Build a Relationship Diagram That Explains a Film’s Conflict
Build a Relationship Diagram That Explains a Film’s Conflict
Ana’s work is inside the workshop, and she needs it before the meeting starts. Bo has the key, but he isn’t leaving until someone covers his shift. Cleo would cover it, except Cleo is waiting on Dani to bring back a borrowed tool. Dani has that tool. It’s damaged. She hasn’t told Eli, who owns it — though Eli already knows.
Five people, one errand, and four conditions standing between Ana and a door. Written out as a paragraph, the scene survives. Put it on a slide as five headshots with lines between them and it can fall apart: the audience sees a crowd, not a problem.
The scene is invented, and so are the five people. I’m keeping them fixed, because the whole point of the diagrams below is that you can check every line against the situation they describe. That’s the test a diagram has to pass: not whether it looks organized, but whether it explains why someone can’t simply act.
So: draw the relationships needed to understand one conflict, not everyone who appears in the story. Label the connections as active relationships — owes, depends on, is waiting on, must persuade. Use direction where power or knowledge is unequal, and split the picture in two when the story moves through time. Keep prose wherever a hard line would claim more certainty than the scene actually provides.
Start with the scene you can’t follow
Before anything gets drawn, name the confusion. Not “who is in this film?” — “why can’t Ana get into the workshop?” That single change of question eliminates most of the cast.
Selection follows from the question. Anyone whose connection changes the answer stays. Anyone whose connection doesn’t, goes. In the workshop scene, the meeting Ana is rushing toward doesn’t need a node. Neither does whoever else is waiting inside. If a person can be removed and your explanation of the scene is still complete, they were never load-bearing.
This is where the cast chart fails, and it fails honestly. A cast chart answers a real question — who is in this film? — which is why it keeps getting used. Here’s the same five people as a cast chart:
Ana Bo Cleo Dani Eli
└──────┴─────┴───────┴──────┘
all five work in the same workshop
Every line on that chart means the same thing. It tells you they belong to the same story, and it will still be true in the last act no matter what anyone does. There is no way to read it aloud as an explanation. You cannot point at an edge and say “this is why Ana is stuck.”
A genealogy has the same property, which is why the family tree is the most common wrong diagram in a film deck. It’s accurate, it’s complete, and it tells you the one thing you probably already knew: these people are related. A complete social network answers a different question than the one the scene raises, and it can actively make the conflict harder to find, because now the audience has eleven connections to hold and no reason to prefer any of them.
Replace roles with sentences
Take the cast chart apart and rebuild it from obligations. A role is a category. An obligation has a direction and a cost.
“Bo and Cleo are colleagues” is a category. “Bo needs Cleo to cover his shift” is a condition: it can be met or unmet, it has a subject and an object, and its failure has a consequence for Ana. Both statements can be true at once — that’s fine — but only the second one does any explanatory work.
Here is the workshop scene as a dependency diagram, with each edge written as a sentence:
Ana
│ needs the key from
▼
Bo
│ needs the shift covered by
▼
Cleo
│ needs the tool returned by
▼
Dani
│ is holding the tool that belongs to
▼
Eli
Eli ── knows the tool is damaged
Dani ── believes Eli does not know
Read the chain aloud and it comes out as four ordinary sentences. That’s the standard to hold yourself to. If an edge can’t be spoken as a sentence with a subject and a verb, the audience has to decode it in silence, which spends the attention you wanted for the conflict.
Notice what got dropped. Nobody is labeled “workshop member,” because Ana’s problem isn’t a lack of membership. What stayed is direction and specificity: Ana needs the key from Bo, not Bo from Ana; Cleo needs the tool Dani is holding. Reverse either arrow by accident and the diagram would still look tidy while saying something untrue.
Two more habits are worth forming here.
First, put social facts in the caption, not in a second kind of line. Bo and Cleo work the same shift, which is genuinely why Bo asks Cleo and not someone else. That’s useful. But if you express it as a dotted line, the key as a double line, and the shift swap as a thick one, the audience is now reading a legend instead of a conflict. Write “works the same shift with Cleo” beside Bo’s name and keep the arrows doing one job.
Second, keep prose for anything the scene leaves open. Some relationships in a film are genuinely uncertain at the point you’re explaining — someone’s loyalty, whether a promise was meant, what a character suspects. A line on a diagram asserts. Where the story hasn’t decided, don’t let your drawing decide for it. A short sentence underneath the diagram costs you two lines of slide space and prevents you from making a claim the film never makes.
Point the arrows, and split the picture in two
Most interesting relationships are unequal. Ana needs Bo’s key; Bo does not need Ana’s anything. Cleo needs the tool Dani is holding; Dani needs nothing from Cleo. If you draw these as undirected lines, you flatten the exact imbalance that creates the scene, and the audience can no longer tell who is waiting on whom.
Arrows carry that. But arrows do two different kinds of work, and it’s worth separating them.
The first kind is action and obligation: who needs what from whom, who must be persuaded, who can refuse. Ana’s chain is all of this.
