On the Problem of Predicting Real World Characteristics from Virtual Worlds
Muhammad Aurangzeb Ahmad, Cuihua Shen, Jaideep Srivastava and Noshir Contractor
Predicting Real World Behaviors from Virtual World Data,
Muhammad Aurangzeb Ahmad, Cuihua Shen, Jaideep Srivastava, Noshir Contractor Editors,
Abstract Availability of massive amounts of data about the social and behavioral characteristics of a large subset of the population opens up new possibilities that allow researchers to not only observe people’s behaviors in a natural, rather than artificial, environment but also conduct predictive modeling of those behaviors and characteristics. Thus an emerging area of study is the prediction of real world characteristics and behaviors of people in the offline or “real” world based on their behaviors in the online virtual worlds. We explore the challenges and opportunities in the emerging field of prediction of real world characteristics based on people’s virtual world characteristics, i.e., what are the major paradigms in this field, what are the limitations in current predictive models, limitations in terms of generalizability, etc. Lastly, we also address the future challenges and avenues of research in this area.
They describe the ideas behind the “Mapping Principle,” and that this mapping between virtual and real behaviours cannot necessarily be assume to hold for all cases.
Representations in different worlds may change the results; not much of a problem for me due to lack of need for non-representative avatars, but may impact on fidelity of representation arguments for identification with avatars.
Configuration of the world influences the nature of the mapping, as the pvp and pve tasks may change levels of aggression or cooperation in the VW or game.
They cite the “Proteus Effect,” behaviours are affected by the appearance and intended nature of the avatar. Could this be used to empathise with your boss or underling in an organisation in BPM?!? Could be exploited in a training manner. Maybe it brings out perceptions of what another job entails, so that perceptions of roles can be tested.
They bring out the difference in using data to drive the notion of theorising about behaviour; reduces reliance on a priori concepts about behaviours.
They suggest to augment data logging with surveys, in order to determine other states outside of the game to be included in the modelling; straight forward but has to be said. Remotely logging people does not provide context with regards to their behaviours.
Generalisation across virtual worlds is still an unaddressed problem.
They note that people may lie when answering surveys, which can be uncovered by log data, eg. time spent playing games. But this is the same for every scenario, one imagines, especially with social repercussions.
They suggest actually using the worlds before testing them; in some ways a classic case of simulating your experiment before you actually carry it out. They suggest a minimum time of world use before the research can be published, and suggest it should be enforced at conferences at in journals.
Economics – virtual world economic patterns commute to real world patterns eg. Castronova. A black market exists for virtual money.
Epidemiology – citation of the Corrupted Blood incident in WOW. Spread of disease was similar to human real viruses, including remote location commencement and travelling escaping people bringing the plague to the other regions. They note that the cost of infection is trivial compared to real life, thus behaviours are different; bringing up a rationale cost benefit analysis model as a basis for action.
Deviant Clandestine Behaviours – this has a strong mapping due to social factors in being found out so to speak. Gold farming is noted as a major example in the virtual world cases. Some behaviours, such as thinning of social networks, have been found to be similar in virtual world and in real world gang examples.
Mentoring – mentoring networks are different from other social networks. But there is a limited form of mapping to the real world.
They do note that the online worlds allow the testing of hypotheses that are note possible in the real world, but they may not be applicable fully to the real world, as noted in other case studies.
They finish with an interesting comparison of a theory based approach to virtual world behaviour predictive modelling, or a data driven, bottom up modelling approach, based upon data mining for correlations to particular behaviours.
They note that the mapping principle is the overarching assumption in this work, but caution is urged due to the often lack of generalisation between the mapping found in different virtual worlds.
15. Hall,M.,Frank,E.,Holmes,G.,Pfahringer,B.,Reutemann,P.,Witten,I.H.:TheWEKAdata mining software: an update. ACM SIGKDD Explor. Newsl. 11(1), 10–18 (2009)
16. Johnson, N.F., Xu, C., Zhao, Z., Ducheneaut, N., Yee, N., Tita, G., Hui, P.M.: Human group formation in online guilds and offline gangs driven by a common team dynamic. Phys. Rev. E. 79(6), 066117 (2009)