OxGame project in Cameroon

Submitted by Michael Rastall 4th October 2013 16:35

The OxGAME project is a 12-month pump-priming project funded by the OUP John Fell fund at the University of Oxford.

Our aim is to explore and evaluate the role computer models can play in anthropological research, particularly with respect to asking questions about the future of subsistence farming in Cameroon. This links to similar methodological approaches used in other research on the impacts of climate change on the forest communities of Cameroon.

Our approach is to use modelling4all project software called the Behaviour Composer to build simulations of farming so that farmers, agricultural experts and policy-makers can easily interact with the software. In particular we are interested in recording narratives as people make, run and explain simulations.

The original idea for the project came about because Professor David Zeitlyn wanted to pilot the use of Sim City as a way to discuss farming. This would have had the advantage of greatly improved visualisations (compared to NetLogo) but it would have been too difficult to work out how to adapt one of the open source version of Sim City e.g. Lin City. Also the Behaviour Composer is designed to help people build models by decomposing models into easy-to-understand components (an analogy might be Lego).

A second field trip is planned for February 2014 and we have a lot of work to do before we go:

Port an enhanced version of the NetLogo Somie model built in Cameroon during the first trip to the Behaviour Composer.
Setup and test running the BC using just a local wireless router and using a powerful laptop as a server (no Internet in remote locations!)
Write new logging features and trial the soon-to-be-released Review functionality which will let us record and replay simulation runs
Last, but by no means least, setup and design workshops with colleagues in Cameroon

With respect to item 4, the focus for the OxGAME project will be on capturing the narrative as different stakeholders build, run and explain models. It is worth noting that the models we are building are not particularly agent-based in nature. The decisions the farming agents make are largely made by the people as they interact with the model. That is to say, we are trying to build simulations that are very simple micro-worlds – something akin to Sim city or Civilization V but where the world resembles a real place in Cameroon.

An agent-based model is being created for the COBAM project and the fieldwork will address the complementary needs of the 2 projects as the teams will work together to apply a range of the tools with different stakeholders in Cameroon to explore which computer-based approach works best in different contexts.