Random Writing Prompts: Using Chance to Create Characters, Settings, and Conflicts

You have a caretaker, an empty building, and a wonderfully particular description of the light. You have moved three sentences around. The caretaker still has nothing to do.

Add one unexpected instruction: return something without admitting it was taken. The corridor becomes a route with witnesses. The caretaker's keys become useful or incriminating. That beautiful patch of light might make it impossible to cross the lobby unseen.

A random prompt can introduce this kind of pressure. It gives you something to respond to when your own choices have begun circling the same possibilities. The result might become an opening, an experiment, or a paragraph you discard after discovering a better idea inside it.

The useful part is what you do with the surprise. A character, a setting, and a conflict become a story situation when somebody wants something and the circumstances make that desire difficult, costly, or revealing.

Give chance a particular writing job

Before drawing a prompt, finish this sentence: For this session, I want to explore… A character's voice is a different project from finding a plausible way out of chapter eight.

When beginning a story, you can leave several decisions open. A person, a place, and a small problem may be enough to produce the first action. You do not need to randomize an ending before discovering what the opening makes possible.

For character exploration, keep the person you already have and vary what they want or what they might lose. An orderly caretaker returning a borrowed key is one character study. The same caretaker deciding whether to lend that key to someone they distrust is another. Their response reveals more than another adjective on a character sheet.

For setting development, try changing who uses the place or what suddenly becomes unavailable there. A greenhouse seen by its gardener, an inspector, and someone sheltering from the rain offers three different maps of what matters. A setting prompt becomes useful when it affects movement, attention, or choice.

When creating conflict, combine a goal with resistance. Someone wants the room cleared; someone else needs its contents. Conflict can be quiet: two reasonable people may need incompatible things at the same time.

To escape repetitive ideas, preserve a familiar premise and change one habit. If every disagreement in your stories ends with somebody walking out, draw a circumstance that requires both characters to stay. Then discover what they do instead.

For technique practice, choose the craft task yourself. You might practice dialogue in which a request is never stated directly, or description filtered through one character's priorities. Let the prompt supply subject matter while the technique remains fixed.

For an existing draft, protect what the story has established. Draw possible costs, interruptions, resources, or discoveries rather than replacing the protagonist or world. Generate three alternatives, write a consequence for each, and choose the one that fits. An unexpected visitor might redirect a scene; an unexplained spaceship probably will not help a carefully grounded family drama.

These are different reasons to use prompts, not diagnoses of a single condition called writer's block. Sometimes the difficulty is a missing decision. Sometimes it is fatigue, uncertainty about the story, or a need to revise what came earlier. A prompt is one available tool.

What research can tell us about the surprise

There is credible evidence that a constraint can sometimes help people produce work judged more creative. In two experiments, Catrinel Haught-Tromp asked participants to write two-line greeting-card rhymes. Rhymes requiring a particular concrete noun received higher creativity ratings than less-constrained rhymes. One experiment supplied the nouns; the other required participants to use nouns they had generated themselves. Haught-Tromp's study offers a useful example of a small content requirement helping a particular writing task. It does not establish that random fiction prompts improve every writer's work.

The broader picture is conditional. Acar, Tarakci, and van Knippenberg's integrative review brings together conflicting findings about constraints across organizational and creative contexts. Restrictions differ in what they limit, how people experience them, and the circumstances in which they operate.

Time pressure can matter too. In product-design experiments, Moreau and Dahl found that input constraints encouraged more creative processing when participants were not also under significant time constraints. That is adjacent evidence, not a prescription for fiction workshops. It gives us a sensible reason to make timers optional when a combination already demands substantial interpretation.

There are also different kinds of creative work inside a writing session. Listing six possible obstacles involves generating alternatives, often called divergent thinking. Making a caretaker, a greenhouse, and a missing parcel belong in the same convincing scene requires selecting and connecting ideas. Research such as Olson and colleagues' Divergent Association Task examines a particular aspect of thinking through relationships between words. It does not score a story's emotional force or narrative coherence.

