Running a real session for 200 people or more

Ask a training provider what they can run for three hundred people and the answer usually falls into one of three shapes. A keynote with an activity bolted on, which is a talk. A small-group game run in twelve parallel rooms, which is twelve separate sessions that happen to share a building. Or a facilitator with a microphone asking three hundred people to turn to the person next to them, which is a lecture with a pause in it.

None of these is a simulation. In a simulation the decisions of one group change the conditions another group is playing in. That is the thing that produces the behaviour worth debriefing, and it is exactly the thing that breaks when you split the room into isolated tables.

Very few providers can put a genuine simulation in front of a large group. It is worth understanding why, because the reason tells you what to ask for.

Why most formats break at about forty people

Three things stop scaling at roughly the same point.

The facilitator can no longer see the room. A good facilitator is taking notes on behaviour — who spoke first, which team abandoned a plan, who was not consulted. At fifteen people that is possible. At forty it is already partial. At two hundred, one person at the front has no idea what happened at table nineteen, and the debrief becomes generic because the facilitator has nothing specific to point at.

The plenary conversation collapses. Above about forty participants the same four extroverts answer every question and everyone else becomes an audience. The debrief is where an experiential session earns its fee, and a debrief nobody contributes to is a summary.

The game's interactions become combinatorial. If every team must negotiate with every other team, twelve teams means sixty-six possible pairings. The session stops being about decisions and becomes about queuing.

A design built for a large room solves all three deliberately. One that was built for twelve people and then stretched solves none of them, and the symptoms show up on the day.

What a design that actually scales does differently

It adds villages, not queue length. The structure is built so that additional teams add negotiating surface rather than waiting time. A team always has someone to trade with, and no round depends on a specific pairing happening.

It pushes the debrief down to the table. The facilitator at the front runs the frame and the timing. The reflection happens at each table against a written prompt, with a nominated reporter, and two or three reports come back into plenary. It is less rich than a thirty-person circle and it is the only thing that works at scale. It also means the design has to make the pattern obvious enough to survive being reported second-hand, which is a real constraint on the game.

It uses the room's own noise as information. In a large-format session you can hear which corner has stopped talking, which is negotiating hard, which has given up. That is a feedback channel a small room does not have, and a design built for the format uses it.

It keeps the physical component count survivable. Three hundred participants is a lot of cards, tokens and boards to distribute, collect and reset. A format that scales has been designed around distribution, not just around play.

What it costs you to run, in practice

Large-format sessions are priced per session rather than per participant far more often than small ones, because the cost of adding teams is much lower than the cost of adding facilitators. That usually makes the cost per head fall sharply as the group grows — which is the opposite of what buyers expect, and worth knowing before you negotiate.

What does rise:

• Room. You need a flat floor, round tables, and enough space between them that teams cannot overhear each other. A tiered auditorium does not work: participants have to face each other, not the front.
• Set-up time. Expect a day on site beforehand for a three-hundred-person build. This is not padding; it is distribution and staging.
• Audio. One facilitator running three hundred people needs the room to hear a timing call instantly. Bad audio does more damage to a large session than anything else on this list.
• Facilitator count. Fewer than you would expect for a design built for scale, more than one. Ask where the thresholds are — a group just past one is the worst number to bring to a quote.

What the day actually looks like

A three-hundred-person session is a logistics exercise wrapped around a game, and the timings are not negotiable in the way a small session's are. A representative full-day shape:

The day before, on site. Tables laid out and numbered, materials distributed to each one, audio tested from the back corner. This is not preparation you can do on the morning with three hundred people arriving.

First thirty minutes: framing and roles. Teams form, roles are assigned inside each team, rules are explained once to the whole room. It is explained once because repeating it for latecomers costs the room twenty minutes and the facilitator their authority.

Rounds, ninety minutes to two hours. Timed, with hard stops called over the audio. Between rounds, a short market or negotiation window where teams leave their tables. This is the noisiest and most productive part of the day, and it is where a design built for scale earns its difference.

A surprise event, usually round three. Conditions change and plans in motion become invalid. At three hundred people this lands as a genuine shock, which small rooms rarely produce.

Scoring and the reveal, twenty minutes. Public, and specific enough that teams can see why they placed where they did.

Debrief, sixty to ninety minutes. Table-level first against a written prompt, then two or three reports into plenary, then the transfer question. Protect this slot. A session that overruns will always overrun into the debrief, because the game has a clock and the debrief does not — and a truncated game with a full debrief beats the reverse every time.

Half-day versions cut rounds, not debrief. Any provider proposing the opposite has misunderstood what they are selling.

The one question worth asking a provider

Ask what happens in round three if two teams refuse to trade with each other.

A design built for a large room has an answer, because it has been run enough times for that to have happened. A small-group game being stretched will either not have considered it, or will tell you the facilitator intervenes — which at three hundred people means it does not get handled at all.

Where HDX fits

Falcon Eye runs from fifteen to five hundred participants on the same design. Villages explore, trade, defend and negotiate across timed rounds, and adding villages adds negotiation rather than dead time. It has been run at conference scale and for single leadership teams, and the debrief structure changes between those two, which is the part most formats do not handle.

Where the group is distributed rather than in one room, Shift runs virtually for smaller cohorts and can be delivered in waves — a different answer to a different constraint, and an honest one: there is no five-hundred-person virtual version and we would not pretend otherwise.

Two things worth reading before you commission anything at this scale: how to debrief an experiential session, because the table-level debrief is where large sessions are won or lost, and when a physical game beats a digital one, because the format decision at three hundred people is not the same one you would make at thirty.

---

Sessions have run in seventeen countries since 2020, including conference-scale rooms. If you have a date, a number and a room, tell us what it needs to leave with.