Problem solving activities for teams

What makes any of these work

The choice of activity matters less than people think. A team learns more from a collapsed paper tower than from an expensive escape room, if four things are true.

• There is a debrief. Without one you have run a game. Facilitation separates a group that enjoyed itself from one that noticed something.
• Taking part is voluntary. Say people may observe instead, and mean it. One person made to perform costs more trust than the exercise builds.
• The constraint resembles their real one: too little information, too little time, two teams who need each other.
• Failure is possible. If everyone succeeds, nothing surfaces. Aim for about half finishing.

The debrief in three questions

1. What happened? Facts only. Who did what, in what order.
2. So what? What does that pattern look like on a normal Tuesday here?
3. Now what? One thing the team does differently, named and owned.

With 45 minutes, run a 20 minute activity and give 25 to the debrief. Not the other way round.

Pick one fast

Quick openers, under 20 minutes

1. Wrong Answers Only

Give the team a real problem. Five minutes to generate the worst possible solutions, competing to be more terrible. Then flip each one: what would make it useful? Groups of four to six. Finish the flip, or you have run a vent.

• Reveals: what people are genuinely worried about. Absurdity gives cover, so risks surface that a serious meeting would suppress.
• Ask afterwards: "Which of these worst ideas is closest to something we already do?"

2. Wisdom of the Crowd

Ask an estimation question with a checkable answer. Everyone writes a private guess; show the average against the spread and the truth. Round two: let them discuss it first.

• Reveals: how well independent estimates hold up, and how fast people drop their own number once someone sounds certain.
• Ask afterwards: "Who changed their answer, and what convinced you: the reasoning or the confidence?"

3. Blind Drawing

Pairs sit back to back. One describes an abstract diagram, the other draws it. Round one, no questions. Round two, new image, questions allowed. Compare drawings and times.

• Reveals: the gap between broadcasting information and confirming it landed. Round two is slower and far more accurate. That is active listening, demonstrated.
• Ask afterwards: "What did round two cost, and what did it buy?"
• Watch out: offer a blocks version or an observer role to anyone this excludes.

4. The Marshmallow Challenge

Teams of four get twenty sticks of spaghetti, a metre each of tape and string, and a marshmallow. Eighteen minutes to build the tallest freestanding tower with the marshmallow on top. It is well known, so ask first.

• Reveals: teams that build small and test early beat teams that plan and assemble once. The marshmallow's weight is the assumption nobody checks until the end.
• Ask afterwards: "When did the marshmallow first go on? What changes if you do that in minute two?"

Longer challenges, thirty to ninety minutes

5. Lost at Sea

Groups of five to seven rank fifteen salvage items by survival value, alone and then by consensus, no voting or averaging. Score both against the expert ranking, and compare each group to its own best member, not to other groups.

• Reveals: whether the group beats its own best individual. Often it does not, and the reason is always that somebody knew and was not heard.
• Ask afterwards: "Whose ranking was closest? Did the group listen, and if not, what stopped it?"

6. The Minefield

Scatter objects across an open floor. One person crosses blindfolded, guided by a partner's voice from the edge, no touching. Round two: three pairs cross at once, so the instructions compete.

• Reveals: what makes an instruction usable. "Your left" not "left", direction plus distance, a confirming pause.
• Ask afterwards: "What made an instruction usable rather than loud? Where does our work sound like round two?"
• Watch out: blindfolds do not suit everyone. Offer closed eyes, a spotter, or the observer job.

7. The Egg Drop

Teams of four to six protect a raw egg from a three metre drop, buying materials from a price list on a fixed budget. Forty minutes, then everyone drops. Cheapest survivor wins.

• Reveals: risk appetite. The budget forces a trade-off between buying a guarantee and running an experiment.
• Ask afterwards: "Did anyone test before the real drop? If not, was it time, or not wanting to fail in front of the room?" The second answer is a psychological safety issue.

8. Escape Room or Puzzle Box

Four to eight people, a sequence of locks, sixty minutes. A commercial venue works, and so does a padlock box in a meeting room, where you control the clues.

