Run a TV-Format Handoff Around a Difficult Episode, Not a Page-by-Page Read

There's a moment in a lot of format handoffs where the room has worked through forty pages in order, everyone has nodded, and nobody has yet asked what happens if the score is level with one round to go. The document has been delivered. Whether it can be used is still untested.

Those are different things, and the gap between them is where most of the trouble lives. A format bible is meant to carry production know-how to people who weren't in the room when the show was made. That is roughly how the FRAPA bible template is described by the practitioner who wrote it — as transferable production know-how, with some elements treated as essential and others as adaptable, and with later local adaptations feeding back into later editions. That description tells you what a bible is for. It doesn't tell you whether a particular bible, in a particular handoff, is doing it.

The way to find out is to stop reading and start asking. Take one ordinary installment and one that goes badly. At each consequential point, ask the receiving team to say what happens next and where they got that from. Then sort what you find into three piles, because they have three different fixes.

Everything below uses an invented format and an invented bible extract, so the method is visible. No handoff session was run to produce this, and no one's paperwork is being described.

Agree the task before you open the document

A handoff exercise needs a subject. Not "review the bible" — that's a reading agenda, and it produces a nodding room. Pick a concrete responsibility the recipient actually holds: preparing a specific episode, briefing the floor team for a stage, deciding what the next role needs to receive and when.

Confirm two things first. Which version of the bible you're working from, because a handoff that resolves questions against an old draft is worse than no handoff. And the scope of the session, in writing, before anyone starts. Two operating situations, the recipient's own role, an hour and a half. That's a handoff exercise.

What it is not: a rehearsal, a build, or a free development pass. The receiving team is being asked to say what they would do and where the supplied material leaves them, not to produce a working episode so the format owner can watch. If the session turns into unpaid production work, both sides have lost the point of it. Owner approval and operational testing come later, if they come at all.

Walk an ordinary episode all the way through

Here's the invented material. A half-hour two-team quiz called Banked, with a four-page handoff extract from what I'll call Format Bible v2.4.

• Signal — eight buzzer questions, one point each.
• Ladder — one run per team. Three questions worth 2, 3 and 4 points; the team chooses the order; a team may stop after any correct answer. An answer that is not accepted ends the run, and points already banked are kept.
• Cut — one question to each team, two points each, Team A first, then Team B.
• Reset — the on-stage scoreboard is cleared to a neutral card while the stage is reset. The running order lists three steps: clear the board, move the lecterns, bring the host to position.
• The Quorum — three questions, two points each. The trailing team chooses the first category and answers it; the teams then alternate, and the team that chose first answers the third. Highest final score wins.
• Reveal — "The result is confirmed before the lights come up."
• Presentation — "Reveal styling and hold duration are set locally."

The ordinary episode runs clean. Signal goes 5–3 to Team A. On the Ladder, Team A takes the two-point question and then the four, and stops, banking 11. Team B takes the two and the four, then misses the three; the run ends and the 9 stays with them. On the Cut, Team A answers correctly and Team B doesn't. Going into the Quorum it's 13–9 to Team A.

The Quorum settles it early. Team B chooses first, as the trailing team, and takes the first question to 11. Team A takes the second to 15. With one question left at two points, Team B's ceiling is 13. Team B answers the third and misses; the reveal confirms 15–11. Nothing is disputed, all three Quorum questions are asked, and the reveal shows a result the room has known since before the last one.

Now run the recipient through it and record, at each point, what they need to decide and where the answer came from.

Eight rows, and five of them resolve cleanly out of the document. That is a decent bible. The three that don't are the read-back on Reset, the confirmation before the lights come up, and reveal styling and hold — and the ordinary episode is precisely the situation in which none of them bite. The result was settled before the last Quorum question, so nobody needed to know who confirms it. The tally was never questioned, so nobody needed a written read-back step.

One row did surface anyway. The read-back after the reset is real — it happens every episode — but it happens because the scorer has always done it, not because the extract says so. That's tacit knowledge, and tacit knowledge is often the best kind. It is also a dependency with an owner attached. If that scorer is ill on the day, the step goes with them, and nothing in the paperwork notices.

