Anyone can write ten questions and read them off a phone. Running a quiz people look forward to is a different job — it's closer to hosting a small event than reading a list. The good news: almost all of it is planning you do once, not talent you either have or don't.
Before the night: the planning that actually matters
Most quiz nights that fall flat didn't fail on question quality — they failed on logistics nobody thought about until it was too late. Settle these before you write a single question:
- Round count and length. Six rounds of eight questions (48 total) runs about 60–75 minutes with scoring breaks — long enough to feel like an event, short enough that nobody's checking their watch.
- Team size cap. Cap tables at 4–6. Bigger teams don't need to be smarter to win — they just need more people who might know any given answer, which quietly punishes anyone who came in a pair.
- How scoring gets tallied. Paper score sheets passed table to table are the single biggest source of quiz-night arguments. Decide upfront how scores get collected and added — by hand, or automatically (more on this below).
- A prize that isn't just "bragging rights." It doesn't need to be expensive. A bar tab, a bottle, or a silly trophy that travels to next week's winner all work better than nothing.
- A tie-breaker plan, decided before anyone needs it. Deciding on the spot in front of a tied room looks unfair even when it isn't.
Writing questions people enjoy answering
The best quiz questions aren't the hardest ones — they're the ones that make a table go "ohh, I know this" a few seconds after reading it, or genuinely argue about the answer. A few rules that hold up across almost every good quiz:
Mix the difficulty on purpose
A round that's all hard questions demoralises everyone except the one table that happens to specialise in it. Aim for roughly: 2 easy (everyone gets these — momentum matters), 4 medium, 2 genuinely hard. The easy ones aren't filler; they're what keeps a struggling table in the game emotionally, not just mathematically.
Avoid anything answerable by reflex phone-check
"What year did X happen" with X being extremely googleable invites cheating and nobody enjoys policing it. Questions with a twist, a connection, or that need reasoning from a clue age much better — and they're more fun even for people who wouldn't cheat anyway.
Specialist rounds should still let non-specialists score
A pure sports or pure music round can shut out half the room. Better: a themed round with a spread of angles (a sports question about a scandal everyone heard about, not just a scoreline) so it rewards the specialist table without being unplayable for everyone else.
Picture and audio rounds are worth the setup effort
A "name the logo," "guess the film from one frame," or "name that intro" round breaks up the rhythm of straight Q&A and is consistently the round people mention afterwards. It takes a bit more prep than typing ten questions, and it's usually worth it.
Team scoring that's actually fair
The most common structural mistake in home-run quizzes: scoring by team total, which silently rewards the biggest table. If Team A has 8 people and Team B has 3, Team A doesn't need to be smarter to win a general-knowledge round — it just needs more lottery tickets. Score by average per player instead, or cap team size hard, and a table of two has a genuine shot against a table of six.
It also matters that scores update during the night, not just at the very end. A leaderboard that only appears once, at the finale, removes the tension that makes people play the last round harder. Showing a running leaderboard between rounds — even just top 3 — measurably changes how engaged the room stays.
Keeping the night moving
The single biggest killer of pace is manual scoring: reading answers aloud, collecting paper sheets, walking them to a table, adding up columns by hand while the room waits and starts checking phones. Every minute spent tallying is a minute the energy in the room drops.
This is the one part of quiz hosting that's genuinely easier now than it used to be. Instead of paper and a calculator, teams answer on their own phones (a QR code or a short link is enough — nobody needs to install anything), and the leaderboard updates itself the instant the last answer comes in. PollSlide was built around exactly this: you write the questions once, present them however you're already presenting (even just from your own screen), and scoring, team averaging, and the leaderboard are handled automatically — so the only thing you're managing live is the room, not a spreadsheet.
If you'd rather stay with pen and paper, that's a completely reasonable choice too — just budget real time for scoring between rounds, and consider a "scorer" role separate from the host so the person reading questions never has to also do arithmetic under pressure.
A finale that people talk about afterwards
The last round is what people remember on the way home, more than the average of everything before it. A few formats that consistently land:
- Last-one-standing: one wrong answer and you're out, sudden-death style, run until one table remains. High drama, works especially well as a genuine tie-breaker.
- Wager round: teams bet a portion of their score on their confidence before answering — it lets a team in last place claw back into contention, which keeps everyone invested until the literal final question instead of checking out once they're mathematically out of it.
- Speed final: same questions, but the fastest correct answer scores highest — turns a close leaderboard into genuine tension in the room.
Whichever you pick, announce that it's the finale before it starts. People play differently — and pay more attention — when they know it's the round that decides the night.
Common mistakes worth avoiding
- Questions with an ambiguous "correct" answer. If two tables can reasonably argue their answer is right, you'll spend the next five minutes refereeing instead of quizzing. Test genuinely ambiguous questions on a friend before the night.
- No warm-up while people arrive. The first ten minutes of a quiz night is often people trickling in and ordering drinks. A light trivia loop on a screen gives latecomers something to do and signals the night has already started.
- Reading every answer explanation at length. A one-line "fun fact" after a reveal is memorable; a two-minute tangent after every single question drags a 60-minute night into 100.
- No plan for a dead heat. Decide your tie-breaker format before you need it — see the finale section above.
Frequently asked questions
How many people do I need to run a good pub quiz?
It works from as few as 2 teams, though the energy noticeably improves from about 4 teams (roughly 15–20 people) upward — enough that the leaderboard has real movement and there's a bit of table-to-table banter.
How long should a pub quiz last?
60–75 minutes is the sweet spot for a casual weekly night — six rounds of about eight questions, with brief scoring pauses. Longer than 90 minutes and attention reliably drops off, even for a good quiz.
Do I need special equipment?
No. A screen or a loud enough voice for questions, and something for players to answer on — paper, or their own phones if you're using a tool like PollSlide. No buzzers or extra hardware required.
How do I stop one table from dominating every week?
Score by average-per-player rather than team total so bigger tables don't have a structural advantage, and rotate in genuinely varied round themes so no single specialty wins every night.
What's the easiest way to run this if I've never hosted before?
Start with six rounds, the difficulty mix above, and a running leaderboard between rounds. The format does most of the work — you mostly just need good questions and a room that can hear you.