01The habit that separates setters from people who think they can set
Every question has two ends: the answer end and the question end. A novice picks up a fact and immediately starts writing at the question end — "What is the name of the…", "In which year did…", "Who was the first person to…" — building forward, hopefully, toward something. A working setter begins at the other end entirely. The answer comes first, on the page, before a word of the question exists. Everything else follows from that.
This sounds like a small procedural quirk. It is not. The order determines everything: what kind of question you end up with, whether it is fair, whether it has the right answer, and whether the question is even necessary.
02Why novices write from the wrong end
The impulse to write from the question end is natural and understandable. You have just learned something interesting — say, that the Eiffel Tower grows taller in summer, because heat causes the iron to expand — and you want to make a question from it. So you type: What happens to the Eiffel Tower in hot weather? Then you check the answer you have in mind: it grows, it gets taller, it expands. And here the trouble starts. Any one of those phrases is defensible. It expands and it gets taller and it grows by up to fifteen centimetres are all correct, and they are not the same answer. You have written a question before you knew what the answer was, and now you have written a question with several answers.
The experienced setter writes: It grows taller (by up to fifteen centimetres) at the top of the card, then asks: what question, asked of these people in this room, will produce exactly that phrase? The answer constrains the question. It tells you which part of the fact to illuminate, which part to withhold, and how specific to be. Write the answer first, and the question almost writes itself. Write the question first, and you spend the next ten minutes tidying up the ambiguity you created.
04What the answer tells you
A correct answer is a lens. Hold it up and look back through it at the question you need to write, and several things become immediately visible.
The first is specificity. If the answer is a number — a date, a distance, a running time — the question must ask for that number in a way that excludes all other numbers. If the answer is a name, the question must point at that name and only that name. This sounds obvious, but question-first writing constantly produces questions that admit a cluster of near-correct answers, because the setter was already in love with the phrasing of the question by the time they checked what it was pointing at.
The second thing the answer reveals is what the question must withhold. Good questions hide the answer in plain sight: everything in the question is relevant, nothing gives the game away immediately. When you start from the answer, you can see exactly what you are concealing and what you are giving as legitimate help. When you start from the question, you frequently include a word or phrase that makes the answer obvious to anyone paying attention — or, equally bad, you withhold information that would actually have been fair to provide. The answer-first method forces you to manage that balance consciously.
The third thing is redundancy detection. Starting from the answer, you ask: is there any other answer that honestly satisfies the question I am about to write? If yes, the question has a problem before it is even finished. The novice catches this problem later, usually during the quiz itself, when someone shouts a different correct answer from the back of the room. The setter catches it at the desk.
05The shape it gives the finished question
A question written answer-first has a recognisable architecture. It tends to move from context to target: it establishes the world of the answer — the period, the field, the geography — and then narrows, with each clause removing possible answers, until the question ends at the precise empty slot the answer fills. The answer snaps into place. Questions written in the opposite direction often feel loose at the ends, as though the question and the answer are not quite talking to each other.
This also affects length. An answer-first question is almost always the right length, because every clause is doing work: each one is either establishing context the solver needs or narrowing the field of possible answers. A question-first question tends to ramble, because the setter is feeling their way toward an answer they do not yet quite trust. The rambling is the uncertainty made visible. When you edit it out, you are usually cutting the question-first section and keeping the answer-first rewrite.
06Research and the reverse search
Writing the answer first is not only a compositional technique; it is also the natural posture of the setter doing research. When you search for material, you are not looking for interesting questions — you are looking for interesting answers. A name, a date, a place, a number that will surprise and satisfy: that is the unit of raw material. The question is constructed around it afterwards.
This shapes how setters read. They read looking for the kind of thing that can be the answer to a question — something specific enough to be unambiguous, surprising enough to be worth asking, checkable enough to be defended. If you read for questions first, you will find only what you were already looking for; the best material, sitting three sentences away from the obvious peg, will escape you. Setters who describe their research process almost always describe it as collecting answers, not questions. The file they keep is a file of answers. The week of construction is largely the work of fitting questions around the answers already banked.
07When the answer changes the question's difficulty
Answer-first writing also gives the setter direct control over difficulty in a way that question-first writing does not. If you have the answer in front of you, you can look at your question and ask: at what point does a solver know they are heading for this answer? If the answer announces itself immediately from the first clause, the question is too easy for an experienced room. If the answer could not plausibly be inferred until the final word, the question may be too hard — or, worse, it may feel cheap, as though the question was designed to deceive rather than to challenge.
This calibration is impossible to do accurately if you do not know the answer. You cannot tune a question to its answer when the question is already written and you are hoping the answer will cooperate. The answer comes first because difficulty lives in the gap between what the question tells you and what the answer requires. Only the setter who holds both ends at once can measure that gap and decide whether it is the right size for the room.
09The discipline in practice
None of this means a setter always writes questions slowly, or that good questions cannot emerge quickly. It means the first habit a new setter acquires should be to write the answer down before writing a single word of the question. On a card, on a screen, in a notebook — whatever the system, the answer occupies the top line. Everything written below that line is accountable to it. If the question, once drafted, could produce a different answer from the one at the top, the question gets rewritten. The answer never changes to suit the question; the question changes to suit the answer.
That asymmetry is the whole discipline in miniature: the answer is the fact, the thing out in the world, fixed. The question is language, malleable and revisable, and it serves the answer. Reverse that relationship and you are no longer a setter constructing questions. You are a writer free-associating around topics and hoping the facts will line up. Sometimes they do. Often they do not. And you only find out during the quiz, in front of a room full of people, when it is far too late.
Absolute Trivia is an independent publication about how quiz questions are written. It is not a quiz service, question bank or league.