Data Science

How Do People Become Connected? Community vs. Social Networks 2

727i2A698852917EF381Michael Wu, Ph.D. is Lithium's Principal Scientist of Analytics, digging into the complex dynamics of social interaction and online communities.

He's a regular blogger on the Lithosphere and previously wrote in the Analytic Science blog. You can follow him on Twitter at mich8elwu.


In my last post, I outlined some basic differences between social networks and communities from a social anthropological perspective. If you didn’t see that post I recommend you take a quick read of my previous article: Community vs. Social Network, as I’ll expand on that thinking here.


Today, I will continue our mini-series on the dynamics and interplay between communities and social networks. You will recall from my previous blog post that individuals in a social network are held together by pre-existing interpersonal relationships. Today, we will investigate how those pre-existing relationships were established on the first place.


RelationshipLifecycle_small.pngLifecycle of Relationships

One of the areas I touched on previously was that social networks connect every single person on this planet. Moreover, research has confirmed the validity of a popular urban myth: six degrees of separation urban myth. Recent data and calculation suggests that most people are actually within 6 to 7 degrees of each other, so this myth still holds up in the modern web 2.0 era. Therefore you can, theoretically, reach and connect to any one of 6.8 billion people on this planet in relatively few steps.


In reality, people don’t connect to the whole world. In fact, people don’t even connect with everyone who is living in the same city, going to the same school, or working in the same company as them. What prevents people from connecting? To understand this, let’s break down the lifecycle of any relationship into three stages:

  1. Creating the Weak Ties: This is the first stage in any relationship
  2. Building the Tie Strength: This cultivates the weak ties into strong relationships
  3. Maintaining the Relationship: This will prevent strong relationships from eroding and reverting back to weak ties.

Note: The kinds of connections that I will focus on are bidirectional, mutual and reciprocating ties, because these are the ties that can be developed into strong relationships. For these to happen, both entities must agree to connect for the tie to form. The entities that are connected by these ties are usually people, but they may be organizations, companies, or even countries.


The Desire to Connect is a Basic Requirement

In all three stages, both entities must have the desire to further the relationship. If any party becomes uninterested or finds the relationship not worthwhile, the process will halt and the relationship will not move to the next stage. That is, the tie may never be created, or the tie may remain remains weak (if it has already been created), or the tie strength may be weakened and revert back to a weak tie (if it has already been developed).


So how do people choose which tie to form, which one to develop and, which to maintain? This is a non-trivial problem, and is currently a subject of intense research. Many social network formation models have been proposed and studied recently. All of them are based on Game Theory, a very challenging branch of mathematics that models rational human behavior and strategic choices. Note: The great mathematician, John F. Nash (who was the subject of the Hollywood movie: A Beautiful Mind) received the Nobel Prize in Economics for his research on Game Theory. Clearly, a full treatise on this topic is way beyond the scope of this blog.


For simplicity, you can think of people’s choice as a result of a cost-benefit analysis of the action they are about to take. What does that mean? For example, in creating a weak tie, both entities will analyze the cost (or risk) and benefit of creating such a tie, as long as both feel that the benefit out weights the risk, they will proceed and create the tie. This is a very interesting topic, so I will revisit this topic with greater details in a later post. For now, just remember that all three stages of relationships development involve a choice that depends on the two persons’ desire to connect. If people don’t want to connect, no ties can be created.


Creating the Weak Ties

Besides the personal desire to connect, there are environmental factors that can affect tie formation by limiting people’s ability to reach each other. Clearly, if the environment precludes two persons from ever encountering each other, then there is no way a tie can form between them. There are basically only two mechanisms that people can meet and connect.

  1. Communities (online or offline)
  2. Social Networks


Mechanism 1: Community

For centuries, community has always been the place where people congregate and it is the place where social ties initially form. Years of social anthropological observation tell us that the majority of relationships in our social network were first established in some sort of communities. Certainly, most of my friends are people that I grew up with (in my neighborhood community), went to school with (in my campus community), my colleagues (in the same professional community), and fellow researchers (in the same research community). These are people who share some common communities with me at some point in time, and these communities can be geographical, cultural, interest-based, and/or institutional.


Therefore, sharing a set of communities (both online and offline) in common becomes the primary factor that will determine whether people can encounter. If the two people do not share any common community, then it would be impossible for them to form ties via this mechanism. The more communities these two people have in common, the greater the chance for them to encounter each other. This will increase the probability of tie formation between them.


Mechanism 2: Social Network

The second mechanism that people can meet is through our friends, colleagues, relatives (i.e. our personal social network). Again, a social anthropologist would say that this is again nothing new. Indeed, humans have been doing this since they were caveman. What is new though, is that Social network services (SNS) made it very easy for people to explore and discover the relationships that are normally unknown to them. Let me illustrate this with an example.

Let’s say I have a friend, Dave. I know Dave pretty well, but I may not know all of Dave’s friends. Jen is a friend of Dave, but I don’t know Jen. If I don’t see Dave and Jen hanging out together in any social context, I may never know that Dave and Jen actually know each other. But, with SNS, I can explore Dave’s network and discover that he’s connected to Jen. Subsequently, I may ask Dave to introduce us and become connected to Jen. This scenario is one that many people traditionally called social networking – the technological enablement of which has spawned our whole industry.

Although SNS greatly facilitates the process of social networking, there are still limitations. Even though it is possible to reach anyone on this planet through social network in about 6 or 7 degrees, in practice it is pretty difficult to actually reach people who are more than 3 degrees apart from you. So the primary factor that will determine whether people can connect through social network is the network distance (the degrees of separation) between the two persons.


