The Session Shape
Three Acts, Five Beats — A Complete Story in Two Hours
A container for framed scenes. Without one, even excellent scene framing drifts.
Drawing from Field, McKee, Morningstar (Fiasco), and Johnn Four (Five Room Dungeon). Full credits at the end.
I. Why a Session Needs a Shape
Scene Framing tells you how to build a scene. Scene Zero tells you how to start a session. This document tells you what goes between them — how a sequence of well-framed scenes becomes a complete story rather than an open chain.
You can build perfect sentences, write tight slugs, enter at the "but" — and still find yourself ninety minutes in with no sense of where the story is going or how it ends. The problem isn't the scenes. It's that scenes need a container.
The Session Shape borrows two containers and welds them together. The Three-Act Structure (from screenwriting) tells you when things happen. The Five Room Dungeon (from Johnn Four, 2006) tells you what kind of scene each beat should be. Together they answer both questions.
II. The Five Beats
These are the five scenes a session must hit. Other scenes are connective tissue.
| # | Beat | What it is | Why it matters |
|---|---|---|---|
| 1 | Threshold | The crossing-over scene. Character commits and the world reveals its first cost. | Without commitment, no story. |
| 2 | Challenge | A non-combat problem. A puzzle, social obstacle, or moral test. Not a fight. | Forces variety into the session. |
| 3 | Tilt | The reversal. The worst complication arrives as fact. Mandatory. | The hinge. Without it, the session coasts. |
| 4 | Climax | The hardest scene of the session. Direct confrontation with what the Tilt revealed. | The scene the story has been pointing at. |
| 5 | Resolution | The aftermath. Three sentences. What happened, what it cost, who they became. | A story without an ending isn't a story. |
The Five Room Dungeon's classic order is Entrance with Guardian · Puzzle · Setback · Big Climax · Reward. Adapted for solo narrative play, those become Threshold · Challenge · Tilt · Climax · Resolution. The structural function is identical. The names are tuned for story rather than dungeon.
The Tilt and the Setback are the same beat. Two completely independent traditions arrived at the same midpoint reversal. That's strong evidence it's load-bearing.
III. The Three Acts
The five beats anchor the session. The three acts give them a time container.
| Phase | Scenes | Time | Beats Hit | Interludes |
|---|---|---|---|---|
| Act 1 — Pursuit | 3 | ~30 min | Threshold (1) → development → Challenge (2) | 0 |
| The Tilt | 1 | ~10 min | Tilt (3) | 0 |
| Act 2 — Fallout | 4 | ~50 min | Connective scenes → Climax (4) | 1–2 |
| Act 3 — Resolution | 1–2 | ~20 min | Resolution (5) | 0 |
Total: ~10 scenes, ~110 minutes of play, ~10 minutes of journaling and frame-writing. Two hours, give or take.
Five beats are required. Other scenes fill the space and let you breathe. You always know where you are because you can name which beat you've hit and which is next.
IV. Act 1 — Pursuit
Three scenes, ~30 minutes. Your character actively pursues the want established in Scene Zero (or the want that opens this session, if you're continuing). Frame each scene using the Protocol. Use your oracle for everything you don't know about the world. Connect scenes via "but" and "therefore."
The goal of Act 1 is not to succeed. It's to make success harder. Each scene should answer the current question but generate a new one.
Required beats in Act 1:
- Scene 1 — Threshold. The character commits. They cross from "considering" to "doing." The world responds — usually by revealing a first cost they didn't anticipate.
- Scene 3 — Challenge. A non-combat problem. Negotiation, deception, a moral choice, a puzzle. Forces a different kind of scene from Scene 1.
Scene 2 is connective — it deepens whatever Scene 1 opened up.
Act 1 Signal: When the path forward feels genuinely threatened — not impossible, but precarious — Act 1 has done its job. Move to the Tilt.