Now hand them the episode that goes wrong

Second situation, same format.

Signal goes 5–3 to Team A. On the Ladder, Team A takes the two-point question and stops, holding 7. Team B takes the two, then the four, then attempts the three-point question last. The answer isn't accepted against the card, so the run ends and Team B keeps 2 + 4 on top of its 3 from Signal: 9. On the Cut, Team A answers correctly and Team B doesn't. Entering the Quorum: 9–9.

Team B has now attempted all three Ladder questions, which matters in a moment.

Before the Quorum begins, a message reaches the producer from the format team's question editor. Team B's rejected answer appears on the corrected accepted-answer list; the card in studio didn't include it. If the ruling is applied, Team B goes to 12 and leads by three. If it isn't, the score stands at 9–9.

Ask the recipient to move. Three points come back empty, and they're empty in three different ways.

Who decides whether a closed stage's score can be reopened? Nothing in the extract. The question editor can confirm the key; whether the show's score changes mid-record is a rules question, and the local team has no authority to answer it. Team B's run is spent as well: it ended on the last of the three Ladder questions, so even a ruling in Team B's favour has nothing left to replay. The only remedy on the table is a changed score.

Who confirms the result for the reveal? The extract says the result is confirmed before the lights come up. It doesn't say by whom, to whom, or against what. In the ordinary episode this was invisible. Here the confirmed figure is 12–9 or 9–9, and the sentence will be leaned on hard.

How long does the reveal hold, and how is it styled? The extract hands this to the local team on purpose. This is not a defect. Someone has to write it down anyway, because "we'll figure it out on the day" is how local discretion turns into a decision nobody made.

The recipient should stop at the first point rather than pick an answer. A confident improvisation in the control room becomes an unofficial rule by the end of the run, and nobody agreed to it.

Three kinds of gap, three different repairs

These three empty points look alike — the recipient couldn't answer any of them — and they need three different responses.

A missing instruction gets added. The read-back after the reset belongs in the running order, as a step: the scorer reads the running tally to the producer, who states it back and checks it against the card on stage. That's a new step 4 on the Reset page, and adding it renumbers what follows, so the change record needs to say which version it lands in.

An ambiguous instruction gets rewritten, not explained. "The result is confirmed before the lights come up" survives in a document because everyone in the room assumes the same reading. The repair is to name the actor, the action, the moment and the evidence: before the lights come up, the scorer reads the final tally to the producer; the producer states it back and confirms it against the scorer's sheet; the host receives the confirmed figure from the producer. Four specifics in one sentence. If the sentence can't absorb them, add a line and cross-reference the scoring page.

A local choice gets recorded locally. Reveal styling and hold duration stay with the receiving team. What goes into the local production note is the choice itself, the person who owns it, and the boundary it sits inside — the figure displayed has to be the figure the producer confirmed. That's the whole entry. Sending it up to the format owner as a gap wastes everyone's time and slowly turns a flexible format into a rigid one. More instructions are not automatically a better handoff.

Then there's the fourth thing, which isn't a gap in the local paperwork at all. The question of whether a closed stage's score can be reopened belongs to whoever holds rules authority under the agreement — a format-owner decision, not a local one, and not one to guess at from the outside. Ownership and permitted adaptation are governed by the actual agreement; the practitioner description of a template doesn't settle it. Because the Ladder run was exhausted, though, the recipient can say what the decision would and wouldn't produce: a corrected figure, not a replayed question.

The tie has a second layer worth noticing. If the ruling applies, Team B leads 12–9 and the Quorum's ordering rule works exactly as written: the trailing team, Team A, chooses first. If the ruling doesn't apply, the extract's "the trailing team chooses the first category" has no trailing team to name. So the owner list has one item that must be answered now and one that only becomes live depending on the answer. That's worth writing down in that order rather than as two flat questions.

