Team building games for small groups

What changes when the group is small

A small group is roughly three to twelve people: one team, one table, one conversation. Four things about that size decide whether an activity works.

1. Everyone is visible, so the stakes go up. A game that is mildly awkward for one person in a room of thirty is brutal for one person in a room of five.
2. You cannot run teams against each other. Splitting six people into two threes mostly creates a grudge. Compete against the clock, a benchmark, or the group's own previous attempt.
3. One dominant person is not a minority. In a group of five, whoever talks most often produces half the words in the room. Structure that assumes a crowd will absorb them instead hands them the session.
4. The debrief finally has room to be good. With six people you can hear everyone properly in fifteen minutes, which is impossible at thirty. Everything below is material for that conversation.

A working rule for small groups: if the activity takes longer to explain than to play, it is the wrong activity.

Pick one fast

Under ten minutes, inside a normal meeting

1. Count to twenty

The group counts aloud to twenty. No order agreed, no pattern, no signalling. If two people speak at once, restart at one. Cap it at five minutes.

What it reveals: how this team handles airtime. Some groups are wrecked by two people diving for every number; others by everyone waiting so politely that nothing happens. Ask afterwards: who did you find yourself waiting for, and does that happen in our meetings? It depends on hearing overlaps, so it fails on video and is unfair to anyone with a hearing impairment.

2. Anonymous answers

Everyone writes an answer to one prompt on an identical slip. Fold, shuffle, redistribute, read aloud, guess the author. Prompts that work: work you are proudest of that nobody noticed, a skill that has never come up here, the thing you would fix first if you ran this team.

What it reveals: how well the group actually knows each other, which is usually less than assumed. Anonymity lets a quiet person be heard without having to hold the room. Ask afterwards: which answer surprised you, and what had you assumed instead?

3. Explain the job next to you

Each person takes ninety seconds to describe what the person on their left spends their week doing. The subject stays silent, then corrects the record in thirty seconds.

What it reveals: invisible work, usually the same person's every time. It also shows where people think a colleague's job stops, which is exactly where handovers fail. Ask afterwards: what does someone here do that we have all been quietly benefiting from?

4. The 15% round

From the Liberating Structures set. One question: with no more budget, no more headcount and nobody's permission, what could you change about how this team works? Two minutes of silent writing, then one contribution each before any discussion.

What it reveals: how much of the team's stuckness is genuinely external. Groups usually find a third of their complaints sit inside their own discretion. Ask afterwards: which of these starts on Monday, and who does it? Skip the silent writing and the fastest talker sets the agenda.

Twenty to sixty minutes, in a room

These run as one group with no breakouts, which is why they suit this size.

5. Paper plane factory

Three rounds of three minutes making paper aeroplanes that must fly three metres. Round one, the team organises itself. Round two, impose one fold per person and a compulsory pass. Round three, they redesign the process. Count only planes that fly.

What it reveals: nearly every group builds a pile of half-finished planes in round two and finds the defects at the end. Ask afterwards: where at work are we passing on nearly-finished things and discovering the defects last?

6. The decision autopsy

Take one real decision from three to six months ago, ideally one that went sideways but not catastrophically. Reconstruct it on a whiteboard: what we knew, what we assumed, who actually decided, who was in the room, who found out later. Then ask what we would do now.

What it reveals: the gap between how the team believes it decides and how it does. Ask afterwards: at what point could a different question have been asked, and what would it have cost? Run it only if you can resist defending your own past decision.

7. The empty chair

Put one empty chair at the table and name it: the customer, the night shift, the new joiner, the team downstream. During a real discussion, anyone may move to it and speak only as that party for a minute. Rotate.

What it reveals: whose interests go unrepresented here, and how thin the team's picture of them is. Ask afterwards: what did the chair say that we could not say in our own voices, and why?

8. The opinion line

A line across the room, strongly agree to strongly disagree. Read a statement with real tension in it for this team: we ship too slowly, we say yes to too much. People stand where they honestly sit. Interview three, then invite anyone who has moved to move.

What it reveals: whether anyone here changes their mind in public. In many teams nobody moves, which is a finding in itself. Ask afterwards: what would you need to hear to move one step? Keep statements about work rather than values, and offer a seated version so it is not a mobility test.

9. Ten-minute teach

One person a week teaches the group something they are good at that has nothing to do with the job. Knife skills, reading a balance sheet, how a heat pump works. Ten minutes, questions allowed, no slides.

What it reveals: competence people never get to show at work, and it changes how the group listens to them afterwards. It is also cheap soft skills practice, since teaching is where you find out whether you can explain anything. Keep it voluntary and let people pick their topic.