Nor does greater strangeness automatically mean greater value. Orwig and colleagues found that creativity ratings in three divergent-thinking tasks stopped increasing beyond a certain computed semantic distance. Their research summary concerns those tasks and measures. It cannot tell a writer exactly how unusual a fictional combination should be.

For practice, the implication is modest: give yourself something concrete to work with, then notice whether the constraint creates possibilities or consumes all your attention. A prompt should leave you room to make consequential choices.

Build categories that contribute something

A list of unrelated nouns can be enjoyable, especially for freewriting. If you want a scene, it helps to separate ingredients by the work they do.

• Character: Whose habits, knowledge, obligations, or blind spots shape what happens? A role plus one unfinished situation often gives more purchase than a name alone.
• Setting: What can happen here, and what does the place make difficult? Include working conditions, other people, or a change in the environment.
• Goal or motivation: What does someone want now, and why does it matter to them? Recovering a coat is an immediate goal; keeping a promise explains its importance.
• Obstacle: What prevents the easy solution? Resistance might come from another person's needs, a practical limitation, or the character's own reluctance.
• Object: What can be handled, exchanged, concealed, damaged, or misunderstood? An object can change the action instead of merely decorating it.
• Tone or atmosphere: How will the material feel? The same lost coat can support brisk comedy, tenderness, or unease through different details and rhythms.

Conflict emerges from relationships among these ingredients. An obstacle is not necessarily a villain. Rain can delay a delivery, but a recipient who refuses to accept it creates a different kind of scene. Choose the kind of resistance you want to explore.

Be specific about the pressure, flexible about the answer

“Write about a mysterious person” leaves nearly every decision untouched. Try: “A regular customer asks for a receipt for something the shop has never sold.” Now somebody has a reason to speak, a detail to investigate, and several possible interpretations.

At the other extreme, naming the customer, explaining the deception, identifying the culprit, and prescribing the ending leaves much less to discover. A useful prompt often supplies a detail and an unresolved relationship. Its backstory remains yours.

Specificity is a choice, not a ladder where more is always better. An experienced writer may want only “a promise changes hands.” A beginner may appreciate a person, a place, and an immediate task. Start with enough structure to act; add another ingredient only when it earns its place.

The instruction attached to a prompt matters as well. In a noun-to-verb experiment, Prabhakaran, Green, and Gray found that an explicit creativity cue changed the semantic distance of participants' responses. For our purposes, it is useful to distinguish content from a craft instruction: “a misdelivered parcel” supplies material; “make its meaning change during the conversation” tells the writer what to attempt. That exercise design is an editorial application, not a tested result from the experiment.

Turn three lists into a workable situation

Here is a small reusable set. Choose one entry from each column separately; the rows are lookup positions, not preassembled stories. The conflict column combines a goal with resistance to keep the initial system manageable.

Suppose you select the bicycle mechanic, the greenhouse, and the error that benefited a friend. Before drafting, write one sentence connecting them: the mechanic has repaired the greenhouse's delivery bike and notices that a friend's prize-winning plants have the wrong labels.

That connection is invented after selection. The mechanic's practical knowledge, the inspection, and the friend's advantage now bear on one decision. You could instead make the error a delivery invoice or a mistaken invitation. The ingredients allow several stories.

A quick test is to complete: Because this person is here and wants this, they must… If the sentence has no consequence, change the relationship before collecting more ingredients. A greenhouse need not appear in every paragraph, but its conditions should affect something.

Build lists around the kind of fiction you want to explore. A flat phone battery can create an obstacle in a contemporary story; in a medieval setting, it needs reinterpretation or replacement. You could choose a period-appropriate obstacle list after choosing the setting. That preserves some surprise while making the later selection depend on the earlier one. If you want surrealism, you may prefer to keep the collision. Compatibility is relative to the task.

Count combinations without counting imaginary masterpieces

Five choices in each of three categories give:

5 × 5 × 5 = 125 possible entry combinations

This counts all triples when every character can be paired with every setting and conflict. If selections are independent and each of the five entries has equal probability, one particular triple has probability:

1/5 × 1/5 × 1/5 = 1/125 = 0.8%

The same counting rule handles unequal lists: five characters, four settings, and six conflicts produce 5 × 4 × 6 = 120 entry combinations.

