Optimal social density?

According to some interesting research out of Facebook the degrees of separation between different nodes in the social network is decreasing. In but a few decades, if these numbers are right, we have gone from 6 degrees of separation to somewhat more than 4. Now, the interesting question is if this is a good thing or not. Are lower degrees of separation (or let’s call it sloppily higher social density) a good thing or not?

Obviously, making the case that it is good seems easy. We could argue that a more connected world will be more peaceful, that knowledge will be shared more easily and that we will all be able to understand each-other better. But that seems to imply that the lower the number, the better. This seems obviously untrue, though, right? Imagine a world in which everyone is connected to everyone else. That would be the same as not being connected to anyone since any interaction is still limited by basic human nature.

So here is a question: what would be the optimal social density for a general purpose online network? There are a number of different possible answers to that question.

  • Any density that follows from the total number of nodes in the network and a first set of connections that do not violate the Dunbar number. You could imagine that social networks are most useful when you can use and connect with all of your contacts in a natural way.
  • Any density that follows from the scarcity of attention. I.e. you could imagine a social network as an attention consumption device, and the density should be determined by the attention you are willing to invest in it, with a ceiling value of your total attention overall.

It is also possible to argue that the overall number is without interest, and that what we should focus on is the number of hops within cliques. What happens with nations, companies and other clique like networks over time? Do we see the same reduction in the number of hops there? 

Is it possible to think of values or advantages with systems that have really high degrees of separation? 7 and up? I sometimes wonder what society would be like if the degrees of separation would be the double of what they are today, like 12 or more. How would such a society work? Better or worse? As some of the comments point out in the RWW article we may live in such a society, and we may only have the degrees of separation of the people who have joined FB. That set is surely biased. This is true, but still allows us to ask what optimal social density looks like and what determines it. It seems that there is a cut-off rate somewhere, where the lack of separation leads to social meltdown.
I wonder if there are any natural experiments that we can look at where we could see two groups with different degrees of separation compete about the same resources, and whether we can see if one social density is superior to another. Or if there would be a way to simulate such an experiment to see what social densities are optimal.
What would that look like? I guess you could imagine an equation that sort of took into account the benefits of social density and then we would have to model the costs. One way to do it would be to say that the speed of signals depends on social density. but that the accuracy of the signal depends on separation. That is not quite right, though. As we know signals often deteriorate if they have to travel through a number of different nodes in a network, so lower social density would also increase the unreliability of the signal. Maybe there is something there, and maybe attention needs to factored into the simulation too.
I wonder if there is a sci-fi story about to civilizations with completely different social densities at war? Ok, off to check that. Enjoy your Friday, folks.
Be Sociable, Share!

4 Kommentarer.

  1. There is one major problem with this. Namely, people have very different criteria for who they add as a friend on social networking web sites.

    Some people will “friend” just about anyone who asks and tons of people who don’t, while others only “friend” people they have met in real life, and yet others only “friend” their best friends. Some people add colleagues to their friends lists, others keep work and personal life strictly separated in their online presence.

    Any attempt at trying to use a social distance metric derived from social network web sites’ data is going to have a very hard time compensating for peoples’ different ideas of what it means to add someone as a “friend”, and without that compensation, any conclusions are at best going to be very vague, and likely not useful for anything important.

  2. Yet still the network will have this quality no matter what, and it is quite possible to think about how different networks with different qualities compete. I think you are speaking about the way the links are generated, whereas I am not sure that matters for the analysis.

  3. Facebook is still “only” 800 million users. It may be that other technologies, like cheap mobile phones, are also decreasing the average degrees of separation but I would bet 6 is still a more accurate global figure than 4. Like you said, attention is obviously limited by human nature. And I think that has to be factored in somehow to make a useful measure.

    A naive computer simulation could perhaps give some, fun if not accurate, answers.

    Did you find any sci-fi stories?

  4. Pingback: What is your care/share-ratio? | Nicklas Noterar

Kommentera gärna!