Pitch a Science Series Through What a Demonstration Lets Us Observe

A pitch page reads:

Episode 4 opens tight on a table lamp. A coin goes down on the table. Then a cereal box. On the wall behind them, the coin's shadow is wider than the box's. "Nothing about size," the narrator says, "is what it seems."

Three things are happening in five lines, and only one of them is a shot. The shot is fine. The claim at the end — that size is not what it seems — has walked from an arrangement on a table to a statement about how people see, and nothing in the sequence carries it there. An editor skimming twenty pitches may not catch it. A physicist watching the rough cut will.

The fix is not to soften the language until it means nothing. It is to write the pitch so that the observation, the conditions and the explanation appear in the right order, and to state plainly what the sequence cannot settle. That is most of the difference between a demonstration that carries an argument and an impressive image with a voiceover stapled to it.

The example below is deliberately small and harmless: a lamp, an object, a wall. Shadows are cheap, safe, and hard to fake. Change the physics to whatever your episode is about; the reasoning is the part you keep.

Name the question before choosing the spectacle

Where most pitches put "an amazing experiment," put the uncertainty.

Two very different things get called a question, and pitches blur them constantly. The first is the viewer's question — what the audience walks in already wondering. "Why is my shadow so long at the end of the day?" That is a hook. It earns the first thirty seconds, and it does not need to be unresolved in science. A question the viewer can answer along with you is often the friendliest opening an episode has.

The second is a research question: something nobody in the room has the answer to. If your episode is built on one of those, the pitch has to say who would answer it, by what means, and how long that takes, because a shoot day is a terrible place to discover the answer is eighteen months away.

There is nothing wrong with the first category. It becomes a problem only when the pitch dresses it up as the second. "Scientists are only beginning to understand why a shadow grows" is a sentence you should be prepared to defend with a name and a published result. "This is why a shadow grows, and here is a way to see it happen" is a sentence you can defend with a lamp.

The working question here is the first kind, and it has the useful property of a plausible wrong answer:

Does a bigger shadow necessarily mean a bigger object?

The assumption the sequence is designed to test is that it does. It does not, not without knowing where the light is and where the surface is. That gap between the assumption and the mechanism is what an episode can build on, and it is a stronger engine than "shadows are amazing," because it can be phrased as something a viewer might get wrong.

One line of the pitch should carry the honesty about which kind of question this is. Something like: Shadow geometry is standard; the optics reviewer confirms the framing and the wording of the claim. Whether a question is settled is a specialist's judgment, not a writer's. Name the expertise you need even if you have not cast the person yet.

Decide whether the sequence is a demonstration, a model or an investigation — and where an analogy fits

Four words get used interchangeably in treatments, and they promise different things.

A demonstration shows a phenomenon or an accepted explanation under conditions you chose. Both halves matter. It is real, and you selected the conditions — which is what makes it televisual and also what limits what it settles.

A model represents some properties of a thing and drops others on purpose. A diagram is a model. So is an animation, a scale build, a simulation.

An analogy borrows a relationship from a different situation. It carries one relationship, not the whole apparatus.

An investigation asks a question whose answer arrives by examination. You pitch it without knowing the ending.

Mode by mode, that looks like this:

Apply that to the shadow question and you get three genuinely different sequences.

The filmed demonstration shows an arrangement a reviewer has chosen: a small object placed near the light source, a much larger object placed close to the wall, both of their shadows landing where the camera can see them. It can establish that in this arrangement, the smaller object threw the broader shadow. That is it.

The graphic is a top-down diagram of the same arrangement: source, two objects, wall, and lines showing how light from the source spreads and gets interrupted. It can make the mechanism legible in a way no camera can, because a camera cannot photograph lines of light. It is a model, not a recording of invisible events, and the pitch should say so. Someone chose how many lines to draw and where to put them; the diagram probably assumes a single bright point and clean straight edges. Real lamps have a glowing surface with width, and that width is what makes the edge of a shadow a gradient rather than a line. A good pitch lists those simplifications in a line, the way you would label a map.

