| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | |||||
| 3 | 4 |
5 | 6 | 7 | 8 | 9 |
| 10 | 11 | 12 | 13 | 14 | 15 | 16 |
| 17 | 18 | 19 | 20 | 21 | 22 | 23 |
| 24 | 25 | 26 | 27 | 28 | 29 | 30 |
| 31 |
« Information, networks, and markets | Main | How networks are used to fight the overt drug market in High Point, NC »
23 January 2008
Right now I’m at Sunbelt 2008 in Florida, and there have been some very interesting talks on the first day.
Johannes Putzke (pictured above) presented very interesting work on MMO game social networks. His collaborator, Marius Cramer, built an German MMORPG that developed a large following that enabled them to collect detailed interaction and performance data on players. By examining a random selection of 55 players over a period of 3 months, they found that demographic factors significantly impacted network formation, with women being more desirable partners than men and older individuals accumulating more ties. However, the structure of the network actually had no discernable impact on player performance.
In fact, I saw this trend echoed throughout a few presentations today. These studies were often replications of previous work, yet the results they found were much weaker than others that have been reported. It did appear, however, that when this information was combined with demographic or other job-related data that predictive power increased dramatically. This is in line with some newer results that are emerging in the field as more diverse environments are being studied, namely that you cannot just look at social network structure. You need to dig deeper into the data at the node level to obtain a greater understanding of the individuals embedded in these networks. In fact, the data at this level may be even richer than that at the network level, as well as more easily interpretable.
Posted by Ben Waber at January 23, 2008 4:51 PM