Those counts need interpretation. If a five-entry character list contains the same wording twice, it has only four distinct character labels. With five distinct settings and five distinct conflicts, there are then 4 × 5 × 5 = 100 distinct label combinations, although the picker still has 125 underlying entry combinations. The duplicated character label also receives twice the probability of a single occurrence under equal entry weights.

Weights change frequency too. Give one of five distinct characters weight 2 and the others weight 1: that character receives 2/6 = 1/3 of character draws. Paired independently with one particular setting and conflict from uniform five-entry lists, its triple has probability 1/3 × 1/5 × 1/5 = 1/75, about 1.33%. All 125 triples remain possible because all weights are positive.

Repeated draws can repeat a triple. Making 125 draws does not guarantee seeing all 125 possibilities. Removing selected entries, filtering incompatible pairings, or changing weights changes the procedure; the original equal-chance calculation may no longer apply.

Most importantly, 125 combinations do not mean 125 coherent or worthwhile ideas. A smaller set whose ingredients suggest decisions may be more useful than a huge collection of interchangeable curiosities.

Keep, reinterpret, replace, or reroll

An awkward pairing deserves a little attention, but it does not deserve obedience.

First, see whether you can describe one plausible connection. “Greenhouse” and “stagehand” might meet because a theatre borrows plants for a set. Perhaps the person delivering them recognizes something that should not have left the building. The apparently awkward pairing may contain the scene's most useful relationship.

You can also reinterpret an ingredient. A greenhouse could appear in a photograph, a memory, or a disputed redevelopment plan while your existing scene stays in a kitchen. Make that reinterpretation do work: the photograph might reveal who knew about the plan.

If one ingredient breaks the task, keep the others and change that one. For a dialogue exercise, an obstacle that immediately separates both speakers may be inconvenient. Replace it with a reason they must negotiate. You can choose a replacement deliberately or draw another; these are different operations, and both are legitimate.

A reroll may produce the same result. Decide whether you want another unrestricted draw or a different option. For the latter, temporarily leave out the rejected entry, then restore it when you want the full list available again.

One optional practice rule is to attempt a connection before requesting a replacement. Another is to draft one short passage before judging the whole premise. Neither rule requires continuing with unwanted subject matter. Change or reject a prompt that crosses a boundary, conflicts with the task, or asks for a form of participation you cannot use. No personal explanation is necessary.

Random selection generates candidates. Choosing which ones deserve your attention is part of writing.

Three exercises that put the ingredients to work

The following exercises, combinations, and samples are original illustrations, not participant results. Each practices a different skill. Use the suggested lengths as a comfortable stopping point, and leave the timer off if it gets in the way.

Recover the Coat: write an opening with immediate pressure

Objective: Begin a story through action, competing claims, and a small consequential choice.

Prompt categories: Character, setting, and conflict. Supply a specific object after seeing the combination.

Illustrative combination: A volunteer packing donated coats; a town square while a festival is dismantled; recovering a borrowed object accidentally sold to someone who considers it theirs. Here, the object becomes a coat borrowed to keep the volunteer warm during the sale.

Instructions:

1. Select one ingredient from each column of the table, or use the combination above.
2. Decide what the viewpoint character wants in the next few minutes and why the other person's position makes sense.
3. Write one opening paragraph of roughly 120–180 words. Include a physical detail that affects the action and one brief line of dialogue if useful.
4. End with an action that complicates the situation. Leave the larger outcome unresolved.

Illustrative opening:

Nell recognized her sister's coat by the yellow stitches under its left sleeve, just as the woman wearing it stepped over the last coil of bunting. Nell had borrowed it for the cold morning shift, left it on the chair behind the donations table, and promised to bring it home without festival mustard on the cuffs. “That one wasn't for sale,” she called. The woman turned and held out a receipt, its corner damp against her glove. Five pounds. Nell had written the price herself. Behind them, the trestle tables folded with a sound like applause. Nell opened the cash tin. The notes were gone; the charity treasurer had collected them ten minutes earlier. “It's my sister's,” she said. “And mine now,” the woman answered, pulling it closed over a thin cardigan. At the bus stop, a driver lifted his hand. Nell put the tin down and stepped onto the bus before the woman could.

