{{
`${userDetail.user.displayName}${userDetail.user.cachedBungieGlobalDisplayNameCode ? `#${standardizeBungieGlobalCode(userDetail.user.cachedBungieGlobalDisplayNameCode)}` : ""}`}}
{{getUserClans()}}
11/22/2014 1:12:10 PM
1
First I get list of all the followers for each person who posts here then arrange everyone in a graph (which you can imagine being like [url=http://anet.sourceforge.net/docs/user/fig2.png]this[/url]). I then calculate various statistics from it, one of which includes [url=https://en.wikipedia.org/wiki/Betweenness_centrality]how well a given node (an individual user) is connected to all the other nodes[/url]. I then use the value generated for a given node as a rank which determines how large it appears in the visualisation (the picture).
The followers a given user connects to don't have to have posted in this topic. I also don't consider the followers of followers.
English
-
So it's all about how many of your followers are also following the the other people in the graph?
-
Not exactly. It's [i]how well[/i] you can connect with the other users. For example, if you were a major highway and all your followers were regular streets, you would have a high centrality to the network because the shortest path to get from one street to another would likely be along your "highway", so-to-speak.