Recheck the difficult step, and keep what the recheck finds

Once the Reset step and the confirmation sentence are revised, run the awkward episode again — only the awkward one, not the whole walkthrough. There is no reason to re-test the rows that already worked, and a full second pass invites the room to redraft things that were fine.

The recheck does two useful jobs. It confirms the confirmation sentence now answers the recipient's question in both branches: whoever ends up on 12, the recipient can name who confirms and against what. And it catches something the first repair missed. The new Reset step assumes the tally is settled when it's read back. In the awkward episode, the ruling arrives after that read-back and changes the number, and nothing in the revised step says who verifies it again. The step needs one more clause. That isn't a failure of the repair; it's what rechecking is for, and it only shows up because the recheck uses the awkward case rather than the tidy one.

What the recheck does not fix is the ruling itself. The recipient still stops there. The difference is that the list of things stopping them is now one item long instead of three, and it's an item they weren't entitled to answer.

Put the answers where the next person will find them

A revised reference, a short list of open items, and the difficult step rechecked. If that's where the session ends, most of the value evaporates by the following week, because the resolution lives in a meeting recap that nobody opens.

Keep the change record next to the thing it changes. Page or step reference, what it said, what it says now, which version it lands in, and who approved it. The open items go on a separate short list with the owner named and the decision that's still outstanding, phrased as a question someone can answer — not "tie-break needed" but "when teams are level entering the Quorum, who chooses first?" The question editor who sent the corrected key should see that list; so should whoever runs the next handoff, because the next team will hit the same two places.

Then leave the rest alone. Everything the recipient could already locate is already working, and the fix for a format bible that's functioning is not a longer format bible.

The point of the whole exercise is narrow and worth restating: the handoff works when the receiving team can point at the page that tells them what to do next. Where they can't, you've learned something the page-by-page read was never going to tell you.

Frequently asked questions

Why run a handoff around a difficult episode instead of reading the bible page by page?

A page-by-page read produces nods; a concrete exercise shows whether the bible can be used. Take an ordinary installment and one that goes wrong, ask the receiving team at each consequential point what happens next and where they got that, then sort the findings by the kind of repair they need.

What should be agreed before the document is opened?

Pick a concrete responsibility the recipient actually holds, confirm which version of the bible is being used, and write down the scope before starting, such as two operating situations, the recipient's own role, and an hour and a half. It is not a rehearsal, a build, or a free development pass; owner approval and operational testing come later, if at all.

What kinds of gaps appear, and how are they repaired?

A missing instruction gets added, such as a read-back step after Reset that renumbers what follows. An ambiguous instruction gets rewritten with actor, action, moment, and evidence, not explained. A local choice gets recorded locally with the choice, its owner, and its boundary. A rules question goes to whoever holds rules authority under the agreement and is not guessed at locally.

Why should the local team stop at the corrected-answer message in the difficult episode?

The message can confirm the rejected answer belongs on the key, but whether a closed stage's score can be reopened is a rules question under the agreement, not a local call. Because Team B's Ladder run was exhausted, a ruling would produce a corrected figure, not a replayed question. If the score stays 9-9, the Quorum's trailing-team rule has no trailing team to name.

What should be rechecked and recorded after the repairs?

Recheck only the awkward episode. That confirms the confirmation sentence answers both branches and catches that the new Reset step needs a clause for what happens if the ruling arrives after the read-back. Record the change next to what it changes with page or step, old wording, new wording, version, and approver; keep a short open-items list with an owner and an answerable question; leave working material alone.

Further reading

• The Host Also Competes. Who Controls the Game? — Useful when the empty point is about authority—who confirms the reveal, who reopens a settled score—because it separates who decides, who knows, and what the result counts for.
• Edit a Series Bible for the Collaborator Who Has to Build From It — When a bible sentence could be a rule, a sample, or an open question, this shows how labelling each stops the next reader guessing.
• What Did Your Unscripted Format Test Actually Prove? — If the recheck tempts you to claim the revised step now works, this keeps the claim inside what a single awkward run can actually show.