The problem
The bullwhip effect is easy to define and hard to understand. You can reproduce the curve in an exam without ever having felt why it happens.
The classic Beer Game solves that, but needs four people, a facilitator and an hour. So it happens rather rarely more than once across a whole degree.
Approach
One human station, three run by the computer. The player makes the same decision as in the real game, the order quantity each period, but without having to gather three other people first.
What it does
After a season the debrief shows a bullwhip chart per echelon, a base-stock benchmark run on exactly the same demand, and coaching tied to the player’s own decisions rather than to general advice.
Everything runs in the browser. No account, no server, no data leaving the machine.
Outcome
The game is live and freely available. It became the first of the games that have since grown into Operations Lab.
What I learned
That the debrief teaches more than the game. Being compared with an optimal policy on exactly the same demand is uncomfortable in the right way, and that is where the understanding sticks.