Disaster simulation and scenario planning constitute a critical practice to help understand the consequences of catastrophic events, and to be better prepared for them. Lucas Roebuck picked up some of my early work around the uses of SimCity in urban simulation, planning, and city culture to consider if "SimDisasters" could help us be prepared for the real thing." Roebuck argues that some of the things done at SimCity provide lessons that can be applied on the real world. However, like I proposed in "A City is not a Toy"(PDF 1MB) I believe that SimCity offers little value other than as cultural stimulation and interpretation, and that its uses as a simulator are not only limited but counter productive. Disaster simulation is progressing in much more useful and challenging fronts. SimCity does offer interesting insights but they are not related significantly to disaster preparedness.

It is true that, like Roebuck argues, in SimCity variables like fire department placement, resource funding, and zoning are significant aspects, including that densely constructed residential areas might be more likely to have fires spread rapidly from structure to structure, or that the pattern of flood zoning affects vulnerability to a hurricane. But already the way the game conducts itself around these hazards is strongly interwoven with the cultural setting it promotes, and relies heavily and logically in the playful nature of the premise. The game falls very short of showing actual consequences of disasters, how to actually rebuild after one of them, and lacks any real attention to detail in meaningful preparedness. For instance, the use of building styles is essentially based around aesthetics, and there is no way to match investment in new structures with more efficient and resilient building codes, where buildings might be more expensive to plan in the short term but over the long term are not only more efficient and sustainable but that provide better public services for the communities they house. Wouldn't that be an interesting mechanism to instill in the simulation of the game?

Fortunately there are many examples of more intriguing and useful developments in disaster modeling. For instance research developed by a couple of Spanish universities is contributing to the broad range of existing fire modeling tools that help predict and control flames, and offer critical information to fire departments. A combination of standard video imagery and infrared recording, is taken from several angles and then processed digitally to generate a 3d model that is then compared to a fire database to predict the most likely behavior of the fire under the live circumstances being recorded.

On a far more experimental side we can also find examples like the one developed by the geographer Paul Torrens at Arizona State University. His intriguing, troubling, and stereotypical work to model crowd behavior opens an interesting debate and brings us back to question back the premises of research and software development. While it is true that large multitude behavior is difficult to model, his assumptions in some instances are essentially wrong. For example, he assumes that "when trying to evacuate, people start to run and panic. Jams will occur and the evacuation doesn't proceed as efficiently as it might otherwise." This scenario used to develop a prototype to test crowd dynamics following outbreak of a fire in a dense part of a city with only a single point of evacuation chooses to ignore all the existing, extensive and robust research of the last fifty years, like Lee Clarke explains well. "People usually behave quite reasonably when their supper club catches fire or their plane crashes." In fact much of the psychological analysis points to the challenges that crowd inaction pose to disaster response. And when panic does occur, it is essential to understand the mechanisms, such as closed emergency doors or security guards using pepper spray on patrons during the evacuations, to try to analyze that crowd behavior. And yet, this modeling opportunity might be a far better entry point than SimCity to try to engage into a debate around ways to analyze human behavior in large urban settings. In fact, it may represent rather well what is the thinking behind trying to expand some of the shortcomings of SimCity into other modeling patterns, and the sociocultural bias behind them. It is particular concerning how these efforts aim to be translated quickly into notions of crowd control, curbing demonstrations, and herding communities to accommodate centralized top-down efforts.

On another recent contribution, Oswald Devisch also revisits some of my observations to further analyze the convenience of having planning professionals using computer games as learning tools. Devisch starts to talk about the limitations both of SimCity and SecondLife to expand on the idea that the games need to be opened, broken apart, and remashed by users to make the connections that would make a far more interesting learning environment. He talks of an environment that would still have tremendous limitations but also calls to recognize the great potential of collaborations, and having open conversations around shared simulations to expand the range of possibilities.

SimCity retains a tremendous interests as a cultural manifestation of city making, and of interpreting urban environments. It is in this cultural facet that I find SimCity a fascinating subject of study. I believe it is of importance to approach the thinking between its appropriation of urban dynamics and how it offers the player a framework to engage with that ecosystem, and to what extent it can be modified. However, much like it does with its approach to city making, while it can be an early playful offer to engage with planning, it does lack the capacity, intended or assigned, to be a reliable tool in education and modeling. That conclusion I believe is shared by its interpretation of disasters, that clearly belongs to a highly mediated idea of what constitutes a disaster, and how it evolves as interpreted by heavy handed cultural paradigms, and thus can not be used to model and learn about actual preparedness and response to disasters. SimCity may be able to help us expand the understanding of the cultural structuring of disasters, and while I believe that this is an important aspect of preparedness it is also an entirely different endeavor.

Images courtesy of SimCity/EA, Paul Torrens/ASU, and Grupo de Robótica, Visión y Control US/UPO.