House Parties and Social Networks
When a 17-year old throws a party in his family’s mansion and invites “a few friends” over, things have the potential to get a little crazy. According to this post on an online forum, one such case ended up in over 2000 people present at the residence and about $70,000 in damages at the end of the night. How could a few friends lead to 2000 people in such a short time? Well, here we just have to think about the ties between the boy and the few friends he invited, and then these friends and their ties to others. Considering the structure of social networks, Twitter, in particular for this case, it is actually pretty clear how the scale of this house party escalated so quickly.
Using the analysis of strong and weak ties we can explain why news of this party lead to it getting out of hand in just a matter of hours. The 17-year old invited, let’s say, 10 people over initially. We can make an assumption that his ties to all of these friends were strong, and that he has weak ties or no ties at all with anyone whom he didn’t personally invite. Now let’s say 3 of the people he invited have friends (strong ties) who weren’t invited. In the interest of keeping these guests happy, the boy may have allowed these 3 friends to invite a couple of their friends over as well. The problem that arises here is that, for whatever reason, the boy and these acquaintances (weak ties, via Strong Triadic Closure Property) are not exactly close. Due to this, these people may have viewed posting information about the get-together turned party on Twitter. According to this article on Twitter statistics in 2014 (when the party was thrown), the average twitter user has about 200 followers. Using this, we can easily see how word about the gathering spread. If each of the secondary guests was to tweet about the party (#MansionParty was commonly used that night), then that would be 3 friends*2 secondary guests each*200 Twitter followers = 1200 people who knew about it, and then there are also retweets and other actions that could be taken into account.