Currently we have a main experiment called ‘Simple’ with a couple of variations.
This is a so-called price-offer market setup, where players are firms with capacity constraints, selling a single product to an unknown virtual demand (the aggregate demand is fixed), and the firms’s sole decision is to syncronously set the price for the next time step. In this experiment, players compete in order to capture the demand of a buyer played by the computer (fixed and unknown).
This is also known as a Bertrand-Edgeworth setup, where equilibrium prices are not possible. In other words, in the competitive environment it is not possible for firms to decide on a constant single price which maximizes their profit.
Frankly, this was in itself a discovery for us. Until we came across this problem, we were used to think that prices outght to reach an equilibrium price, and the reason it does not reach this state, would be due to uncertainty in the market players.
Our contention is that simple behavior can set up reasonable prices in a market, and that this behavior can be characterized by very few parameters.
For more detail instructions on the setup of the experiments, please refer here.
To read more detailed explanation of the ‘Simple’ experiment, read from this pdf.