Office Party Games
Icebreakers and team games for the rest of the year — away days, new starters, end-of-quarter drinks, and the fifteen minutes at the start of a workshop.
What to play with a team
Ordered from lowest commitment to highest. The first two work in a meeting; the last two want a proper slot.
1. This or ThatOffice questions
2+ players5–15 minOne phone
The five-minute option, and the one to reach for at the start of a meeting. Rapid-fire pairs answered on instinct — it takes no explanation, produces no losers, and warms a room up faster than asking everyone to share a fun fact about themselves.
2. Would You RatherOffice questions
2+ players10–30 minOne phone
The best general-purpose team icebreaker there is, because the structure forces participation without demanding disclosure. There is no correct answer so nobody can be wrong, and no way to opt out with "I don't mind" — everyone picks a side and explains, which is where you actually learn how a colleague thinks.
Not a game, but the thing that makes the rest of an away day work. Paste the attendee list, pick a number of teams, and it deals them out fairly in a couple of seconds — no captains, no picking, nobody chosen last in front of the room.
3–12 players10–20 minOne phone
The best game here for a team that already knows each other well. Everyone receives the same word except one person, and the group questions each other to find the impostor. It rewards paying attention to how your colleagues actually talk, which is a more interesting skill than trivia.
5. Most Likely ToWork questions
4–20 players10–20 minOne phone
Very funny with an established team and a poor choice for a new one, since every question depends on shared history. Use the work bank, which stays on workplace habits. If half the room joined this quarter, play Would You Rather instead and come back to this next year.
3–10 players20–40 minOne phone
The one to book a proper slot for — twenty to forty minutes, real scoring, and a genuine competitive shape. It is the best of these for a team-building session with an actual budget of time, and it rewards writing a convincing lie rather than knowing the answer, which levels the field completely.
7. Never Have I EverOffice questions
3–15 players15–30 minOne phone
Only with the office bank, which rewrites the game around unread inboxes and meetings that should have been emails. Set the rule that anyone can pass without explaining before you start, not after somebody has needed it.
The utility that quietly gets the most use: who presents first, which team goes next, whose turn it is to run standup, where the team lunch is. Paste the options in and let it decide.
Why most icebreakers fail
Almost everyone has sat through a bad icebreaker, and they fail in one of two identifiable ways. Either they demand disclosure — tell us something nobody here knows about you — which puts a person on the spot in front of colleagues with no graceful way out. Or they demand performance — a fun fact, a two-minute story — which advantages the extroverts who did not need the icebreaker and punishes everyone who did.
The ones that work do neither. They give everybody the same small, low-stakes decision to make at the same time, and then let the conversation come out of the disagreement rather than out of anybody's willingness to volunteer. A question with two answers and no correct one does this better than almost anything: you cannot be wrong, you cannot pass, and the interesting part is the reason rather than the answer.
It also matters that the game ends cleanly. A good icebreaker is over in ten minutes and everybody knows it is over. The ones people dread are the ones with no visible end, where the room slowly realises this is going all the way round the table.
Fifteen minutes at the start of a meeting
This is the most common real use, and it has tight constraints: it must need no setup, no explanation, and no equipment beyond the screen already being shared. It must stop dead when the meeting needs to start. And it must work for the people who joined the call two minutes late.
This or That fits exactly, because it is designed to be interrupted — every pair is independent, so stopping mid-game costs nothing. Would You Rather works when you have a little longer and want people talking rather than just answering. Both run from one shared screen with no per-person setup, which is what makes them practical for a hybrid room where half the team is dialling in.
Away days and longer sessions
A full away day is a different problem: you have hours rather than minutes, and the real risk is not awkwardness but fatigue. Nobody can do six hours of structured activity. The useful pattern is one long game with an actual competitive shape, placed after lunch when attention dips, and short warm-ups between sessions rather than more long ones.
Fake News Factory is the best fit for that long slot. It has scoring, it runs twenty to forty minutes, and it rewards writing a convincing lie rather than knowing the answer — so the person who wins is not the person who happens to know the most trivia, which is what makes it feel fair to a mixed team.
| Occasion | What to play | Length |
|---|---|---|
| Start of a weekly meeting | This or That | 5 minutes |
| New starter's first team call | Would You Rather | 10–15 minutes |
| Away day, after lunch | Fake News Factory | 20–40 minutes |
| Team that knows each other well | Odd One Out | 10–20 minutes |
| End-of-quarter drinks | Most Likely To | 10–20 minutes |
| Splitting into breakout groups | Team Generator | Seconds |
Splitting into teams without the playground problem
Letting two people pick teams is the traditional method and it is genuinely bad at work for the same reason it was bad at school: it is slow, it is a popularity contest, and somebody is chosen last in front of everyone. At work it has an extra cost, because the picking usually tracks seniority and existing cliques, which is the precise thing a team activity was supposed to loosen.
A random deal takes two seconds and removes all of it. Nobody was rejected, because nobody was choosing. If a split genuinely looks lopsided, shuffle again — but agree how many re-rolls are allowed before you start, or you have quietly reinvented picking teams by hand.
New starters and mixed-seniority rooms
Two situations need extra care. A room with new starters cannot play anything that depends on shared history — Most Likely To is very funny with a settled team and actively alienating when a third of the room joined last month, because they can neither answer nor be answered about. Play something history-free and come back to it once they have been there a while.
A room with a big seniority spread has the disclosure problem instead. People match the level of openness the most senior person sets, which means a manager answering a personal question candidly has not modelled vulnerability, they have raised the floor for everyone below them. The fix is not to ban managers from playing — it is to use question banks that are about work, where the ceiling on how personal an answer can get is built into the questions.
Remote and hybrid teams
These all work over a video call without modification, which is the main reason to prefer them to anything physical. One person shares their screen, reads the question, and the group answers into the call. There is nothing to post to anyone and nothing to install.
The one adaptation worth making is for hybrid rooms specifically, where some people are together and some are dialling in. The people in the room will answer faster and louder, and the remote attendees will progressively stop trying. Use the chat as the answer channel for everybody, including the people sitting next to each other — it feels artificial for about one question and then it works.
Common questions
- What are good icebreakers for a work team?
- Would You Rather with the office question bank is the most reliable: there is no correct answer, so nobody can be wrong, and no way to opt out with "I don't mind". This or That is the faster option when you only have five minutes. Both avoid the two things that make icebreakers fail — demanding personal disclosure, and demanding a performance.
- What can we play in fifteen minutes at the start of a meeting?
- This or That, which is built to be interrupted — every pair is independent, so stopping when the meeting starts costs nothing. It needs no setup or explanation and runs from whatever screen is already being shared.
- What works for an away day?
- One longer game with real scoring after lunch, when attention dips, plus short warm-ups between sessions. Fake News Factory suits the long slot: twenty to forty minutes, and it rewards writing a convincing lie rather than knowing trivia, so it feels fair across a mixed team.
- How should we split a team into groups?
- Randomly, with the Team Generator. Letting people pick teams is slow, tracks existing cliques and seniority, and leaves somebody chosen last in front of the room. A random deal takes two seconds and nobody was rejected because nobody was choosing.
- What should we avoid with new starters in the room?
- Anything that depends on shared history. Most Likely To is very funny with a settled team and alienating when a third of the room joined recently, because they can neither answer nor be answered about. Play something history-free and come back to it later in the year.
- Do these work for remote teams?
- Yes, with no modification — one person shares their screen and the group answers into the call. In a hybrid room, use the chat as the answer channel for everyone including the people sitting together, or the in-room half will answer faster and the remote half will stop trying.