The investigation is narrower and more interesting than either. Suppose you want to know how much the soft edge matters — how wide the fully dark part of the shadow stays as the object moves toward the source. That question has no answer until someone records positions and measures widths against a stated brightness threshold. You cannot cut to it after a nice take and reveal a number. Either you plan the measurement or you leave the question out.

Then there is the analogy, which is a device rather than a mode. A natural one here is a slide projector: put the slide farther from the screen and the picture grows. The shared relationship is real and worth stating — distance scales the result. But the projector also stops being useful at a specific point, and the pitch should say where: a lens bends light back together, so it can focus, shrink, and invert an image. A blocked light source does none of that. It only spreads. An analogy that names one shared relationship and one place it stops is doing its job. An analogy left open is a trap you set for your narrator.

Connect what appears on screen to the claim

A sequence that reasons has four parts, and they go in this order: what was observed, under what conditions, why that is the proposed explanation, and what further evidence a stronger conclusion would need.

The observation is what a viewer can see and describe without knowing why. The coin's shadow is wider than the box's. Note the camera's role here. A straight-on shot of the wall shows a result and hides the variable that explains it; a side view, or a diagram with the distances on it, lets the audience see the thing doing the work. If the audience cannot see where the lamp is, they cannot evaluate anything you tell them.

The conditions are what was held fixed, what was varied, and how you handled the awkward parts. Shadow width is not a crisp number when the lamp has a glowing surface an inch wide. The edge is a gradient, because different parts of that glowing surface are blocked at different points. So the pitch needs to say what "width" means: the fully dark region, or the outer edge where the shadow disappears, and who decided.

The explanation is the mechanism, stated without borrowing authority from the shot. Light leaves every part of the source and travels in straight lines. An object between the source and the wall removes a cone of that light, and the cone widens with distance. An object near the source blocks a wider cone than a larger object near the wall. That is the whole engine of the surprise, and it contains no claim about what viewers think.

What a stronger conclusion would need is the part most pitches omit, and it is the part that protects the series. To claim that shadow width varies with distance in a particular way, you need recorded positions and measured widths — a set of them, not one. To claim that people misread object size from shadows, you need a study with participants. That is evidence about people, and it is a different production requiring a different design. It is not something a narrator can assert between two shots.

Watch the leap from this happened here to this explains every instance. A filmed arrangement can show that shadow width depends on placement, which is enough to make the point that width alone does not reveal an object's size. It cannot show that people use shadows as a size cue in daily life, and it does not cover every lighting situation — a sun, an overhead panel, a room with light bouncing off four walls. Each of those is a condition you did not test.

Write uncertainty into the proposed episode

Three words keep a pitch honest, and they should appear, in the document, where you can see them: proposed, observed, established. A sequence is proposed until it is filmed. A result is observed once it is on record. A mechanism is established when a specialist agrees it is standard. Mixing the three is how a treatment written in a development office becomes a claim in a voiceover.

The uncertainty goes in as specific items, not as a mood. Which sequence is proposed. Which explanation needs review, and by what kind of reviewer. What remains unknown and is worth saying out loud on screen because the audience will enjoy knowing it.

Then write the branch for when the shoot does not deliver. Suppose the coin's shadow does not come out dramatically wider, or it does but the camera flattens the distances so it looks rigged. You have three honest routes and one dishonest one. You can investigate the discrepancy: perhaps the lamp was wider than the diagram assumed, or the wall was closer than you thought, and adjusting and recording the change is genuinely better television than the planned version. You can revise the explanation if the reviewer's geometry suggests a different emphasis. Or you can narrow the segment — film the part that works and drop the narration that overreached. The dishonest route is a second take that hides the distance between object and lamp. The tell is that the conditions vanish from the final cut. A "surprising" result is not surprising to the people who know the setup, and losing them is expensive for a series that wants them.

One boundary belongs in every pitch of this kind. Anything involving heat, pressure, high voltage, lasers, fast machinery, or chemicals is not something a treatment writer designs. Its place in a pitch is the question it would answer; the design belongs to people qualified to design it, with their own approvals, their own risk assessment, and their own schedule. This article stays inside the part the writer owns: the question, the observation, the connection between them, and the limit.

