Policy and Complex Systems
Serious Gaming
Serious games can be designed to gain insights into policy problems . These types of games are a simulation of real world systems and events where players get the chance to make decisions about virtual events . The purpose of a serious game is to put actors in situations representing reality , in order to understand their decision process and then study the possible outcomes of the aggregation of those decisions . These situations can be virtual ( i . e ., computer games ) or real ( i . e ., a setting that would represent the real world situation ).
Games provide the possibility to learn on multiple levels . While the involved players may learn from the contextual information provided by the game or the decision making as it takes place during the game , useful material is gathered by the designers of the game to solve the underlying policy problem ( Raybourn and Waern 2004 ). Serious games do not necessarily have to be computer-based ; they could also be role-plays among human players .
While serious gaming ( SG ) reflects a great part of real actor decision making , compared to other models , the population that can be considered for a game is normally much smaller than the real population and therefore unreliable to extrapolate to real world scenarios .
Agent-based Modeling
Agent-based modelling ( ABM ) is a computer-based modeling approach that enables the exploration of the consequences of complex assumptions ( Janssen and Ostrom 2006 ). In ABM , models are inherently bottom-up and decentralized . Therefore , ABM describes those situations where the standard methods of predictive policy analysis are least effective ( Moss 2002 ).
With ABM , it is possible to design irrational agents with incomplete information in relatively uncertain situations .
The main advantage of ABM over other modeling approaches is that it captures emergence , linking individual behavior to system level behaviors . This results in a natural representation of a system ’ s global behavior as well as adding more flexibility to possible outcomes ( Bonabeau 2002 ). However , since ABM is a bottom-up approach to problem solving and the global behaviors of the system are emergent outcomes rather than being implemented into the system ( Epstein 2006 ), techniques for representing policies as top-down structures into the simulation are neither common nor straightforward .
Functionality of the Tools for Policy Analysis
The tools introduced in this section
are all used for policy analysis . In fact , in most cases the policy analyst uses a selection of these tools along with the non-computational methods . Therefore , highlighting which part of the policy analysis cycle each tool supports will be informative for choosing an effective combination of the tools . Table 1 shows where in the policy analysis cycle the aforementioned tools can be helpful . The letters in the second column show which requirement of Section 3 is addressed .
NEM can be used to parameterize the policies and make an association between different dimensions using equations . NEM does not support the identification of the resources , processes , characteristics , and boundaries of the system . However , once they are detected , it also helps parameterize these attributes for problem attributes . This also holds for the evaluation measures and policy definition . A limita-
68