Game theory explains the spread of trends better than the epidemic theory
The ‘epidemic’ theory is typically used to explain the spread of a trend – stating that increased exposure to a trend will mean that more people will adopt it. When a trend hits a node that has many connections, then the trend will explode, or ‘go viral’. However, theoretical research in game theory has shown that this might be an oversimplification of the ‘viral’ phenomenon – just because someone is exposed to the trend, does not necessarily mean that they will adopt it and pass it on through their links and connections.
Researchers have created a game where a player bases his decision (whether to adopt or not) on the decisions of those around him/her. For example, with Facebook, you would look to see how many people around you were already on Facebook. If the number was high, then the chances of you joining would also be high, and if the number was low, then the chances of you joining would be lower. The game is set up so that you have a higher payoff by doing what others do as opposed to going against the grain, where you would have a lower payoff.
However, researchers pointed out that in a perfect world, no one would ever adopt anything new as changing would never be the best strategy if no one else was changing. To make this more applicable to real life, the game can be modified so that some nodes did not have all the information, and so would be unaware of what his/her neighbors did and go against the grain. This also accounts for those people in society who do what they want, regardless of what is more popular within their neighbors. When the structure was played around with, they found out that sometimes, nodes with many connections actually slowed the spread of the trend down – the main difference between the game theory approach to the epidemic approach. This was because nodes that were highly connected are under more influence and pressure from their neighbors. If some of their neighbors were not adopting the trend, the node would have to think independently of its neighbor’s actions if he/she wanted to adopt the trend and play its weaker strategy. On the other hand, well-connected nodes that had neighbors who adopted the trend would then adopt the trend because it would be his/her best strategy.
This game theory approach would apply mostly to websites or trends that require some sort of loyalty or commitment. Youtube videos or blog-posts may only need exposure in order to spread, but deciding between social networks or operating systems (eg. Mac versus PC, AT&T versus Verizon – where communication or transferring information may be easier if you had the same system as your neighbors) are largely dependent on game theory and the actions of your neighbors to which you are connected.
Source –
http://arstechnica.com/science/news/2010/11/game-theory-explains-why-some content-goes-viral-on-reddit-digg.ars