Running a live quiz without losing the room

Three things decide whether a live quiz holds a room: how fast you move, how hard the questions are, and whether the score still means anything by question ten. You can set all three at your desk, before anyone joins.

Put the timer where the thinking is

Most decks use one timer for every question, usually twenty seconds. That shape fits almost nothing. A recall question ("which year did this happen") is answered in four seconds or not at all, and eighteen of those twenty seconds are dead air where people check their phones for other reasons.

Match the timer to the work. Recall gets ten seconds. Anything requiring a calculation or a comparison gets thirty to forty-five. If you are asking people to read a chart before answering, the timer starts when they have had a chance to look at it, not when the slide appears.

The signal that you have this wrong is easy to read live: if the answer counter hits full well before the timer expires, every remaining second is costing you attention.

Ask questions that split the room

A question everyone gets right teaches nothing and generates no conversation. A question everyone gets wrong is usually badly worded rather than hard. The useful zone is the one where the answers split, because a split is a discussion you did not have to manufacture.

In practice that means writing distractors that are actually plausible. "Which of these is the capital of Australia" with Sydney, Melbourne, Canberra and Paris asks the room nothing. Drop Paris. Now it is a real question, and the reveal is worth talking about.

If you want a reliable split, ask about something people believe confidently and incorrectly. Those are the questions people remember a week later.

Decide what the score is for

Scoring changes behaviour, and it is worth being deliberate about which behaviour you want.

Speed-weighted scoring, where answering faster is worth more, produces energy and a moving leaderboard. It also punishes the careful thinker and rewards the person who guesses early. That trade is fine for a warm-up or a social quiz and actively bad for an assessment.

If you are checking understanding rather than running a game, turn the timer off or lengthen it substantially, and let the standings be about accuracy. The room reads the difference immediately.

Use the standings as punctuation, not wallpaper

Showing the leaderboard after every single question flattens it. By question six nobody is looking. Showing it three times in a twenty-question quiz, and saying something about it out loud each time, makes it an event.

The thing worth narrating is rarely who is winning. It is movement: someone who was outside the top ten a moment ago and is now fourth. That is the story the room actually reacts to.

Have an answer for the phone that will not connect

In any room of thirty people, one phone will be on a captive-portal wifi, out of battery, or otherwise not joining. Decide in advance what you do, because deciding in the moment costs you two minutes and the room's attention.

The usual answer is pairing: that person plays with a neighbour. Say it once at the start, before it happens, and it stops being an interruption.

What this looks like in practice

A twenty-question quiz that works usually has a spread of timers between ten and forty seconds, four or five questions where the answers genuinely split, standings shown three times, and a stated rule for people who cannot join. None of that is about which tool you use. It is about deciding, before you are standing in front of the room, what each question is for.