Why we built something different from conversation card decks

You have probably had one in the house at some point. A small box of questions — a gift, or something picked up on the way to a weekend away — with a promise on the side about going deeper. You shuffle it, you pull a few, and for one evening it does exactly what it said it would. You talk about something that isn't the boiler or next Thursday.

We like those decks. A box of good questions asks nothing of you: no account, no charging cable, no unspoken question about why someone's phone is out. It is one of the very few objects that reliably gets two people to sit down facing each other.

We still kept running into the same four walls with them, and eventually built something else. This is what the walls were — and where the box is still the better answer.

A deck is a starting gun, not a road

A box of prompts is designed for the first evening. It has to work cold, for two people who have not done this before, which means every card has to stand on its own. Nothing can assume you have already warmed up, because the shuffle might have put that card first.

That is a real design constraint, and it has a cost. The question you draw ten minutes in and the question you draw ninety minutes in arrive with exactly the same weight, even though the two of you are not the same two people at those two moments. By the end of the evening you are usually ready for more than the deck knows how to hand you — or, just as often, you are tired and it hands you something enormous.

You only meet each card once

This is the part that surprised us most. A deck is finite in a way that feels generous at first and then quietly runs down. Once a question has been asked, it has been answered. Come back to it in a month and you are not discovering something — you are recalling what you said last time and checking whether it still holds.

And because you naturally pull the interesting-looking cards first, the deck gets weaker as you go. The fifth evening is working with what the first four left behind. Nothing is wrong with the box; it simply spent its best material early, the way any fixed set of prompts has to.

The deck cannot tell what kind of night it is

Some evenings you have three hours and an empty house. Some you have twenty minutes before one of you falls asleep. Some you are both a little raw and want to be close without being examined. A shuffle cannot tell the difference, so somebody has to.

What that looks like in practice is one of you quietly editing: pulling a few cards out of the pile beforehand, sliding one back under the box, steering. It works. But whoever is doing it has stopped playing and started curating, and curating is a different experience from being surprised. If the same person ends up doing it every time, only one of you is actually in the game.

Saying no out loud costs more than it should

When a printed card lands wrong, there are two ways out: answer it anyway, or say no in front of someone you love. Either is fine once. Over an evening they add up, and the second one adds up faster than people expect — not because anyone is being difficult, but because declining something out loud is a tiny act of exposure, and you are already doing a lot of that.

We wanted a version where passing is an ordinary move. Not a negotiation, not a moment. Just: next.

So: four things we wanted instead

It should remember. If you have already answered something, it should not come back around unless you ask for it. The pool you draw from should get smarter about you, not thinner.

It should have a shape. An evening has an arc — you warm up, you open up, you land somewhere. A tool that knows roughly where it is in that arc can hand you a light question early and a real one later, instead of leaving the ordering to chance.

You should be able to set the ceiling before you start, and lower it halfway through without a conversation about it. How far tonight goes is a decision the two of you make once, out loud, when nothing is happening yet — which is a much easier moment to be honest in than the middle.

And it should keep going. The hundredth evening should be as well-supplied as the first. That is the one thing a finite box, however good, structurally cannot do.

Where the box on the table still wins

We are not trying to talk you out of a deck, and it would be silly to pretend the printed version has no advantages. It has several, and they are not small.

No screens, for one — and if either of you has a rule about phones in the bedroom, that rule is a good rule and a card deck respects it. It needs no battery, no account, no signal, and it survives being thrown in a bag on the way to a cabin. It works with more than two people without anyone thinking about it. And it is giftable in a way software is not: handing someone a physical box says something before it is ever opened.

If what you want is an object that makes two people sit down facing each other with nothing else in their hands, buy the deck. It is genuinely good at that. What we built is for the part that comes after — the fifteenth evening, and the fiftieth.

Keep exploring

A tool only matters if you actually reach for it more than twice. If you would rather borrow questions than buy a box, we keep a running set in questions to ask your partner. And if what you are trying to fix is less about tools and more about how far apart a few busy years can quietly put two people, that is its own subject — we wrote about it in reconnecting after years together.

Common questions

Is this just a card deck in an app?

Not quite. The cards are the surface, but the useful part is what sits underneath: what you have already seen drops out of the pool, the evening follows a rough arc instead of a shuffle, and you set how far it goes before you start rather than card by card.

Do we need to install anything?

No. It runs in any modern browser at private-game.app, and can be added to your phone home screen if you would rather it behaved like an app. There are iOS and Android builds too, but the browser version is the whole thing.

Can we keep it gentle?

Yes. You choose an intensity ceiling at the start of every session and it holds for that session. Keep it at the lower tiers and you get an evening of questions and small, warm prompts — nothing explicit appears unless you have deliberately raised the ceiling.

What if one of us is more enthusiastic about this than the other?

That is normal and it is worth naming rather than working around. Set the ceiling to where the less-keen person is comfortable, not to a compromise between you, and treat skipping as a free action — no explanation owed. A tool that only works when both people are equally eager is not much of a tool.

How long does an evening take?

A short session runs around fifteen cards and a longer one around twenty-two. In practice that is anywhere from twenty minutes to an hour, depending entirely on how much you talk between them — which is the point.

Private Game · How to play · Game tiers explained · Intimate couples card game · Questions to ask your partner · Gift guide for couples · FAQ