The paragraph connects its details. The sale explains the receipt; the missing money prevents an easy refund; the departing bus creates a decision. The buyer has paid and needs warmth. Nell has made a mistake and owes someone a promise. Neither needs to become a villain for the scene to move.

Adaptations: Begin with three sentences: recognition, resistance, action. For another pass, write from the buyer's viewpoint without changing the facts. Experienced writers can retain the situation but change its tone, noticing which details make the exchange comic, tender, or threatening. An opening does not have to resolve the problem to have accomplished something.

The Box by the Door: let dialogue carry two agendas

Objective: Explore voice and subtext while two characters pursue different immediate goals.

Prompt categories: Two characters, one goal for each, a shared circumstance, and an object. Draw from a character list twice, setting aside the first entry if you want distinct roles. Choose each immediate goal separately, then adjust one if the goals create no friction.

For a small goal list, try “clear the space,” “keep an object in use,” “delay a decision,” or “obtain permission without asking directly.” Useful circumstances include an inspection due, an audience arriving, or transport about to leave. The writer establishes the relationship between the selected goals.

Illustrative combination: A retiring caretaker wants a mislabeled box removed before an inspection. An apprentice stagehand needs its contents for that afternoon's performance. They have one trolley; the loading van is waiting. Beneath a label reading DISPOSAL, the box contains a handmade moon from an old production.

Instructions:

1. Give both speakers a practical objective and a reason they cannot simply agree. Keep the shared situation fixed.
2. Write 12–16 spoken lines. Let vocabulary and attention distinguish the speakers; elaborate accents are unnecessary.
3. Leave one underlying concern unstated. Include an action that changes the meaning of a later line.
4. End when the negotiation shifts. A complete resolution is optional.

Illustrative scene:

“That box goes with the van,” Mr. Vale said.

“It says disposal,” Iona said. “It doesn't say when.”

“The inspector comes at twelve. The door must open all the way.”

“The moon's in there.”

“Then the moon is obstructing a fire exit.”

“Ours split during rehearsal. Yours has a wooden frame.”

“Mine has a split too. Back seam. You won't see it from the stalls.”

“You remember where.”

“I remember everything people put back broken.”

Iona lifted the box onto the trolley. Under it lay a folded brown coat and a thermos with VALE painted down its side. She moved them onto the bottom shelf.

“That's my going-home pile,” he said.

“The trolley's going upstairs first.”

“The van won't wait for your scenery.”

“We only need the moon till four.”

“Who's bringing it down?”

“I finish at four,” she said. “You could show me the seam.”

Vale speaks in deadlines, instructions, and remembered damage; Iona looks for exceptions and ways to keep things useful. His coat and thermos connect the scenery's removal to his own departure. Her final offer can be heard as a repair request and an invitation to stay. The scene leaves that second meaning available without explaining what either character feels.

Adaptations: Reverse the goals so the apprentice wants everything cleared and the caretaker needs the moon. Rewrite four lines while keeping the intended subtext. With a partner, draft privately from the same combination, then compare what each version leaves unsaid. Read aloud only if both writers want to; a silent exchange works too.

A Doorway That Stays: build a world through consequences

Objective: Develop an invented world's daily life and tensions from connected rules.

Prompt categories: Setting or environmental condition, social rule, and unusual object. For each, make a short list and select one entry. Alternatives might include a town divided by seasonal water, a requirement to share storage space, or a tool whose benefit creates an inconvenient obligation.

Illustrative combination: A town whose streets rearrange at sunrise; every household must maintain a public passage; a brass token holds a doorway's connection to the same street for one hour.

Instructions:

1. State how the unusual condition works well enough to reason about it. You can invent the rule without explaining its ultimate cause.
2. Derive an ordinary adaptation, an institution, and a cost from that condition. Connect them with “because,” “therefore,” or equivalent reasoning.
3. Decide who benefits from the social rule and who finds it burdensome.
4. Write a short world description that ends with an exception or dispute capable of starting a story. Check that changing one rule would affect the others.

