What to run when you have to break the ice and nobody wants you to

Why the fun fact fails

Almost every icebreaker asks the same thing in a different costume: say something about yourself, in turn, to a room of people you do not know. Two interesting facts, a fun fact, something nobody would guess. It is the same request.

That request is hard in a very specific way. You are being asked to perform being interesting, cold, with no context, while eleven people wait — and the people who find it easy are the ones who least needed an icebreaker. Everybody else spends the first three turns silently auditing their own life for something acceptable, and misses the first three answers entirely.

The games below get to the same place by a different route. Nobody is asked to be interesting; they are asked to have an opinion, or to guess, or to be obvious. The talking happens as a side effect, which is the only way it happens comfortably.

Start with a game where the group answers, not the individual

Most Likely To is the one to open with, every time. A prompt appears — most likely to survive a zombie apocalypse, most likely to argue with a self-checkout machine — and everybody votes for whoever it describes. You do not have to say anything, prepare anything, or reveal anything.

What it does is teach the room about itself very fast. A four-nil landslide tells you more about how a group sees somebody than an hour of fun facts, and the person it lands on almost always agrees, which is where the laugh comes from. Votes stay hidden until the reveal, so nobody follows the crowd.

Think Alike is the other one that works cold. Everyone answers the same prompt and the most common answer scores, so the skill is being predictable — the exact opposite of every other icebreaker. It levels a room in a way nothing else here does, because the person who is quick with a joke has no advantage at all.

Then something with a little more exposure

Once the room is warm you can ask for more. Two Truths and a Lie is the classic, and it works far better as the second or third game than as the first — by then people have something to draw on and are not reaching into a void.

Hot Seat is the sharpest of these and the one to be careful with. One person answers, and everyone else writes questions anonymously. The anonymity is what makes it work: a question you have to put your name to in front of new colleagues is a much blander question, and a blander question gets a blander answer.

It is also the one with a real off-ramp — every question has a Pass, and it costs nothing. Run it with a group that has met once and the passes will be frequent, which is fine. Run it with a group that has worked together for a year and nobody will use them.

What to do about the remote ones

A video call makes all of this worse, because the natural correction — reading the room, noticing somebody has gone quiet, letting two people talk at once — is exactly what a call removes.

The fix is to pick games where everybody acts at the same time rather than in turn. Most Likely To and Think Alike both do this: everyone answers simultaneously and the results land together, so there is no turn order to wait through and no awkward gap while somebody unmutes.

Avoid anything with a performance in it for a remote group. The quarter-second of latency turns a shared laugh into a scattered one, and a joke that lands late does not land.

Fifteen minutes is enough

The most common mistake is running an icebreaker for too long. The purpose is to get the room to a state where the actual thing can happen — the meeting, the party, the evening — and once it is there, the icebreaker is in the way.

Two short games is almost always right. One where nobody is exposed, one where somebody is a little, and then stop while people are still enjoying it.