Designed experiments are crucial for understanding cause-and-effect relations: humans can grasp linear relations for one factor at a time (“the wetter the road, the longer it takes to bring the car to full stop”) but struggle with interactions (road humidity and temperature combined can form ice) and even more with non-linear effects. Typical production and service processes have tens of potential causes and multiple interactions at play. Not even experts can “guess” what factors truly call the shots and how.
At the same time, designing and executing experiments demands a significant level of diligence and know-how. As machines are interrupted, multiple versions of a website set up or customers served in different ways, experiments also tend to be costly. They must be “first time right”.
Interestingly, before going for “the real thing”, teams can prepare in ludic ways – which is where simulated designed experiments come into play: a facilitator sets up and knows the situation; participants then develop strategies to figure out and predict the system behavior.
We explain how that can be done to save time and money – and to have fun together as a team.