Your Best-Written TV Script Is Not Your Most Representative. Which Should You Send?

The script you're proudest of is usually the one that looks least like the show. That isn't a coincidence. You wrote the other one the way you write most weeks — serviceable, on-model, moving the crew through the work — and you wrote this one because you wanted to prove something, which meant stripping away the ordinary machinery and seeing whether anything held up without it. It held up. Now the request asks for one script, and the one that shows your best writing is the one that least resembles what the recipient is buying.

The reason this decision feels like taste is that "best" is being treated as a property of the script. In a submission it isn't. It's a property of the match between the script and the thing actually being evaluated, and the thing being evaluated is named in the request you were sent.

So, the short answer. Read the request for its object and its purpose. If the purpose is writing ability, the exceptional script is generally the stronger sample, provided it's eligible and provided you say what it is in one factual sentence. If the purpose is the ordinary operation of the series, the representative script is more informative, and the cost is that your best work stays in the drawer. If the request names a pilot and neither script is a pilot, both candidates leave the table and the dilemma you were about to spend a week on doesn't exist.

The rest of this is how to get from that sketch to a choice you can defend out loud.

The requested object comes first

Before there is a comparison, there is an object. A completed episode script, a short episode example, a developed synopsis and an original pilot are four different things. Writing one of them beautifully does not make it the answer to a request for another. If a recipient asked for a short example of your writing on their show, a completed script isn't a more generous version of the same answer; it's a different answer to a different question.

The same goes for restrictions, which are sometimes explicit and sometimes just implied by the sentence structure. "Any completed script you've written" is open. "One complete episode of Harbor Works" names a series. "Your original pilot" names a form. "A pilot outline at first stage, a full pilot script for applications for additional funding" makes the object depend on where you sit in a process — one bounded example of that last pattern is Screen Ireland's Television Drama Development route, where the page as checked on 9 September 2026 listed a pilot outline for the first stage and a completed pilot script for later funding applications. That's one Irish funding scheme, not a description of what recipients generally want, and it's a dated observation rather than something re-verified while writing this. Treat it as evidence that the requested object can change with the stage, then go back and read the page before you rely on it.

Two consequences follow. First, an ineligible favourite cannot win this comparison, no matter how good it is. Second, if you haven't yet settled whether you're being asked for an original pilot or for a sample episode of an existing series, that question has to be answered before this one means anything, because it changes which scripts are even on the table.

What a script is evidence of

Once the object is settled, quality comparison becomes possible. The useful move here is to notice that a script is evidence in two directions — about the writer and about the series — and that those two kinds of evidence come apart easily.

Start with the writer. Ask what skills this particular episode actually puts on the page. A confined two-hander will show a reader how you build a turn, hold subtext, keep a scene escalating when there's nowhere to cut. It will not show how you construct a guest character whose problem is public enough to occupy an hour, how you handle a procedural sequence with actual procedure in it, or how you intercut four strands in a week where nobody's problem gets resolved. If the reader needs those things, the confined script cannot supply them however excellent it is. Nor can a note supply them. You can describe a skill; you can't put it on the page from the cover of an email.

Then the series. A reader who doesn't know the show builds a mental picture of it from your sample, and the properties that determine that picture are fairly consistent across requests. The recurring relationships and the configurations they normally appear in. The locations, including whatever standing set the show lives in. How much the guest carries. The structural device the show runs every week — the job, the case, the appointment, the shift. And how much prior context the script assumes the audience already has.

Be careful about turning this into a score. A grid with "craft: 9, representativeness: 4" looks rigorous and answers nothing, because the weights come from the request, not from the scripts. The request usually names one purpose, not a formula, and you're better off comparing both scripts against that single purpose than against an invented rubric.

There's also a category error to avoid: quality and representativeness are different kinds of evidence. An excellent exception tells a reader about your ceiling. A solid ordinary episode tells them about your baseline. Neither implies the other, and if what's wanted is the baseline, no amount of ceiling closes the gap.

One variant is worth naming, because it changes the shape of the trade-off. If your two candidates aren't episodes of the same series — your strongest script is an original pilot or an episode of something else, and your representative one is an episode of the show in question — then the exception isn't misrepresenting anything, because it was never an episode of that show. What it can't do is demonstrate that you can run this show's ordinary week. Eligibility is a separate question, and a prior one.

