Almost every trivia format assumes a crowd. Teams, rounds, a host with a microphone, answer sheets swapped between tables. Take all of that and try to run it with two people and it collapses immediately, because the person reading the questions cannot play, and the person playing has nobody to confer with.
Which is a shame, because two is a good number for trivia. Two people can play on a couch, in a car, waiting for a table, at eleven at night when neither of you wants to start a show. The bar for a good two-player format is not that it scales down from a pub quiz. It is that it works on its own terms.
Six that do.
1. Alternating quizmaster
You ask ten, they ask ten, and you both keep score.
This is the obvious one and it is better than it sounds, provided you fix its one real flaw. The flaw is that the person asking gets to choose the difficulty, and the temptation to pick things you know your opponent does not know is enormous. If you both give in to that, the game becomes an exchange of obscure grievances and stops being fun by question six.
The fix is a source rule. Both of you pull from the same pool, a book or a deck or an app, and nobody hand-picks. You are the reader, not the author. Once the questions are outside your control, the asking half of the game becomes genuinely enjoyable, because you get to watch someone work through something you just learned thirty seconds ago.
Best for: long drives, where one of you is holding the phone anyway.
2. Simultaneous silent answer
Both of you see the question. Both of you write an answer down without saying anything. Reveal at the same time.
This is the format that most resembles what makes trivia good in a group, which is the moment where you find out someone else in the room thought the same weird thing you did. Speaking answers out loud destroys that, because the first person to talk contaminates the second. Writing preserves it.
It also handles the skill gap better than anything else here. When answers are simultaneous, a stronger player cannot dominate the airtime, and a weaker player is never answering in the shadow of someone else's confident guess. Both of you are wrong alone or right alone.
Best for: two people of noticeably different ability who still want a real contest.
3. The steal
One question at a time, alternating who gets first crack. Get it right and you score two. Miss it, and your opponent gets a shot at it for one point.
The steal is the mechanic that makes head to head trivia tense rather than parallel. In every other format on this list you are mostly playing against the questions. Here you are playing against a person, because their failure is your opportunity and you can see it happening in real time.
Add a wrinkle if you want it sharper: on a steal attempt, a wrong answer costs you a point. Now the second player has a decision instead of a free roll, and passing on a steal becomes a legitimate move.
Best for: competitive pairs. Do not run this format with someone who does not enjoy losing.
4. The category draft
Before you start, lay out eight or ten categories. Take turns picking. The catch is that you are picking categories for your opponent to answer.
Everything interesting about this format happens before a single question is read. You have to model what the other person knows, which means you have to admit what you think they are bad at, out loud, to their face. It is the most social two-player format on this list and it is very funny with someone you know well.
A gentler version has you drafting for yourself, which turns it into a handicap system: the stronger player drafts last, or gives up their first pick entirely. That version is worse as a game and better as a way to keep an uneven match close.
Best for: people who have played together enough to have read on each other.
5. Confidence bidding
Before each question is read, you both privately commit a number from one to five. That is what the question is worth to you, and only to you. Right answer scores it, wrong answer loses it.
This sounds fiddly and it is, for about four questions, and then it becomes the most interesting format here. You are no longer just answering. You are estimating how much you know before you know what you are being asked, which is a different and honestly harder skill. Categories help: announce that the next one is geography, then take the bids.
The scoring swings are large, so it fixes the runaway leader problem automatically. Someone down twelve points can bid five on the next three questions and be back in it, which is exactly the property you want from a two-player game where there is no round structure to hide behind.
Best for: a rematch, when the first game was not close.
6. The running season
You do not sit down together at all. You both answer the same set of questions whenever you get to them, and compare afterward.
This is the format that fits real life, and it is the one most people have never deliberately tried. Nobody has to be free at the same time. Nobody has to host. You answer during a lunch break, they answer on the train, and the result is waiting for one of you later.
What makes it work over weeks is a cumulative score. A single asynchronous round is a curiosity. A running total across a month is a rivalry, and a rivalry is the thing that actually gets two people to keep playing something. The score is doing the work that a shared room does in every other format on this list.
Best for: friends in different cities, or two people whose schedules never line up.
Making an uneven match worth playing
The problem with two players is that there is nowhere to hide. In a team of six, a gap between the best and worst player is absorbed by the group. At two, one person is simply winning, every time, and after about the fourth night the other person quietly stops suggesting it.
Handicaps help but they have to be structural rather than charitable. Nobody enjoys being spotted ten points. What works better is asymmetry that both players can see the logic of: the stronger player answers on a shorter clock, or gives up their category picks, or has to answer without the first letter hint that the other player gets. The result is still a real contest, and the weaker player's wins still count as wins, which is the whole thing you are protecting.
The other honest option is to stop playing against each other and start playing against the questions together. Same questions, one shared score, try to beat last week's total. It is a completely different game and it is the right call for a lot of pairs.
Keeping score when there are only two of you
Two-player scoring has the opposite problem from pub quiz scoring. In a room of eight tables, the danger is a runaway leader. At two, the danger is that the score stops meaning anything, because a single session is short and neither of you writes it down.
The fix is to make the unit of competition bigger than the session. Keep a running total on a note in your phone, or on the back of an envelope stuck to the fridge, and let it accumulate. Ten questions on a Tuesday is nothing. Ten questions on a Tuesday that puts you three ahead going into the weekend is a different object entirely.
Within a session, the mechanics from the pub quiz world mostly do transfer, just smaller. Escalating point values work: last five questions are worth double, which means a session is never decided at the halfway mark. A joker still works, though at two people it is less a strategic decision and more a small thrill. Wagering works extremely well, which is why confidence bidding earned its own entry above.
What does not transfer is partial credit, and you should not attempt it. Two people negotiating whether an answer was close enough, with nobody neutral in the room, is how a nice evening becomes a discussion about honesty. Right or wrong, decided by whatever source you are pulling from, no appeals. If you want the full version of the scoring argument, our piece on trivia night scoring systems works through it at pub quiz scale, and most of it shrinks down cleanly.
Where the questions come from
Every format above has the same dependency, which is a supply of questions neither of you has seen. That is a bigger problem than it looks. A trivia book runs out. A deck of cards gets memorized. Whoever does the sourcing ends up doing unpaid work, and when they stop, the game stops.
Rotating question pools are the only real solution, and it is why we built the app the way we did. Trivia goes stale faster than people expect, and a two-player game feels it first, because you are burning through the same pool twice as fast with half the people.
If you want the asynchronous format above without building the scaffolding yourself, Trend Trivia's challenge feature is exactly that: you send a friend the same set of questions, you both play when you have a minute, and the result is there when the second person finishes. If you want to sharpen up first, nine habits that actually stick is a better use of a week than grinding question sets, and writing your own questions is worth trying once, if only because it permanently changes how you read someone else's.
Pick one and actually run it
If you have never done this deliberately, start with simultaneous silent answer. It needs no equipment beyond something to write on, it survives a skill gap, and it is the fastest way to find out whether the two of you enjoy this at all.
Then, once you know you do, go straight to confidence bidding. It is the one that will still be interesting in six months.
Challenge one person, not a room
Send a friend the same set of questions and both play whenever you get to it. Free on iOS and Android.
