The article shows how people design mechanisms in an economic experiment. They found that about 30% of participants' offers were optimal, while a similar percentage were not. Over time, the optimal offers became more common, following a specific pattern identified by previous research.