Is the departure the show's own?

Before you decide that your best script is atypical, check that against the series rather than against your impression of it. You're the wrong person to guess, and also the right person, because you've read every script. Does the show do a confined episode every season? Does it have a formal move it returns to — a two-hander, a night shift, a single-location hour, an episode that leaves the main set entirely? If a bottle episode is part of the show's regular repertoire, then your two-hander sits inside the show's own pattern, and "unrepresentative" gets more complicated. A script written in one of the series' established modes represents the series better than a script that's formally ordinary but wrong in some other way.

That other way matters too. Structural ordinariness is not the same as representativeness. A guest-centred episode built around someone who would never recur can misrepresent a show whose guests are usually local and repeat. A finale-shaped script that assumes twenty episodes of setup is as unlike the normal week as a bottle episode, and harder to notice, because it has all the usual locations in it.

So the question isn't "does this depart?" but "does this depart from something the show doesn't already do?" — and if the answer is that the show does it, the case for the exceptional script gets stronger, not weaker. This is also where you should be honest about cause. The scripts that feel like showcases feel that way because you wrote them to be showcases. That's not a flaw. It's the mechanism. The question is only whether the result is still recognisably the show.

Two scripts, two requests

The series and scripts below are invented for this comparison, and their quality is stipulated for the exercise rather than judged from a script anyone read.

Harbor Works is an existing one-hour drama about a municipal repair crew. One writer has two unproduced spec episodes of it, both completed: "Night Line," at 58 pages, and "Seventh Street," at 62.

To apply the test from the previous section: Harbor Works doesn't run the confined two-hander as a mode it returns to. Its ordinary week is several jobs across several locations, with a guest carrying one of them. So "Night Line" departs from the show's own repertoire rather than sitting inside it.

Now two requests. The first: send one complete script as a writing sample; any episode you've written is fine. The second: send one complete episode of Harbor Works; we want to see how you'd handle the show as it normally runs.

For the first, the object is writing ability, and the only thing a reader can check is what's on the page. "Night Line" is the stronger evidence for that. In one location with two people and no device to hide behind, the dialogue, the turn and the subtext are all visible and all attributable to you. The rationale you're applying is simple: this request is buying craft, craft is what this script shows, and the alternative shows less of it.

The context sentence, if one is allowed, does the work the script can't: "'Night Line' is a two-hander in one location over one night; it assumes season-two history the reader won't have, and the show normally runs several locations and a guest story." That's a description, not a defence. It doesn't apologise for the departure, and it doesn't pretend the episode is typical. It gives the reader an accurate frame so they don't conclude the writer thinks every week looks like this. Without it, you'd be letting a reader form a false picture of the series you're claiming to be able to write. The assumed history is part of that same frame: it limits what the episode can say about the series, which this request isn't asking about, and it leaves the scene work the request is buying where the reader can see it.

For the second request, the reader has said what they want to see, and it's the baseline. "Seventh Street" is the answer. It shows the engine — three jobs, four locations, six regulars colliding, a guest with a public stake, and the school crossing that never gets finished. The loss is real and specific: shorter dialogue scenes, no sustained two-hander, so the script understates your best writing. You're choosing to lose the demonstration of craft in order to give the demonstration of operation, because operation is what was asked for.

Notice that the context sentence here is optional, and that the discipline is different. If the request permits a note, one neutral line of orientation is plenty — "This one runs the show's normal pattern rather than a departure" — and its job is description, not advocacy. If no note is permitted, nothing in the package is dishonest, because the representative script needs no explanation. The asymmetry is the point: a context sentence is required when your sample is the exception, and unnecessary when it's the baseline. Mentioning the other script in a cover note isn't a way around this. "My stronger writing is in a different episode" is a claim with nothing behind it in the envelope, and a reader who wanted to see that script will ask for it or won't.

When the request names a pilot

Third request, same writer: send your original pilot.