V. The Tilt
At the forty-minute mark, or after scene 3 — whichever comes first — stop and apply the Tilt. This is mandatory. Do not skip it. Do not soften it.
The Tilt is the structural hinge. Everything before it is setup-through-action. Everything after it is falling toward an ending you couldn't have planned.
Without a forced Tilt, the oracle will let you coast. A coasting session is drift with better-framed scenes.
Ask one question: what is the worst complication that could arrive right now?
It arrives. Write it in your journal as a fact, not a possibility. Build the first Act 2 scene from that fact.
The Tilt is not a setback you recover from in the next scene. It's a fracture. Something that was true at the start of Act 1 is no longer true — a relationship, a plan, a resource, an assumption. Something breaks and stays broken until the Resolution decides what to do with it.
Applying the Tilt by System
| System | How to Apply the Tilt |
|---|---|
| Mythic GME 2e | Set Chaos Factor to 8. Roll a Random Event. Take the result literally — it happened, right now. |
| Ironsworn / Starforged | Pay the Price — hard. No weak hits. The worst reasonable outcome occurs and enters the fiction as fact. |
| Kal-Arath | Roll on the Peril or Complication table. The result enters the fiction immediately. |
| Any oracle system | Ask: "Does the worst complication arrive right now?" Roll at maximum difficulty, then ask what it is. |
Worked Tilt — Kael at the Derelict
What was true at the start of Act 1: Kael had a clean salvage claim, the element of surprise, and a clear extraction route.
The Tilt question at the 40-minute mark: What is the worst complication that could arrive right now?
The roll: A second Syndicate ship drops out of jump on a parallel vector — not a patrol, a cleanup ship. They aren't here for the wreck. They're here for the enforcer's mistakes.
What that fractures:
- Her extraction route is now blocked.
- The enforcer is no longer her primary obstacle. He's a peer in a worse problem.
- The logs in her pocket may be evidence the cleanup ship was sent to destroy.
Act 2's opening scene builds from this fact, not from speculation about it.
The Variety Check — Run This Before the Tilt
The Five Beats enforce structural variety. Before you apply the Tilt, run a quick check:
- Have at least three different scene types shown up so far? (Combat, social, exploration, intrigue, etc.)
- Has the Challenge beat actually been non-combat? If you defaulted to a fight when the oracle pointed at a puzzle, you've collapsed two beats into one.
A session where every scene is a fight isn't a story. It's a fight scene with intermissions. The 5RD discipline forces texture into the session even when your oracle keeps suggesting the same thing. If your variety check fails, the Tilt is the moment to course-correct: make the worst complication something non-combat — a betrayal, a revelation, a moral fracture.
VI. Act 2 — Fallout
Four scenes, ~50 minutes. React to the Tilt. Frame each scene using the Protocol, but now the want from Scene Zero is under pressure in a new way.
Don't try to immediately fix what the Tilt broke. Let consequences compound for at least two scenes before the story starts resolving. The want may change shape in Act 2 — that's not failure, that's story. Senna may find that getting the witness to safety matters more than the original vow. Kael may find that the logs contain something she didn't expect. Let the Tilt reframe what winning looks like.
Required beat in Act 2:
- Scene 8 (or thereabouts) — Climax. The hardest scene of the session. Direct confrontation with what the Tilt revealed. This is the scene the story has been pointing at since the Threshold.
Scenes 5–7 are connective. One or two of them can be interludes — the breath between the Tilt and the Climax.
Act 2 Signal: When the consequences of the Tilt have fully landed and you can see what the story is actually about, you're ready for the Climax.
Interludes in Act 2
An interlude is a scene without a frame. No Want, no Obstacle, no Stake — just a character, a moment, and what's on their mind. (Full structure in Scene Framing.)
Interludes belong in Act 2 specifically. They don't belong in Act 1 (they stall the pursuit) and they don't belong in Act 3 (they dilute the ending). In Act 2, they do real structural work — they let the Tilt's consequences settle emotionally before the Climax.