Keep the subject review separate from production approval. A science reviewer says whether the claim is warranted. The production side says whether it shoots, costs, and insures. When those two conversations merge, "it films beautifully" quietly starts answering "is it true."

An annotated sequence

Here is the whole thing as a pitch block, with each line doing one job.

Question. Does a bigger shadow necessarily mean a bigger object?

Status. Proposed sequence. Not filmed. No measurement has been taken.

Proposed observation. In an arrangement to be chosen with the optics reviewer and filmed in one unbroken side-on view, a small object near the light source throws a broader shadow on the wall than a much larger object set close to the wall — the reverse of what the two objects' sizes would lead you to expect.

Observed so far. Nothing. This is a description of what the sequence would show.

Explanatory connection. Light leaves every part of the source and travels in straight lines. An object between the source and the wall removes a cone of that light, and the cone widens with distance, so placement matters as much as size. The reviewer confirms this framing before the narration is drafted.

Model on screen. A side-on diagram of source, objects and wall, not footage, with its simplifications listed: single bright point, straight edges, no room light. The diagram carries the geometry; the film carries the instance.

Analogy, and where it stops. A slide projector enlarges a small slide as the distance grows. It shares one relationship — distance scales the result — and it stops at the lens, which can focus, shrink and invert what a shadow cannot.

Limit. The sequence can show that shadow width depends on placement, so width alone does not reveal an object's size. It cannot show that viewers misjudge size from shadows. That requires a study with participants, and this episode does not have one.

Required review. Specialist assessment of the question, the arrangement, the diagram and the inference, separate from production approval.

Two notes on putting this to work. The definitions above are working vocabulary, and different teams draw the lines slightly differently — it is worth checking them against the way the receiving department and its science advisers talk before you commit to them in a document that will be read as a promise. And a maker's account of a previous shoot is usually more useful than any style guide for telling you which of these distinctions survives contact with a studio floor and a limited day.

The pitch that makes it through review is the one whose last line is narrower than its first image, and says so on purpose. That is not a smaller pitch. It is one that can be shot, defended, and watched by the people who know the answer without losing their respect.

Frequently asked questions

Why can an impressive demonstration fail to carry its pitch?

A striking image can be fine while the narration leaps from an arrangement on a table to a claim about how people see. The sequence needs the observation, the conditions, the proposed explanation, and a statement of what further evidence a stronger conclusion would require.

What is the difference between a viewer's question and a research question?

A viewer's question is what the audience walks in wondering; it can hook the opening even if science has settled it. A research question is unresolved and needs a named person, method, and timetable. A pitch should not dress the first as the second by implying scientists are only beginning to understand something ordinary.

How do a demonstration, a model, an analogy, and an investigation promise different things?

A demonstration shows a phenomenon under conditions you chose, so it carries 'in this arrangement.' A model represents selected properties and drops others on purpose. An analogy borrows one relationship from another situation. An investigation asks a question whose answer arrives by examination. Each strains when it claims more than that.

What four parts should a sequence that reasons contain?

Observation: what a viewer can see and describe without knowing why. Conditions: what was held fixed, what varied, and how awkward parts such as a shadow's soft edge are defined. Explanation: the mechanism, such as light travelling in straight lines and a blocked cone widening with distance. Limit: what stronger evidence would be needed, such as recorded measurements or a participant study.

What safety and review boundaries belong in a science pitch?

Anything involving heat, pressure, high voltage, lasers, fast machinery, or chemicals is not designed by a treatment writer; the pitch can name the question, but qualified people own the design, approvals, risk assessment, and schedule. Science review—whether a claim is warranted—should also stay separate from production approval, so that 'it films beautifully' does not answer 'is it true.'

Further reading

• Pitch a Factual Explainer Series Around Questions, Not a List of Subjects — After you have a viewer question, this shows how to build an episode spine of evidence and revised understanding instead of a list of subjects.
• Keep Searching—or Make a Small Visual Experiment? — When the open question needs a measurement rather than a better shot, this shows how to scope the smallest test without letting a stand-in answer for different material.