The second kind is knowledge. In the workshop scene, knowledge runs against the obligations. Dani is hiding the damage, but Eli already knows. That means Dani is operating on a false belief — she thinks she has a secret, and she doesn’t. This isn’t a relationship between Dani and Eli in the ordinary sense. It’s a fact about Dani’s head. Draw it as a statement (“Dani believes Eli doesn’t know”) or leave it in the prose. What you must not do is promote it to an arrow: “Dani conceals the damage from Eli” would put a line on the picture that the scene contradicts, since Eli already knows. Nor should you draw a line from Eli to Dani labeled “doesn’t know” — that line is false for the same reason.
Once knowledge and obligation live in the same picture, you’ve probably got two moments in the story, not one. That’s the case here. When Eli tells Dani the damage is already known, her false belief goes — and that is all that changes:
Ana
│ needs the key from (unmet)
▼
Bo
│ needs the shift covered by (unmet)
▼
Cleo
│ needs the tool returned by (unmet — the tool is still out)
▼
Dani
│ is holding the tool that belongs to
▼
Eli
Dani ── believes Eli does not know ← this line is gone
Both Eli and Dani now know the tool is damaged.
Look at how little moved. The line about what Dani believes is gone. The knowledge is now shared. Everything else is exactly where it was: the tool has not been returned, Cleo has not accepted the swap, Bo has no coverage, and Ana still has no key.
That’s the most common diagram error in a before-and-after slide — and it’s a factual error, not a design one. Someone draws the “after” state with Dani’s problem solved, sees a cleaner picture, and slides it into the same frame as the resolved ending. But removing an obstacle is not the same as completing an action. Eli forgiving the damage does not carry the tool across the workshop. Understandable hesitation is not a returned object.
The rule that prevents this: never stack an old unresolved state and its replacement in one image. If the story changes a dependency, or a belief a character is acting on, draw two pictures. If it doesn’t, draw one — but then make sure every line in it is currently true.
Put the diagram back against the scene
The diagram earns its place when it lets someone explain the next scene. So test it that way, with the actual scene, not with an opinion.
Hand the picture to someone and ask why Ana can’t just go in. A weak diagram produces “because Bo won’t give her the key.” The chain above should produce something longer: Bo can’t leave until Cleo covers the shift, Cleo is holding out for the tool, the tool is stuck with Dani, and Dani is stalling because of the damage she thinks is still hidden. That answer names four separate places the scene could turn, which is exactly what a writer or director needs in a room.
Then run the second state. After Eli speaks, the same reader should be able to say what changed and what didn’t: the belief is gone, and nothing structural with it. If the diagram produced a falsely smooth answer there, it wasn’t explaining the story — it was explaining the story you’d rather have.
Compare against the cast chart at this point, using the same five people and the same scene. A neater network is not automatically the clearer one. The chart is tidier. It also can’t answer the question.
Two honest limits on this method. The diagram does not replace characterization — it shows you that Dani is stalling, not whether she’s ashamed, or stalling for reasons of her own. Once you know the shape of the dependencies, the scene still has to supply motive, tone, hesitation, and all the things that make a person a person rather than a node. Keep a paragraph of prose next to the picture, and keep it in the prose’s voice.
And the comparison in this article has not been put in front of readers. The retelling test I just described is the test I’d want run — give someone the diagram, read the following scene aloud, see whether they can explain why Ana is stuck and what would have to change. That check hasn’t happened here. What you have is an invented scene, two diagrams, and an argument that they line up. The argument is checkable: every arrow above can be read against the workshop paragraph at the top, and none of them should be doing work that paragraph doesn’t support.
Start with the confusion, label the connections as sentences, point the arrows, split the timeline when the story turns, and keep the sentences the film hasn’t earned yet. The diagram is finished when someone can read it out loud and then use it on the next scene — including the scene where the misunderstanding clears up and nobody has moved.
Frequently asked questions
What question should a relationship diagram answer first?
It should start with the scene you cannot follow—why a character cannot act—not who is in the film. In the workshop example, the question is why Ana cannot get into the workshop. Anyone whose connection changes that answer stays; anyone who can be removed without changing the explanation goes.
Why does a cast chart fail to explain conflict?
Every line on a cast chart means the same thing: these people belong to the same story. That can remain true regardless of what anyone does, and no edge can be read aloud as 'this is why Ana is stuck.' A genealogy has the same problem: it is accurate and complete but only says the people are related.
How should connections be written on a dependency diagram?
As active relationships with direction and cost: owes, depends on, is waiting on, must persuade. Each edge should be speakable as a sentence with a subject and verb, such as 'Bo needs Cleo to cover his shift.' If it cannot be spoken, the audience has to decode it silently.
What is the common before-and-after slide error?
Stacking an old unresolved state and its replacement in one image. Removing an obstacle is not completing an action: Eli forgiving the damage does not return the tool, cover the shift, or get Ana the key. If a dependency or a belief changes, draw two pictures; if not, draw one where every line is currently true.
What are the limits of the diagram method?
It does not replace characterization; it can show Dani is stalling, not whether she is ashamed or why she stalls. The retelling test described—giving someone the diagram and reading the next scene aloud—has not been run with readers. The comparison is an invented scene and two diagrams, and the argument is checkable against the workshop paragraph, not a tested reader response.