Two interludes per session, max. If you find yourself wanting a third, the issue isn't that you need another interlude — it's that your last framed scene didn't generate enough consequence. Re-read your last outcome, find the "but" or "therefore" you missed, and frame from there.
Example — Kael, between Tilt and Climax:
Kael, alone in the derelict's empty bridge, with the captain's logs already in her pocket and the cleanup ship's transponder pinging on her HUD.
She plays the captain's voice for the first time. The recording isn't what she expected. One paragraph. Then cut.
VII. Act 3 — Resolution
One or two scenes, ~20 minutes. The Climax has happened. Now the story ends.
Ask one final oracle question:
Does [character] get what they wanted?
Roll at whatever difficulty your current Chaos Factor, track, or oracle suggests. Take the answer and write three sentences:
- What happened. The final scene's outcome in plain terms.
- What it cost. What is different about the world or the character that wasn't true at the start.
- Who they became. One sentence. Not a moral — just a true thing about this person now.
Then stop. A complete story lives in your journal.
A note on failure. A "no" or partial-fail at this final roll isn't a broken story. It's a tragedy, a cost, or a Pyrrhic victory — all of them legitimate endings. Some of the best solo sessions end with the character not getting what they wanted. Write the three sentences anyway. The story still lands.
Worked Resolutions
Kael got the logs. They named her captain's killer — but also named Kael as the original source of the coordinates that led the Syndicate to the wreck. She has the evidence. She doesn't know yet what to do with it.
Senna got the witness to the garrison. The riders arrived an hour after dawn, too late. The bridge is still out, and whatever the warlord is building continues — but now there is testimony, and testimony has weight.
Vesper freed the prisoner but lost three of the resistance to the warden's alarm. The governor's witness is on the road north. Vesper is the one who put him there, and the one whose name the governor now knows.
A failure-outcome example, for contrast:
Bram never got the name. The woman with the warlord's mark slipped away in the night and took the answer with her. The village stays unavenged. Bram learned that he is the kind of man who, when finally face to face with what he wanted, hesitated.
The next session can continue this character and world, or begin fresh.
VIII. Worked Session — Kael's Ten Scenes
Two lines per scene. The full shape in motion.
| # | Phase | Beat | Scene |
|---|---|---|---|
| 1 | Act 1 | Threshold | Kael cuts through the airlock seal. The Syndicate crew is already inside; her clean salvage just became a heist. |
| 2 | Act 1 | (development) | She finds the enforcer's ID badge in a crew quarters. He's not a stranger — he's the one who denied her claim. |
| 3 | Act 1 | Challenge | The ship's AI is awake and watching. She has to convince it she's the captain's heir, not a thief. |
| 4 | Tilt | Tilt | A second Syndicate ship drops out of jump. Cleanup. Her route is gone. The logs are evidence someone sent them to erase. |
| 5 | Act 2 | (fallout) | The enforcer reaches her first. He's terrified. He proposes they work together to survive. She has thirty seconds to decide. |
| 6 | Act 2 | (interlude) | Alone on the bridge with the logs in her pocket. She plays the captain's voice. The recording isn't what she expected. |
| 7 | Act 2 | (fallout) | The cleanup crew breaches the hull. Kael and the enforcer are pinned in opposite corridors with one extraction route between them. |
| 8 | Act 2 | Climax | She has to choose: leave with the logs and let the enforcer die, or share the route and risk losing everything. |
| 9 | Act 3 | (descent) | They make it out. The cleanup ship is hunting their drift. Kael hides the logs in the enforcer's ship before she leaves. |
| 10 | Act 3 | Resolution | She has the evidence. She doesn't know yet what to do with it. The captain's voice is in her ears, and the man who killed him has a name. |
IX. Scaling
The shape is fractal. Same structure works at scene, session, and arc levels.
