What 300 Wedding Guests Taught Us About Quizzes
We pulled the answer data from every wedding that has run a quiz on QuizYourGuests: 322 players, 5,528 answers, 227 questions. Here is what guests actually know, what they never know, and whether they get bored.
I've written a fair bit of advice on this blog based on watching a handful of receptions and running one of my own. That's fine as far as it goes, but it's still anecdote wearing a suit. So I went and looked at the numbers instead.
Every wedding that has run a quiz on our platform, aggregated: 322 players, 5,528 individual answers, 227 questions. Nothing below is a guess and nothing below comes from anyone's personal quiz. No question text, no names, no couples - just the shape of what happened when several hundred people at several different receptions were handed a phone and asked how well they knew the bride.
Some of it confirmed what I already believed. Two things did not, and one of them is the most useful thing I've learned about writing these questions.
Guests do not get bored
This was the finding I most wanted to be true and least expected to be.
The obvious worry with a multi-round quiz is decay. Round one gets a flurry, round two gets a shrug, by round four you're the person who made everyone do homework at a party. So I measured it: for each round, what share of the players who had joined that wedding actually submitted an answer?
- Round 1: 66.8%
- Round 2: 77.9%
- Round 3: 75.4%
- Round 4: 71.6%
- Round 5: 70.1%
Participation goes up after round one, then settles into a slow drift of a couple of points per round. Across everything, 72.8% of possible answers were submitted. Bear in mind the later rounds come from fewer weddings, so round four and five are thinner samples than round one - but the direction is unmistakable, and it holds across every wedding long enough to have them.
The interesting number is the first one. Round one is the worst-attended round of the night. That fits what I remember from our own wedding: at the start, people don't know the quiz exists yet, half of them are in the car park, and the QR code by the door has been seen by about a third of the room. By round two, word has got round and the tables have organised themselves.
Two conclusions. First, stop worrying about the quiz outliving its welcome - a well-spaced quiz survives dinner, speeches and the first hour of dancing without visible fatigue. Second, do not judge it on round one. If your first round lands at two-thirds participation, that isn't failure, that's the normal warm-up. Put your least essential questions there and save the good ones for when everyone's actually looking.
Overall, guests get about half of it right
Across all 227 questions, guests got 53.5% correct, and the median question scored 57%. That's roughly where you want a quiz to sit: a shade harder than a coin flip, comfortably short of impossible.
The spread underneath that average is the real story. 73% of questions landed between 30% and 90% correct, which is the healthy band. But 20 questions came in under 20%, and 17 questions were answered correctly by literally every single person who saw them. The two ends are where the lessons live.
What every guest knows: the everyday facts
The questions that scored perfectly weren't clever. They were the things a friend would know without thinking: questions about the couple's pet's name or what one of them does for a living scored 77 out of 79 - 97.5%.
I'd assumed identity facts in general would behave this way, but they don't, quite. Middle names, for instance, were a coin flip. The pattern isn't "personal facts are easy", it's "present-tense facts are easy". What the dog is called, what you do on a Tuesday, where you live now: everyone in the room has been told these in the last year. Anything requiring a guest to have retained a detail from before they knew you is a different question entirely, no matter how basic it feels to you.
Which is fine. You want a handful of these. They're the questions that let the plus-one who's met you twice put something on the board, and they cost you nothing.
What no guest knows: numbers and drink orders
Here's the reliable disaster zone.
Questions asking guests to count something about the couple - possessions, tattoos, qualifications, photos on a phone - scored 33.8% across 160 answers, and the worst of them scored zero. Nobody knows how many of a thing you have. You barely know.
Worse, and this genuinely surprised me: questions about the couple's favourite food or drink scored 25.9% - 87 correct out of 336 answers, with a median of 21%. Your comfort takeaway, your coffee order, your dessert of choice. These feel like softballs when you're writing them. They are the hardest category in the entire dataset. Your friends have watched you eat for fifteen years and cannot pick your order out of four options, because the wrong options are all things you'd also plausibly order.
The bottom five questions of all time drew 130 answers between them and got 5 right - 3.8%. Two were counting questions, one was a drink preference, one asked which half of the couple did something first, and one was pop-culture trivia that turned out to be much less common knowledge than its author thought.
None of this means don't write them. A question everyone gets wrong is a good joke once, and the reveal is genuinely funny. Two or three per wedding is a spice. A whole round of them is a diet, and a leaderboard where nobody scores stops being a leaderboard.
The one that changed how I write questions
This is the finding I'd print at the top of the question form if I could.
Take the couple's own story - how they met, when, where the proposal happened. Asked at headline level, those questions scored 84.3%. Asked at fine-detail level - what she was wearing, which venue exactly, what he was doing on his phone at the time - the same category of question, about the same couples, scored 39.7%.
Same subject. Same guests. Half the score.
Everyone in that room knows your story. Almost nobody knows the footnotes, because the footnotes only ever got told to about four people. The difference between a question that lands and a question that dies is very often not the topic at all - it's how far you've zoomed in. When a question feels too easy to you, that's usually because you were there. Resist the urge to sharpen it.
The safest question shape
If you want something that reliably lands mid-table, use the this-or-that construction: which of them is more likely to X, who usually does Y, which of these has one of them never done.
These scored 51.8% with a median of 58%, and, more importantly, they almost never bomb - only one of twenty fell below 30%. A guest who knows nothing can reason their way to a decent guess, and a guest who knows you well gets it right. That's exactly the curve you want.
General trivia, for what it's worth, sits slightly below couple questions overall - 50.2% versus 54.5% - and wedding-tradition trivia specifically comes in at 44.7%, harder than most couples expect. It's still worth including as neutral ground, but don't treat it as the easy filler round.
The shape of a typical wedding quiz
Briefly, because I've covered this properly in the complete guide to wedding quizzes for guests:
- Median 22 players per wedding, ranging from 7 to 83.
- Median 3 rounds of 5 questions, 24 questions in total.
- Almost every wedding opens its first round right at the start of the reception.
- Median gap between rounds: 30 minutes. Median total quiz window: 2 hours 15 minutes.
- Every single wedding used exactly two teams. The bigger team took a median 58% of players, and the most lopsided split we've seen, 68/32, still produced a close finish - team scores are sums of individual scores, so who keeps playing matters far more than who signed up. More on that in Team Bride vs Team Groom.
What I'd do with all this
Write more questions about your present tense and fewer about your archive. Zoom out one notch on every story question. Use this-or-that phrasing when you're unsure. Keep the unanswerable ones as garnish, not the meal. And put your weakest round first, because that's the one fewest people will see.
Caveats, honestly stated: this is a modest number of weddings, they skew towards couples who are the sort of people who put a quiz at their wedding, and I have no data at all on the guests who never joined. I'll rerun this when the numbers double.
Want to see what your guests would see? QuizYourGuests schedules the rounds, splits the teams and scores everything automatically, so you can set it up months ahead and forget about it. There's a free demo on the homepage.