How to make your own Game

Michael Douglas spends the first half of The Game trying to work out whether it's real. He spends the second half acting as though it is.

The switch happens without him noticing. That's the whole trick, and it's the only part worth stealing.

David Fincher made the film between Se7en and Fight Club, and almost nobody talks about it. Douglas plays a wealthy control freak whose brother buys him a mysterious "game" for his birthday. It takes over his life. The stakes escalate until they feel genuinely life-or-death. At some point he stops being an observer and becomes a participant, because the pressure leaves him no room to keep asking.

Once you're acting as if it's real, the distinction stops mattering.

Why "pretend this is real" never works

Every host has tried it. You set up the scenario, explain the rules, and then say some version of: treat this as though it were your actual budget, your actual client, your actual team.

Nobody does. Asking someone to believe something is the fastest way to make them notice they don't.

The film doesn't ask. It removes the option.

Which is the design problem. You can't instruct your way to immersion. Something in the structure has to make people forget to ask the question, and forgetting is a consequence of mechanics rather than framing.

Four things tend to do it.

1. Borrow the stakes, don't invent them

Points, badges and leaderboards are invented stakes. Everyone knows the points aren't real, which means the apparatus is asking for belief all over again.

Borrowed stakes already exist. Your peers are watching you decide. Your boss is on the same table. You committed to a number in front of eleven colleagues and everyone can see it.

Nobody has to pretend a leaderboard matters. They only have to look sideways.

So before adding a game layer, look for the pressure already present in the group and point at it.

2. Make them commit before they know enough

Douglas never has sufficient information. He acts anyway, because stopping isn't available.

This is the Blind Bid mechanic in Rondo. You submit privately, before you see what anyone else chose. There's no time to work out the socially safe answer, so what lands on screen is judgment rather than consensus.

The moment people commit, they have something at stake: their own position, in public, with their name on it.

Information after commitment feels like consequence. Information before commitment feels like a briefing.

3. Let them see something change

Meaningful Choices and Visible Consequences are Principles 2 and 3 in our internal Narrative Design playbook, codified by our LX design lead Fer Franco.

The visible half is the one that gets skipped. A choice with an invisible consequence is a survey.

What changes doesn't need to be elaborate. The other tables' answers appear. The scenario escalates because of the thing you underfunded. A character responds differently to the participant who was blunt with them.

4. Keep the frame quiet

The more you explain your design, the more people watch your design.

Every time you say "the reason we're doing this exercise is…", you hand the room permission to evaluate the exercise rather than play it. Douglas is never told the rules. He's told nothing, which is precisely why he can't stand outside it.

This is the hardest one, because explaining is what hosts are for. The discipline is holding it until the debrief, and then explaining everything, because the debrief is where the learning gets named.

A worked example

A session on prioritizing under pressure. Not a lecture about prioritizing.

Setup. Each table runs a business unit with one budget to split across four initiatives. Sixty seconds.

Commit. Everyone submits privately, before discussion.

Escalate. An Escalation Timer runs. Halfway through, one initiative fails publicly, and it's the one most tables underfunded.

Reveal. All allocations appear at once. The room can see who hedged, who went all in, and who is now quietly explaining themselves.

Debrief. Now you explain the frame, and only now. What did you assume about the timeline? What would have changed your allocation?

By that point nobody needs telling to treat it as though it were real. They already argued about it in public.

The obvious objection

Someone will say this is gamification with better lighting, and that most of it is exhausting. They'd be right about most of it.

The difference is where the stakes come from. A leaderboard is invented, and a room of people watching you decide is not.

The failure mode is worth naming too. Immersion has a ceiling, and past it people stop learning and start surviving. Fincher takes Douglas well beyond that line, which is fine in a thriller and not in a workshop. The line between immersive and overwhelming is real, and it moves depending on who is in the room.

Which leaves the question we keep coming back to

Not what do they need to know, but: what should they have to decide, in front of someone, before they're ready?

That tends to be where the design starts.