A 90-Minute Session — 7 Scenes
For a tight session: cut one connective scene from Act 1 and one from Act 2.
| Phase | Scenes | Beats |
|---|---|---|
| Act 1 | 2 | Threshold + Challenge |
| Tilt | 1 | Tilt |
| Act 2 | 3 | One connective, one interlude (optional), Climax |
| Act 3 | 1 | Resolution |
A 2.5-Hour Session — 12 Scenes
For a longer evening: add one connective scene to Act 1 and one to Act 2.
| Phase | Scenes | Notes |
|---|---|---|
| Act 1 | 4 | Threshold, two development, Challenge |
| Tilt | 1 | |
| Act 2 | 5 | Three connective (1–2 interludes), Climax |
| Act 3 | 2 | Final scene, then Resolution |
A Two-Session Arc — Cliffhanger at the Tilt
Session 1 runs Act 1 and ends at the Tilt — the worst complication arrives, you write it down, and you stop. The Tilt becomes the cliffhanger. Session 2 opens with Act 2's first scene and runs through Climax and Resolution.
This is the single best use of structural Tilt for solo play. The mandatory hinge becomes a built-in cliffhanger that pulls you back to the next session.
A Five-Session Campaign — One Beat Per Session
For a longer arc, each session is one beat:
| Session | Beat | What it does |
|---|---|---|
| 1 | Threshold | Introduce the world. Character commits. World shows first cost. |
| 2 | Challenge | A non-combat layer. Politics, mystery, moral test. The world deepens. |
| 3 | Tilt | The reversal. The session everyone remembers. |
| 4 | Climax | The confrontation. The hardest single session in the arc. |
| 5 | Resolution | Aftermath. What it cost. Who they became. |
Each individual session still uses the full five-beat structure internally, with its own Tilt at the midpoint. Stories nested inside stories — every level has a clear shape and a forced hinge.
This solves the long-form problem. People can write multi-session stories but rarely finish them, because long arcs lack mandatory breaks and drift into open exploration. A forced Tilt at every layer (scene, session, arc) gives even abandoned campaigns a sense of completion at every level.
Quick Card
| Scene Zero | 5 min. Character · drive · first "but" · NPC · stake · obstacle → frame → Go. (See Scene Zero.) |
| Act 1 — Pursuit | 3 scenes / ~30 min. Beats: Threshold → development → Challenge. Things get harder. |
| The Tilt | Mandatory at 40 min or scene 3. Worst complication arrives as fact. Don't soften. |
| Act 2 — Fallout | 4 scenes / ~50 min. Connective + 1–2 interludes → Climax. Want may change shape. |
| Act 3 — Resolution | 1–2 scenes / ~20 min. One oracle roll. Three sentences: what happened · what it cost · who they became. |
The Five Beats
Threshold · Challenge · Tilt · Climax · Resolution.
Five required scenes. Everything else is connective.
The Pacing Rule
Two interludes per session, max. All in Act 2. None in Act 1 or Act 3.
Companion Documents
- Scene Framing — The 60-second protocol for any scene. Run this for every scene in every act.
- Scene Zero — Five minutes from blank page to first frame. The on-ramp into the Session Shape.
Credits & Sources
- The Tilt as a mandatory structural break — Jason Morningstar, Fiasco (Bully Pulpit Games, 2009)
- The Five Room Dungeon — Johnn Four, Roleplaying Tips (2006). Adapted as Threshold · Challenge · Tilt · Climax · Resolution.
- Three-Act Structure — Syd Field, Screenplay; Robert McKee, Story
- Scene as goal · conflict · disaster — Dwight V. Swain, Techniques of the Selling Writer (1965)
- "Prep is play" — Shawn Tomkin, Ironsworn (Tomkin Press, 2018)
v11 — System-agnostic. Compatible with Ironsworn, Starforged, Mythic GME, Kal-Arath, Daggerheart, Traveller, PUM, SUM, and any solo or GM-less system.