<p>On an extended note - we have to ask the question - how applicable is the model to social groups with varying degrees of external contact? In such cases, it no longer describes the group in a near-absolute sense, but the model may then predict a scenario where the majority of people are very familiar with most other people in their houses/classes/groups/etc, even if not with every single person? </p>
<p>On a tangential note, I’ve visited several online forums over the last few years. One of them, CC, does not seem to develop into a particularly close-knit community, because it has well over 150 people and overturn is so high. There are few people from 2004 left posting here, for example. On the other hand, I’ve seen smaller online communities where many people are motivated to stay for years (due to the social associations involved within them, even if not everyone recognizes each other).</p>