One of the worst defects of urban games is the all-powerful role they
reserve to the user: the player usually creates a city from nowhere, d
esigns, like a god, the complete city map, and makes alone all the imp
ortant choices on land control and land use, population flows, financi
al policies, etc. In reality, almost all important decisions concernin
g city management are taken after long and often protracted negotiatio
ns between protagonists. This is the very basis of democratic city rul
e and corresponds fully to the importance of local administration in m
odern industrial countries. Software trying to simulate the evolution
of a city in order to give new insights in the task and to understand
better how local administration works can hardly avoid simulating such
negotiations. Another important goal of urban simulation games is to
demonstrate to the user how a city reacts to external impacts: with a
didactic goal in mind, urban softwares need more than the others to mi
mic the processes and the conclusion of a negotiation. This paper disc
usses the conditions a program simulating negotiations must meet and p
roposes such a software called Negoce. (C) 1997 Elsevier Science Ltd.