when chance is allowed in.

Failure through stupidity. When the reader sees the obvious move and she misses it, failure turns into irritation. Put the difficulty in the problem, not in her wits. To test it, hand the situation to the smartest person you know. If that person would do something different, give her a better plan, or a clear reason for the weaker one: fear, pride, a vow, or ignorance the reader shares. See agency.

Ending the cycle by exhaustion. On the fourth try it works, and nothing on the page says why. Only the count changed. The ending arrives without being earned. The fix: say in one sentence what makes the last attempt possible, then check that an earlier failure supplied it.

Treating three as a quota. Three comes from oral storytelling, and novels have no such rule. Padding a book out to three does as much harm as cramming twelve cycles into a novella. Two well-ordered attempts beat five in random order. Count forfeits, not cycles.

Never letting her win. Constant failure is as predictable as constant success. It reads as punishment instead of pressure, and it throws away the two most useful outcomes. A half-win raises the pressure, because it hands her something new to lose.

"Yes, but" as a tic. The shorthand is a brainstorming tool. Jill Williamson's 2013 post on Go Teen Writers, listed below, uses it to work out what the next scene should be. It was never meant as a sentence pattern. When chapters visibly end on a but, the book looks machine-made. The reader should feel the consequence and never see the formula.

The attempt nobody chose. An attempt the mentor suggested, the plot needed or the author wanted isn't hers, and its failure teaches her nothing. Before every cycle, you should be able to say why she chose this method over the others. An answer like "it was the only thing left" is fair, and late in the book it's the ideal one. But it only holds if the other options existed and were closed.

Saying the lesson out loud. A character who recites what she learned hasn't shown it. The lesson belongs in her next attempt, where she acts differently and the reader does the math. Inner narration that announces the change is the top cause of a middle that feels thin even when it's well built.

A cycle with no goal. Attempts can only fail against something. A middle full of investigating, coping and looking into things has no attempts in it, because nothing is firm enough to refuse her. At the level of the scene, the fix is a scene goal. At the level of the book, the fix is a dramatic question.

Related elements