A round with a shape starts easy, builds to the material you actually wanted to ask, and ends on something everyone can answer. A room that has been through that arc feels like it went somewhere together, and it is the arc doing that, not any individual question. Building one takes about ten minutes with whatever questions you already own, and it is the most transferable skill in hosting: learn it once and every deck, list and box of cards you ever pick up gets better.
The reason it works is the oldest result in this field. Reading the questions in the order you found them, which is what most people do, throws that result away for nothing.
Forty-five minutes of proof that order matters
The famous 36-questions study is usually remembered for the questions, but its real finding is about sequence. In 1997, Arthur Aron and colleagues had pairs of strangers work through 36 questions in three escalating sets of twelve, about 45 minutes in all, while a control group made small talk for the same time. Immediately afterwards, question pairs rated the new relationship as closer than the closest relationship in the lives of 30 percent of comparable students (Aron, Melinat, Aron, Vallone and Bator, Personality and Social Psychology Bulletin, 1997). The questions were not magic individually. Set one is easy, set three is not, and nothing real is asked for until the earlier sets have already brought the cost of answering down.
The design rule hiding in that study: escalation is a property of the sequence, not of any question in it. Which means you can build it out of any questions you have.
1. Sort into three piles by what an answer costs
Ignore topic completely. Ignore how funny each question is. Sort by one measure only: what it costs a person to answer honestly in front of this group. Three piles is enough, cheap, medium and real, and most questions sort themselves in a second or two.
Cheap costs nothing: a vote, a preference, a pointed finger. Someone who has met nobody in the room can still answer it.
- Most likely to have a playlist for this exact situation?
- Would you rather always choose the restaurant or always choose the music?
Medium costs an opinion or a small piece of your actual life, the material you probably picked the questions for.
- Hot take: snack choices reveal more than star signs.
- Tell us: how you ended up living where you live now?
Real costs a genuine answer, the kind a room gives at the end of a good night and not at the start of one.
- Would you rather have more time or more courage?
- Would you rather be completely understood by one person or liked by everyone?
All six of those are from one deck, which is the point: any decent source of questions already contains all three piles, shuffled together and hiding each other.
One caution while you sort: cost is a property of the room, not of the question. "How you ended up living where you live now" is medium for people who met an hour ago and cheap for housemates, because they already know. Sort for the specific room you are hosting, and if you are between two piles, sort down. A real question played early costs more than a cheap question played late.
2. Deal roughly seven, seven and six, in that order
Seven cheap, seven medium, six real, played straight through. The middle pile is what you came for, and the first pile is what makes the middle pile possible: seven easy cards is roughly how long a room takes to stop auditioning and start playing. It always feels like too many when you are building the round, because you are imagining the room already warm. It is not warm at card one. Trust the ramp you built and it will be warm by card eight.
The last pile gets six rather than seven for a reason worth knowing: real cards are slower. Each one starts a conversation the cheap cards do not, so six of them fill the same minutes that seven cheap ones do, and a round that ends on schedule ends while people still want more.
If you have fewer than twenty questions, keep the proportions rather than the counts. A twelve-card round wants four or five from each pile, in the same order. And if you are unsure whether the room wants the third pile at all, build it anyway and watch card fifteen: a room that is leaning in will pull you into the real pile on its own, and a room that is not will tell you just as clearly, at which point the remaining cheap cards are a perfectly good landing.
3. Rotate the shape every few cards
Cost is the vertical axis. Shape is the horizontal one: point at someone, take a side, choose between two things, tell a story. Within each pile, vary the shape so that no run of four cards asks for the same kind of response, because four story cards in a row turns a game into a round of speeches and four votes in a row turns it into a survey, no matter how good the individual questions are. A choice card after a story card also does quiet repair work: it hands the floor back to the whole room, which keeps everyone answering without the host doing anything.
Shape is also how you spend the cheap pile well. Seven easy cards does not mean seven identical votes; a point, a vote, a choice, a tiny story, a side taken, in any order, teaches the room all five of its moves while the stakes are still zero, so that by the time the middle pile arrives nobody is learning rules any more.
4. Put the sharpest card two thirds in, not last
The bravest question you are willing to ask belongs around card thirteen or fourteen: late enough that the room has earned it, early enough that its answers have somewhere to go. A great card at card twenty is a waste, because the conversation it starts gets cut off by the ending.
- Would you rather know how your friends describe you or never hear it?
That is a card thirteen. Card twenty should be one everybody can answer in a phrase, so the round ends with the whole room in it:
- Would you rather leave tonight with a new plan or a new inside joke?
5. Mixing sources? Mix at the pile level
Combining an app deck with a paper deck, or two halves of different boxes, works well with one rule: sort everything into the three piles first, then deal from the piles. Dropping an unsorted handful of favourites into an already-sorted run flattens the arc you just spent ten minutes building, because a real question landing at card three does not read as bold, it reads as misjudged, and it resets the room's temperature instead of raising it.
This method is what the app literally does
Chinwag's staged decks, eleven of them now, are this exact build baked in. Each one is sorted by disclosure cost into three named stages, Break the Ice, Warm Up, Turn Up the Heat in New Friends, and when a staged deck plays on its own, a twenty-card round deals seven from the first stage, seven from the second and six from the third, in that order, with the stage name shown next to the card counter. None of that is required to use the method. It is the method, applied for you, and the stages section of how it works describes exactly what the deal is doing. A paper deck and ten minutes of sorting gets you the same arc by hand.
Order is half the craft. The other half is what makes an individual question worth a slot in any pile at all, and that has a test of its own: how a question earns its place.
See a built round in the wild.
New Friends is free and is the deck the three-stage build was designed on. Play one round and you can feel the seven-seven-six underneath it.
Read nextHow a question earns its placeThe four-part test every card has to pass →