Your Friends Have
More Friends Than You
This is true for almost everyone, in almost every social network, probably including yours. It has nothing to do with how likeable you are. Click into a real network below and watch it happen.
Act one
Click yourself into the network
This is a small circle of friends, generated just now: some people are very well-connected, most aren't. Click any person to see the paradox from their point of view.
Click a person to begin.
Act two
Is it true for almost everyone?
Check every single person in this same network at once. Green means the paradox holds for them: their friends have more friends, on average, than they do.
Loading the network.
You're more likely to be friends with someone popular
Sociologist Scott L. Feld proved this mathematically in 1991, and the reason has nothing to do with charm. Popular people simply show up on more friend lists. If you pick a random friendship connection and look at the person on the other end, you're more likely to land on someone well-connected, for the same reason a random ticket drawn from a raffle is more likely to belong to whoever bought the most tickets. Averaging over your friends means averaging over a sample that's tilted toward exactly the people who have the most friends.
This isn't just a thought experiment. When researchers checked this on Facebook's entire global friendship graph, they found that around 92% of users had fewer friends than their friends' average. It holds almost everywhere real social networks have been measured.
The effect has a genuinely useful side. Researchers Nicholas Christakis and James Fowler used it to catch a flu outbreak earlier among college students: instead of monitoring a random group of people, they monitored the friends those people named. Because named friends skew toward the better-connected and more socially exposed, that group got sick, on average, weeks before the general population did, turning an odd sampling quirk into an early warning system.
- Ask students how big their classes feel, and the average answer is bigger than the school's actual average class size, because more students are sitting in the large classes than the small ones. This is called the class size paradox, and it's the same sampling bias wearing a different outfit.