Illustrative world:

In Orrel, the houses stay where they are, but the streets rearrange at sunrise. A lane that reached the mill yesterday may lead to the washhouse today. Doorsteps meet different roads; the distances between neighbors change. People write destinations on parcels, never directions.

Because a rearrangement can leave whole streets without a useful onward route, every household must keep one passage open from morning bell to evening bell. A shop may offer a side corridor. A small house may have to admit strangers through its kitchen. The council supplies washable runners for the floors; it does not supply privacy. On busy mornings, breakfast is eaten standing against the stove.

Brass tokens keep particular doorway-to-street connections unchanged for an hour, even through sunrise. A token must cool until the following day before it works again. The council lends its small stock to the flour carts, which cannot turn around in the narrow new lanes. Bakers rise before the rearrangement to collect their sacks. Everyone else waits for the children who earn breakfast money scouting the morning routes.

Households beside a token-held doorway receive customers early, so shopkeepers campaign to have the flour stop moved to their street. Residents whose kitchens become public shortcuts petition for the opposite. The same rule that keeps the town passable makes some homes profitable and others exhausting.

On the first morning of winter, a council token appears in the lock of an unlicensed school. Its borrowed street keeps the youngest pupils from crossing a newly lengthened route beside the river. The flour cart is stranded across town. The route clerk has one hour to decide whether to reclaim the token, and her son's boots are already drying in the school's doorway.

The sample establishes its magic, then follows practical consequences: access, privacy, deliveries, unequal benefits, and a disputed exception. The token affects institutions and ordinary habits. Its final use matters because the earlier paragraphs establish scarcity and competing needs.

Adaptations: Replace the magical streets with seasonal flooding and the token with access to a movable bridge. Keep the underlying dispute and see what changes. Write a visitor's instructions, a shop notice, or a scene from the route clerk's viewpoint. Advanced writers can change one rule and revise every consequence that depends on it.

Make room for different ways of participating

A solo writer can keep the entire exercise private. In a classroom or workshop, explain the craft objective before distributing prompts, and offer a way to change unsuitable material without disclosing why. Fiction does not require autobiographical evidence.

Beginners may prefer three ingredients and a brief example. A more experienced writer might bring an existing scene and randomize only its obstacle, or design a constraint that challenges a familiar solution. Let writers choose a useful challenge rather than treating the number of categories as a measure of ability.

CAST's Universal Design for Learning guidance emphasizes choice, autonomy, and varied support. A practical adaptation here is to allow typing, handwriting, private dictation, an untimed draft, or a short initial sketch. Participants can share prose, discuss a craft decision, compare prompt combinations, or keep their work private.

For a paired session, invite both people to draft before discussing. If they choose to exchange work, begin with something specific that registered: an image, a change in the conversation, a place where the goal became clear. Let the writer ask a question before offering fixes. Liz Lerman's Critical Response Process provides a fuller practitioner framework, including neutral questions and permission before unsolicited opinions.

For collaborative storytelling, divide responsibility: one person supplies a character and goal, another supplies setting and resistance. Agree on the connection before continuing, and allow either person to propose a revision. A shared prompt need not require public improvisation.

One reason to compare interpretations is that outside ideas can shape a group's range as well as help individuals. In Doshi and Hauser's study, access to AI-generated story ideas improved average evaluations while making stories more similar to one another. The authors' research repository documents that finding. AI suggestions differ from random slips; the study does not establish that shared prompt lists homogenize workshops. It does make a useful distinction between individual assistance and collective variety.

Keep a prompt system you can actually use

Start with a few entries per category and keep a copy of the original lists. Add new material when it offers a different kind of action or tension. If every goal involves recovering an object, changing the nouns will provide limited variety. Add goals involving persuasion, repair, concealment, reconciliation, or refusal.

Choose your method according to how you like to work. The same printed table can support all of these approaches:

For the dice method, rerolling sixes gives each remaining face probability 1/5, assuming independent fair rolls. Do not map six to a favorite entry unless you intend to give it extra weight.

