Before continuing with the algorithms, we decided to do some further investigation into the variations in the market structures in which the different traders were developed. We reasoned that a greater understanding of the conditions for which the agents were designed could yield important insight into their possibilities in other market structures. The structures we considered were:
The Gode-Sunder ZIC Environment
The Third Week: May 29 - June 2
At the beginning of the third week, we got a revised version of the simulator code. After getting acquainted with the changes, I ran a series of experiments on the ZIC, IBM, and three forms of the GD algorithms. The next step with the code would be to transcribe the additional algorithms (ZIP, Pvt, RAH) from the old code into the new code.
The Cliff-Bruten ZIP-CLIFF Environment
The IBM-ZIP Environment
The Gjerstad-Dickhaut Environment
The Priest-van Tol Environment
The Santa Fe Institute Competition Environment