"Night Line" isn't a pilot. "Seventh Street" isn't a pilot. A pilot establishes a world and its people from nothing; a mid-series two-hander assumes the audience already knows them, and a mid-series procedural assumes they already know the job. Neither can be relabelled into the requested object, and doing so costs you more than the missed opportunity — an evaluator who reads fifty pages expecting an origin and gets a season-two conversation learns something about your care with instructions, which is not what you wanted the sample to say.

So the choice collapses. What's left is one of two honest positions. If you have a completed pilot, it's the object, subject to everything else the request says. If you don't, you don't have this submission, and the options are to write the pilot, to send what you have with an accurate description of what it is, or to wait for a request that matches the scripts you've finished. Which of those is best depends on how that recipient handles partial answers, which isn't something to guess at. Nothing here says a pilot is the better sample in general; it says a named object can't be satisfied by a different one.

The package

One script. The one the request asked for. No second script tucked behind the first, no scene supplement, no five-paragraph note explaining that the enclosed episode is actually a departure and the better one exists elsewhere — that last one is the most common way writers try to escape this trade-off, and it reads as an unwillingness to be evaluated on what you sent.

Check the labels before you upload. If the script is an unproduced spec episode, say so; the reader will work it out anyway, and a mislabel costs trust at exactly the moment you need it. If the request permits one context sentence and your sample is the exception, include it, factual and unadorned — what the script is, what it assumes, how it differs from the series' ordinary shape. If your sample is the baseline, don't add the sentence; there's nothing to correct.

And check the object once more at the point of sending, because the third request in this article is the one people trip over. A synopsis is the object of a different request. A short episode example is the object of a different request. A required pilot cannot be replaced by an unrelated best episode, however good it is.

Choosing the loss you can name

For the two eligible requests above, the answer isn't the same script, and that's the whole lesson. Craft request, exceptional script, one honest sentence about what it is. Ordinary-show request, representative script, and the loss of your best work is a cost you accept rather than explain away. Neither is automatically superior, and neither is the safe answer in every situation.

What you can control is accuracy. Whether a particular reader prefers the exception to the baseline is not something this comparison can tell you, and it's not something to try to fix by sending more. Pick the script the request is buying, name what it doesn't show in the space you're permitted, and send nothing else. The submission that loses your interest is the one where the writer sent an exception with no sentence and hoped nobody would look closely at the series it came from.

Frequently asked questions

How should I decide between my strongest script and my most representative one?

Read the request for what it is actually buying. If the purpose is writing ability, the exceptional script is generally the stronger sample, provided it is eligible and you describe what it is in one factual sentence. If the purpose is the ordinary operation of the series, the representative script is more informative, though your best work stays in the drawer. If the request names a pilot and neither script is a pilot, both candidates are out.

Does one script show both my craft and how I would run the show?

No. A script is evidence in two directions, about the writer and about the series, and those come apart. A confined two-hander can show scene construction, subtext, escalation and dialogue under pressure, but it cannot show guest construction, procedure, multi-strand structure or the show's usual volume. A solid ordinary episode can show the baseline without showing your ceiling.

If my exceptional script is atypical for the series, should I explain that?

Yes, when the request permits a note and the sample is the exception. Include one factual context sentence saying what the script is, what it assumes, and how it differs from the series' ordinary shape. That is description, not defense. If your sample is the baseline, do not add the sentence; there is nothing to correct.

What if the request asks for an original pilot but my best scripts are episodes?

A mid-series two-hander or procedural is not a pilot and cannot be relabelled into that object. If you have a completed pilot, send it subject to the rest of the request. If you do not, the honest options are to write the pilot, send what you have with an accurate description of what it is, or wait for a request that matches finished scripts. Which is best depends on how that recipient handles partial answers, which is not something to guess.

Is a pilot automatically a better sample than an episode?

No. Nothing here claims a pilot is the better sample in general; it says a named object cannot be satisfied by a different one. The request's object comes first: completed episode, short episode example, developed synopsis and original pilot are different things.

Further reading

• Write TV Episode Examples That Show the Range of the Series — When the requested object is short episode examples, this shows how to build a small set that demonstrates range without becoming a different show.
• How to Write a Pilot or Sample Episode Synopsis — If the request names a synopsis rather than a script, this explains how to show an episode's movement and how to label proposed versus scripted material.