Paper can work with or without replacement. Dice do not remember earlier results unless you add a rule. Digital tools vary: a single-list picker does not automatically become a category generator, and a reset may restore eligibility rather than erase history or reload your original wording. Know which change you want.

Use a picker with separate written lists

The Sandbox wheel can supply a selection step, while your notes preserve the three categories. Keep the original character, setting, and conflict lists outside the picker.

1. Load the character list, make a selection, and record the result under Character.
2. Replace the active entries with settings, then conflicts, recording one result from each list. Check that the new list has replaced the old one before drawing.
3. To change only the conflict, keep the accepted character and setting in your notes and draw again from the conflict list. The same conflict may appear again.
4. For a fresh combination, reload the original category lists. For a round without replacement, retain a separate record and check that used entries are no longer eligible; remove them manually if needed.

These are three sequential selections, with accepted ingredients retained in writing. They do not require a built-in category lock or a combined three-wheel control. Paper slips or a suitable dice table can perform the same task.

Check what helped, then revise the lists

After a session, ask what happened relative to your purpose. Did you begin? Find an action worth exploring? Hear two distinct voices? Produce a world rule with consequences? Is there anything you want to keep?

Record a brief answer, along with any replacement you made and whether it improved the fit. Across several sessions, look for patterns. An obstacle that repeatedly clashes with the settings may need rewriting. A tone list that makes everything sound ominous may need more range.

These observations describe usefulness, completion, variety, and your own judgment. They are not standardized creativity scores. A modest passage can fulfill a dialogue exercise; a dazzling premise can fail to help with the chapter you meant to revise. Keep the prompt system and the writer's work separate when deciding what needs adjustment.

Eventually, the slips, numbers, and wheels can disappear from view. Nell still has to get the coat back. Iona has offered a reason for the caretaker to stay. Orrel's clerk has an hour and two obligations. Chance supplied ingredients; each piece begins to matter when someone must choose what to do with them.

Sources and research notes

The references below support the mathematical models, research summaries, or practical guidance discussed here. Worked scenarios and suggested activities are editorial examples unless explicitly identified as recorded calculations.

The prompt lists, exercises, fictional samples, and suggested procedures are original editorial techniques. They have not been tested as an intervention package. All combination counts and probabilities were independently calculated under the stated assumptions.

• Writing constraints: Catrinel Haught-Tromp, The Green Eggs and Ham Hypothesis: How Constraints Facilitate Creativity, 2017, Psychology of Aesthetics, Creativity, and the Arts, 11(1), 10–17. The account uses the author-university copy of the original paper. Its rhyme tasks support a bounded finding, not a general fiction-writing guarantee.
• Context and time pressure: Acar, Tarakci, and van Knippenberg, 2019 is an integrative review, not a meta-analysis establishing a universal optimal constraint level. Moreau and Dahl, Designing the Solution, 2005 concerns creative product design; its account here is limited to the accessible publisher abstract.
• Instructions and association: Prabhakaran, Green, and Gray, Thin Slices of Creativity, 2014, published online in 2013. The creativity-cue finding concerns noun-to-verb responses. It does not experimentally validate the craft instructions proposed here.
• Divergent thinking and distance: Olson and colleagues, Naming Unrelated Words Predicts Creativity, 2021 examines a divergent-thinking measure. Orwig, Luchini, Beaty, and Schacter, 2025 is an accessible conference summary of their published semantic-distance research. Numerical distance thresholds are not assigned to this article's invented premises.
• External story ideas: Doshi and Hauser's repository for their 2024 study supports the distinction between improved individual evaluations and reduced collective diversity. The study investigates AI-generated suggestions, not random prompt tables. Its data were not reanalyzed for this article.
• Participation and feedback: CAST UDL Guidelines supplies educational guidance on choice and support; Liz Lerman's Critical Response Process documents practitioner feedback methods. The workshop adaptations are recommendations, not demonstrated effects of randomized prompts.

The picker instructions use sequential lists and an independent written record. They make no claim of a combined-wheel workflow, category locking, a live-service audit, or measured creative benefits. Whenever used entries must stay out of the pool, check the remaining list before the next selection.