Instructor guide

Run the Beer Game with your class

The Beer Game has been a staple of operations and systems-thinking courses for sixty years — because no lecture on the bullwhip effect lands like causing one yourself. This guide covers everything you need to run it online: no boards, no chips, no spreadsheets, and no student accounts.

How many students do you need?

Each game is one supply chain with four roles — retailer, wholesaler, distributor, factory. A session runs with 2 to 4 human players; the host assigns an AI player to any seat left empty, so an odd-sized group is never a problem.

  • Small group (2–4): One session, one role each. The purest version of the exercise.
  • Seminar (5–12): Pair students on roles — two or three per screen, deciding each order together. The negotiation within each pair is its own lesson in forecasting under uncertainty.
  • Large class (12+): Run several sessions in parallel and let chains compete for the lowest total cost. Comparing results across chains during the debrief shows how consistent the bullwhip pattern is.

For asynchronous or flipped settings, assign the 10-minute solo game as homework and spend class time entirely on the debrief.

How long does a session take?

Budget 60–90 minutes for the full experience. The debrief is not the wrap-up — it is the point. Protect it even if you have to shorten the game (hosts can set the length anywhere from 10 to 100 weeks; 24+ weeks shows the full boom-and-bust cycle).

10 min

Brief the class

Explain the four roles, the weekly rhythm, the two costs, and the no-communication rule. Don't reveal the demand pattern — the surprise is the lesson.

5 min

Get everyone in

Create a session, share the invite link, assign roles (or let players pick), and fill any empty seats with AI.

30–45 min

Play

The game advances week by week as every role submits its order. Resist the urge to rescue struggling teams — the pain is the data.

25–30 min

Debrief

This is where the learning actually happens. Walk through the analytics, reveal the demand pattern, and work the questions below.

How does classroom setup work?

Students never need accounts. The whole flow runs in the browser:

  1. 1

    Create the session.

    Go to /sessions/new, name it, choose the game length, and optionally add a join password. Sign in with a free En Dash account to keep your sessions and results.

  2. 2

    Share the invite link.

    The lobby gives you a link (and code) to project or post in chat. Students open it on any laptop, tablet, or phone.

  3. 3

    Assign roles.

    Assign roles yourself from the lobby or let players choose. Fill any empty seat with an AI player. The game can start once all four roles are covered.

  4. 4

    Set the experience.

    Host settings control supply chain visibility (full, limited, or hidden), an onboarding tutorial, and in-game tips. For the classic experience, use limited visibility and remind players: no talking about orders.

  5. 5

    Play, then reveal.

    After the final week, walk the class through the analytics — order amplification, inventory and backlog time series, cost by role, and a week-by-week replay with full visibility.

One physical tip: seat the four roles apart, or the no-communication rule dissolves on its own. If students haven't seen the rules, point them at the player guide beforehand — or don't, and let the confusion be part of the data.

What should you ask in the debrief?

These ten questions, in roughly this order, take a group from blaming each other to seeing the system. Each comes with the point it exists to surface — the sequence matters more than the wording. The bullwhip effect page makes a good pre-read for you (and a post-read for them).

1.Who or what do you blame for your costs?

WhyOpen here, before any reveal. Players almost always blame a person — the erratic wholesaler, the slow factory. The rest of the debrief dismantles this: the cause was the system's structure, not anyone's competence. Bank the answers so you can revisit them at the end.

2.What did you think customer demand was doing?

WhyCollect guesses from every role before revealing the truth. Upstream players typically describe wild, volatile swings. The gap between their perception and reality is the demand-signal distortion at the heart of the bullwhip effect.

3.The reveal: how many times did customer demand actually change?

WhyOnce. From 4 cases to 8 in week 5, then flat for the rest of the game. This is the single most powerful moment of the session — pair it with the order-amplification chart so players see the tidal wave they built from one modest step.

4.Why did you order more than you needed?

WhySurfaces the panic-ordering mechanics: backlog fear, invisible pipeline inventory, and distrust of the supplier. Most players double-ordered while their earlier orders were still in transit — and can now see exactly when in the replay.

5.What one piece of information would have changed your decisions most?

WhyPlayers converge on the answer fast: real customer demand, visible to everyone. That is precisely the argument for point-of-sale data sharing and collaborative forecasting in real supply chains — they invent the remedy themselves.

6.Which role paid the most, and why is that not a coincidence?

WhyPull up the cost breakdown by role. Costs generally climb as you move upstream, with the factory worst off — furthest from the demand signal, last to feel every wave. Position in the chain, not player skill, largely determined the outcome.

7.If you replayed with the exact same rules, what would you do differently?

WhyDistinguishes strategy fixes (track your pipeline, order steadily) from structural fixes (share information, shorten delays). Both matter — but players discover that better personal discipline only shrinks the whip, it doesn't remove it.

8.What would change if the four of you could talk?

WhyCommunication is banned during play, which mirrors how real supply chain partners often operate — through purchase orders alone. This question opens the discussion of collaboration, trust, and why integration efforts like CPFR and VMI exist.

9.Where have you seen this pattern in real life?

WhyTransfer. Pandemic-era shortages, semiconductor boom-and-bust cycles, restocking chaos in retail — once players have caused a bullwhip themselves, they start recognizing it everywhere. This anchors the simulation to their own industry.

10.Everyone acted rationally. So why did the system fail?

WhyThe closing synthesis, and the game's deepest lesson: local rationality plus delays plus limited information produces global dysfunction. Return to the blame answers from question one — the villain was the structure all along. That is systems thinking in one sentence.

Instructor questions

Do my students need accounts to play the Beer Game?+
No. Students join a hosted session through an invite link or play solo at /play/guest with no sign-up. Only the host benefits from a free En Dash account, which adds saved progress and session management.
How many students do I need per game?+
Each game is one four-role supply chain. It works with 2 to 4 human players per chain — AI fills any unassigned role — and two or three students can share a role at one screen and decide together. For larger classes, run several sessions in parallel.
How long does a classroom Beer Game session take?+
Plan 60 to 90 minutes: about 10 minutes of briefing, 30 to 45 minutes of play for a 36-week game, and 25 to 30 minutes of debrief. Hosts can shorten the game (down to 10 weeks) for tighter slots, though 24 or more weeks shows the full oscillation best.
Does it cost anything?+
No. The game is free to play and free to host.
Can I control what students see during the game?+
Yes. The host can set supply chain visibility to full, limited (each player sees only their own role's data), or hidden, and can toggle onboarding tutorials and in-game tips. Limited visibility is the classic experience.

Your next class could cause its own bullwhip.

Hosting runs on a free En Dash account — it keeps your sessions and results ready for the debrief. Sign in or create one, and your first session takes about a minute. Or play a quick solo game first to see what your students will see.