• Reveals: how information moves under pressure. Teams rarely fail on difficulty; they fail because someone held a clue and nobody asked.
• Ask afterwards: "What did you know that nobody else did, and why did it not get out?"
• Watch out: venues sell adrenaline and will not debrief for you. Book a room afterwards.

9. The Bridge Build

Two teams in separate rooms each build half a bridge to one written spec. They may send notes or a delegate but cannot see each other's work. At the end the halves are joined.

• Reveals: everything about handoffs. The join usually fails on a detail both sides believed the other had confirmed.
• Ask afterwards: "Which assumption did each side think the other had checked? Where does that gap sit between our departments?"
• Watch out: the failure lands hard. Debrief it as a systems problem, or you have made a grudge.

Structured methods for real problems

These need no props and work on the thing the team is actually stuck on, which makes them likeliest to outlast a busy week.

10. The Pre-mortem

Take a live plan. Tell the room it is six months later and the project failed badly. Everyone writes privately why, before anyone speaks, or the first opinion sets the range. Cluster the reasons, give the top three an owner.

• Reveals: objections nobody raises normally, because doubting a plan sounds like disloyalty. The framing removes that cost.
• Ask afterwards: "Which of these did someone already suspect, and what made it hard to say?"

11. Brainwriting 6-3-5

Six people write three ideas each in five minutes. Sheets pass left and the next person builds on what is there. Six rounds, half an hour, over a hundred ideas. Volume, not judgement.

• Reveals: how much output discussion was suppressing. Quiet people produce as much as loud ones, because writing has no queue.
• Ask afterwards: "Which of the strongest ideas began as somebody else's half-finished one?"

12. Six Thinking Hats

The group thinks in one mode at a time: facts, then feelings, risks, benefits, alternatives, and a review of the process. Five to ten minutes each, timed strictly or it becomes normal debate.

• Reveals: habitual roles. The standing sceptic argues the benefits, the optimist names the risks, and both find it hard.
• Ask afterwards: "Which mode was hardest for you? Is that the one you avoid at work, and what does the team lose for it?"

Remote and hybrid

Remote problem solving is not the in-person version on video. Nobody sees a shared workspace unless someone builds one. More in our virtual team building guide.

13. Remote Clue Split

Build a puzzle needing five pieces of information and send each person one piece before the call. Breakout rooms of four to six, fifty minutes, a shared document available but never mentioned.

• Reveals: whether a distributed team builds shared visibility or just talks louder. Most narrate at each other for fifteen minutes before anyone opens the doc.
• Ask afterwards: "How long until someone made a shared list? What changes if that happens first?"
• Watch out: in hybrid, the people in the room have an extra channel. Everyone dials in separately.

14. Desk Object Build

Give a brief, such as "a way to stop handover meetings overrunning". Three minutes, deliberately too short, to build a model of the answer from whatever is on the desk. Twenty seconds each to explain it, then merge the best three ideas.

• Reveals: how a hard constraint accelerates thinking. It gets cameras on without asking anyone to disclose anything personal.
• Ask afterwards: "Which feature survived the merge, and what made that argument land?"

What to avoid

• Forced fun. An opt-out that costs social capital is not one. Offer a real job: timekeeper, observer, scorer.
• Anything that exposes people personally. No confessions about family, money, health or beliefs.
• Formats that reward the already-dominant voice. Open brainstorms go to whoever speaks first and fastest. Write before you discuss.
• Unchecked physical and sensory assumptions. Blindfolds, running, standing, small print, noise. Ask privately in advance and design an equal role, not an exemption.
• Prizes worth winning. A real prize makes the exercise worth cheating for, and teams stop telling you what really happened.
• Treating one afternoon as a culture fix. An activity shows a team its pattern. Changing it happens afterwards, in how meetings run.

These work because a short challenge compresses months of team behaviour into forty minutes and makes it visible. Who takes charge, who withholds, whose objection gets heard: it is the pattern the team already runs at work. That is experiential learning, and why the debrief beats the game.

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.