10. The two-minute hot seat

One person takes the seat and stays quiet. Everyone else names one specific thing they do well, with an example. Two minutes each. The subject says only thank you.

What it reveals: who is seen and for what. Across a few rounds the pattern is stark: one person praised for output, another for holding things together, someone getting vague compliments because nobody can name what they do. That gap is the finding. Warm but exposing, so run it only where there is reasonable psychological safety, let people pass, and not just after redundancies.

Small remote teams

A remote group of five needs no breakout rooms, which removes most of the complexity. It also removes the corridor, so structure has to do that work. Fuller list in virtual team building activities.

11. Asymmetric puzzle

One person shares a screen showing a puzzle. Everyone else holds the instructions and cannot see the screen. Only speech connects them, under a countdown. Bomb-defusal games work this way, but a spreadsheet with a hidden error and a printed rulebook does the same job free.

What it reveals: how precise this team's language is when it matters. Groups fail on vocabulary, not intelligence. Ask afterwards: which word cost us the most time, and where do we use that word loosely at work?

12. One driver, three navigators

Not a game. Take a real piece of work, a plan, a deck, a difficult email, and do it live together for forty-five minutes. One person has the keyboard and types only what the group agrees. Rotate every fifteen minutes.

What it reveals: more about how this team collaborates than any invented exercise, because the stakes are real and nobody is performing. It kills the standard remote failure where four people agree politely and then produce four different documents. Ask afterwards: what did we settle here that would have taken three email rounds?

13. The walking call

Pair people for twenty-five minutes, cameras off, headphones in, both walking outside. Three prompts so it does not stall: what has been harder than expected this month, what are you better at than a year ago, what would you like my help with.

What it reveals: depth a video grid suppresses. People say different things when they are not watching their own face. It is also the only item here that suits a team of three, where every group format collapses into a conversation anyway.

Longer formats worth the time

14. A co-operative board game, run as a diagnostic

Three to five people, sixty to ninety minutes, a game where the group wins or loses together. Hanabi is sharpest: you see everyone's cards except your own, so you depend entirely on the hints others give you. Pandemic suits groups that prefer talking to deducing.

What it reveals: whether people can act on someone else's judgement without re-deriving it. Watch for the player who starts directing everyone else's turns, and whether the group lets them. Ask afterwards: whose hints did you trust fastest, and what made them trustworthy? Say up front that you will debrief it, or it is just a nice evening.

15. The field visit

Take the whole team to where the work lands: a customer site, a shop floor, a support desk, a delivery route. Half a day, no laptops. Each person notes three things they did not expect, then the group compares lists over lunch.

What it reveals: how far the team's private models of the customer have drifted apart. This is experiential learning at its plainest, and it beats any constructed exercise for a group arguing from different assumptions. Ask afterwards: what did we see today that makes one of our priorities look wrong? The same principle drives outbound training at a larger scale.

What to avoid with a small group

• Anything that isolates one person. Avoid games where one person performs while the rest judge, and never make someone the object of a guessing game about themselves.
• Forced fun. Compulsory attendance plus required enthusiasm produces one visibly unhappy participant, and in a group of six that sinks the room. Offer an observer role and mean it.
• Physical contact games. The human knot, trust falls and three-legged anything exclude people with injuries, disabilities and different norms around touch, and here the exclusion is obvious to everyone.
• Personal disclosure dressed as bonding. Family, money, faith, politics, health, relationships. At this size, declining to answer is conspicuous.
• Rapid-fire formats. Quizzes and open brainstorms hand the session to whoever speaks quickest. Silent writing before discussion changes who gets heard more than anything else you can do.
• Using the game to dodge the conversation. If the real problem is one person's behaviour or an unresolved conflict, no activity fixes it, and running one signals that you know and are avoiding it.

How to debrief with a small group

Four questions, in order, and do not rescue the silences. What happened? How did that feel at the time? What does that remind you of in how we work? What will we do differently, specifically, starting when? Only the last one changes anything, and it is the one people run out of time for.

With six people you can ask every question of everyone. Go round rather than opening the floor, start with someone other than the most senior person, and if you are the manager, answer last. Good facilitation here is mostly restraint: the moment you explain the lesson it becomes your observation instead of theirs, and people defend themselves against observations they did not make. To feel a good debrief from the inside, play The Weekly Challenge with your team and compare answers.

Tour De Force runs this as live, experiential training for teams worldwide — online and in person. Talk to us, or play Gamified learning appsThe Weekly Challenge to see the method in ten minutes.