Who’s more powerful on Facebook – 20 year olds or 60 year olds?

Who is more powerful on Facebook?  60 year olds or 20 year olds?

Image: Flickr - Greenchameleon

I’ve been thinking about this while chewing hard on an interesting nugget from the recent ‘Anatomy of Facebook’ study published by the Facebook Data Team and collaborators.   (You can find a summary of the study here and a link to the full article download here.)     The team selected five different age bands in the Facebook population, and for each of these five age cohorts, they looked at what age all that person’s Facebook friends were.

It’s the oldest Facebookers in the sample which have the widest variation in age range within their circle of FB-friends, and the youngest who have the narrowest.

Figure source: The Anatomy of Facebook - summary (Facebook Data Team)

Does this mean that if you want to achieve the widest potential reach for a message,  your best bet is to target 60 year olds to spread it?   That would be delightfully unexpected, if  true.  But I think it isn’t.

Let’s take a stroll together through a sanity check.  For every 60 year old who’s FB-friends with a 20 year old, there is a corresponding 20 year old in the data who’s FB-friends with that very same 60 year old.   For 60 years, the 20 year olds are noticeable, as a part of their friendship circle.   But,  when you look at the age distribution of FB-friendships for 20 year olds,  60 year olds form a vanishingly small part of their FB-friendship circle.

How can both these things be true?  My guess about what’s going on is that the 20 year olds tend to have many, many more friends than the 60 year olds.  I think that’s the main way in which 20 year olds could be important in 60 year olds’ FB-friends’ age distributions, but 60 year olds are not important (numerically speaking) to 20 year olds. This is just a guess but it’s my best guess about how these two facts could fit together.  If you have other ideas let me know what they are!

What’s the consequence of this pattern of connectivity?   We all know that raw connectivity and influence are not the same thing.  If you’re feeling digressive and a bit geeky here’s a fun paper on the topic of how network structural characteristics affect viral distribution patterns by Kitsak et al.   (Have fun but come back soon.)   But without connectivity, there is no path for influence to propagate.   So connectivity is  interesting.   It’s just not the only thing that’s interesting.

Considering the universe from a path connectivity point of view, what can we say about the relative potential power of 20 year olds vs 60 year olds?   From the point of view of outbound communication,  we can guess that as a 60 year old FB-friend it’s probably pretty hard to get the attention of the 20 year olds you’re connected to.  Speaking purely in connectivity terms, you have to fight for attention against all those inbound comms channels from all those 20 year old age-mates.    Your input is one of many.

But look at the information flow from the opposite perspective, and a tantalising possibility emerges.   The 60 year olds have the most broadly balanced feed, in terms of the age range of their information sources.   They will be better listeners, as their mixing deck is better adjusted to a wider range of signals from reality.   They will know a greater diversity of things.  They will be wiser, for structural reasons.   If you think knowledge is power then you should bet on the 60 year olds.   But I bet you knew that already.

What can 68.4 billion Facebook friendships teach us about reality?

The Facebook Data Team recently published a ground-breaking piece of collaborative research, by Ugander, Karrer, Backstrom, and Marlow,  which characterised the network structure of the entire active population of Facebook as of May 2011.  That’s 721 million active users, who between them had formed 68.4 billion Facebook-friend relationships.   You can read a summary of findings here, and the entire article can be downloaded from here.

Computing the network metrics characterising this population provided employment for 2,250 servers, deployed in a Hadoop cluster.  Awesome™.  Some of it was done on a  a 64 GB machine via a stream-based algorithm, and some on a 24-core 72 GB machine via a novel algorithm.   (Still Awesome™.)

Why did they look at all their data, when doing so was pretty difficult?   Personally, I think they fed their whole active user base into their crunchers partly because they could.  Facebook is not an ordinary engineering company, and it is not just an engineering company, but it is definitely an engineering company.

But the reason given in the text is different:

Network completeness is especially important in the study of online social networks because unlike traditionals social science research , the members of online social networks are not controlled random samples, and instead should be considered biased samples.(p.2)

Now, that is really interesting.    It is interesting not because it’s flawed, but because of the way it’s flawed.   Making the Facebook ‘sample’ as big as humanly possible doesn’t help, if what you want to talk about is the structural characteristics of human social networks.  The whole population of Facebook is a biased sample if you view it as a sample of human social relationships.  The logic of the authors’ assertion that you need to look at the whole population because using just a sample is biased is wrong.   Even if you use the whole population, the entire Facebook universe is still biased as a sample if what you want to talk about is ‘human social networks’.

However, if what you want to talk about is the nature of Facebook-friendships, being able to look at the social graph patterns  from the whole population of active Facebook users is just amazing.   (And, of course, Awesome™.)  But there’s simply no need to confuse Facebook with reality, and say the reason you are looking at your entire data set is to avoid sample bias.  Looking at the whole shooting match doesn’t help.

The authors go on to say:

…the most accurate representation of our social relationships will include as many people as possible.  We are not there yet, but in this paper we characterise the entire social network of active members of Facebook in May 2011… (p. 2)

Again,  a pivotal bit of confusion revealed by choice of language.  The study reveals the structure of Facebook-friendships, and Facebook-social-networks.    It does not study “the entire social network of active members of Facebook”.

The question of how people form and maintain relationships using Facebook is a  truly fascinating one.  So, too,  is the question of how the design of Facebook facilitates and influences these fundamental human processes.   The answers are important for a variety of  reasons, theoretical and applied, personal and commercial.  They are important to a huge range of stakesholders in the system in addition to Facebook itself.  So I’m really glad Facebook thrashed thousands of servers to do this study, and I’m gladder still that they published it.   I am looking forward to cherry-picking some of the tastier findings in future posts.

But I will always have social connections who aren’t connected to me on Facebook.

So, can games learn from TV? (And if so, what…?)

By popular demand (whose name is Ed (Hi Ed.  – Ed)),  I’m doing a quick post about the talk Sorrell from ScreenPop gave last week at the Evolve conference in London.

Before I went, I made some excited noises about the Sorrell talk because my guess was that he was going to be talking about “social game TV”.    Just think.   If social TV is hot and social games are hot then just think how hot “social TV games” could be.  Ooh.  (And natch it could also be utter crap, if you do it badly.  But the upside is real.)

Only one hitch: I was wrong, wrong, wrong.    Not that a social game layer on top of TV wouldn’t be awesome™.  It could be.

But social TV games just weren’t the tack Sorrell took in his talk.   He did a pretty  straight rant-flame about what games can learn from TV.   His take, as I understood it, is that TV is way more successful than games in terms of the value of the industry, and the number of eyeball-hours, probably always will be unless games get their act together and listen up to the following words of wisdom:

• TV understands how to use familiar music to cue emotions – games are pretty rubbish at this by comparison
• TV uses story loops well, if games combined story loops with rat loops [i.e. compulsion loops] nobody would ever leave the house again
• the games industry will never succeed until it hires more women, 15% isn’t enough
• TV has cracked the recommendation engine problem, it’s called channels, and 80% of shows are still watched this way, real-time at broadcast time
• there’s a special experience to do with live events, although ‘live-ification’ is a  word only used by particularly horrible TV execs
• no game justs gives you a task with no choice, but turn on the TV and you’ve done everything you need to do, and passive is good.

All very interesting, in an intentionally controversial but still content-ful way.  Just not what I was expecting.

So, in the question period I asked Sorrell what he thought of social TV.    He rolled his eyes a bit,  in an “oh dear oh dear not that question again” kind of way, but said that if anyone was going to do it,  it would be  Zeebox, one reason being that they allow unofficial apps.

Les jeux sont faits….