In Summary

So what have we learned so far about how people become connected?

1. Weak ties can pretty much form anywhere. They are created:

     a. In communities (which are everywhere) and

     b. Through social networks (which cover the entire planet)

2. The formation of weak ties between two people depends on

     a. Their desire to connect (this is a very interesting topic that I will cover in greater depth in a future post)

     b. The amount of communities they shared in common

     c. The network distance (degrees of separation) between them

Next post, I will look at the second stage and try to understand how weak ties are developed into strong relationships. In the mean time, this is a pretty meaty topic, so I welcome any comments, questions or thoughts you might have.




Great article, Michael! This could easily be material from some humanities course on relationships, rather than a blog post on a technology company's online community. I know this is social media technology, but this still struck me as fascinating. I enjoyed reflecting on my own experience in both online and offline communities, seeing how weak ties are formed, and grow stronger or not. Way to boil it down...

Data Science

Hello Mike


Thanks for your interest and for following my blog.


Yeah, science and knowledge are also connected just as the internet, people, and pretty much everything in this world. There is a lot to be learn from the scientific development from other disciplines. I've seen it several times moving from Mathematics, to Physics, then to Neuroscience, and now to Social Analytics. That was actually one of the intent of this blog: To offer a different perspective of social media (the social anthropology perspective), and see what we can learn from that.


Although this post is the 2nd installment, it is still laying the ground work and more on the introduction side. The next two post will get very interesting, espectially for software development. I think there are a lot to be learn there from software design. Hope we can all make use of these knowledge.


Thanks again for commenting.


Oh, we're going to have to talk Neuroscience next time I'm out there. Totally fascinated with that lately. Of course, strictly from a layman's perspective.

Interesting, but surprisingly uni-dimentional: closer relations are those frequently updated because of dedicated investment)? That's a bit sad, is it?


I love the effort, but you might want to enrich that with what Carolyne Haythornwaithe ( and Bernie Hogan ( did around how closer ties are those who use many tools or pertain to several contexts… What she said about latent ties is enlightening, too :


Now that, thanks to smartphone, we have data on where people are, spend their time, etc. there is considerable rsearch to separate interactions (implicitly with colleagues) with attachement (including trust in long-lost friends).

Data Science

Hello Bertil,


Thanks for the comment.


I'm quite curious what you find "surprising" or "bit sad". Uni-directional ties, such as fans who follow some celebrities, are unlikely to develop into strong ties because the celebrity may never get to know the fan. So it is very difficult to have trust in this type of one sided relationships. So personally, I don't think that is surprising at all. I mention that only bi-directional, mutual and reciprocating tie will have more chance to be develop into strong relationships, which is what you called "close ties" right? Unless I miss understood what you meant by "close tie." My impression just from the term is that weak ties are acquaintances, and close ties are people that are close to you, you trust them, and have strong relationships with them.


I'm not saying that weak ties are not valuable. Both strong ties and weak ties has their values. In fact this will be cover in the next installment of this miniseries. I hope you will be back next time to see if it will address your "surprise."


Thanks for the reference. I will definite check out the academic papers you referenced.


Your point about trust in long-lost friends is very interesting. According to my simple model on the 3 stages of relationship, if you simply don't do anything to a strong relationship, that is sufficient to revert it back to a weak tie. But this weak tie is much easier to revive compare to a weak tie that has just been created, because there has been strong relationship before and people will remember that. So people's memory definite plays a role in who they trust.


Data Science

Hello Mike


Maybe I should start a miniseries on the similarity between social systems and neuroscience. There are actually quite a bit of similarity between them.


Yup, anytime you come out to the Bay Area, just let me know. I'm always happy to talk Neuroscience, in fact any science. Smiley Happy

Thanks Mike for a great post. I look much forward to your next post where you are going to explain how weak ties are developed into strong relationships. Would be pleased to continue with discussions on twitter and LinkedIn



Data Science

Hello Hans,


Thank you for the comment. I'm glad you find this work interesting.


Thanks for all the retweet and for getting connected on LinkedIn. Being a community evangelist, I'm sure you have a distinctive point of view, and I'm quite interested to hear about them. Most of my blog post are written for a general audience, so that details and technicality are often striped out. But I'd be happy to discuss this topic in greater depth with you. And we can do it here on Lithosphere (a community about everything community related), so others can also learn from our discussion or pitch in their ideas.


Thank you for stopping by and see you again next time.


Game theory is not the only way to approach grouping in social networks.  Consider also the behavioral economists such as Dan Ariely.  Along the behavioral line, consider that people are choosing to associate with people, which may have no other goal.  See The Hyper-Social Organization by Gossieaux and Moran.



Data Science

Hello Bill,


Thanks for the comment.


I totally agree that game theory is only one way to approach social network formation. And I know quite a bit about the predictable irrational behavior that Dan Ariely talked about.


I also agree that people may choose to associate with others with no real goals or motives. This may simply be due to homophily, triadic closure, or happen to be in a certain geographic location, etc. But I am not certain if these ties will developed into strong relationships though. Some of them may (this is not a physical law), but I think the number may be pretty small.


I will check out the reference you mentioned "The Hyper-Social Organization" by Gossieaux and Moran.


Thanks again for commenting. Hope to see you again on lithosphere.



Lithy Awards 2017

The winners in digital CX have been crowned!

See the winners!!