Cheating in Online Games: A Social Network Perspective
http://www.cse.usf.edu/dsg/data/publications/papers/cheating_TOI.pdf
In this article, researchers study the social networks and behaviors of cheaters vs non-cheaters through the Steam gaming community. Their findings are interesting: firstly, that the positionality of cheating players is not markedly different from non-cheating players (they do not occupy some strange relationship with players in the graph), and secondly, that abundance and geographic location of cheaters does not correlate well with maps of human population density or popularity of Steam. Another interesting finding of theirs is that cheaters (cheating behavior) are inherently tied to social experiences. Cheaters are less likely to own single player games than non-cheaters, and more likely to own only multiplayer-only games.
This article relates closely to Chapter 19 from the book, since they measured the behavior of cheating as it passed from person-to-person through social mechanisms. Rather than a simple q-threshold (which likely isn’t practical for modeling many social behaviors) the uptake of cheating behavior was seen to correlate strongly with the number of friends who cheated. This is similar to our assignment where we found a q-threshold that was variable: the value that the behavior had to the person was based not only on the fraction of friendly nodes who had already adopted it, but also on the total number of friendly nodes that had adopted it (behavior is more attractive if you have 10 friends that are doing it out of 20, than if you had 3 friends that were doing it out of 5). Information about how social trends such as cheating flow across networks helps us to understand preventative measures that can be taken to prevent them, such as, in this case, isolating cheaters from playing multiplayer games or from acquiring more friends.