SCIENCE
How Friends and Strangers Shape What We Do
Thu Jul 10 2025
People often copy what others do, especially friends. But does having many friends help or slow down the spread of new ideas or behaviors? Some think that if people are more likely to copy their friends, then behaviors will spread faster and wider in groups where everyone knows each other well. Others believe that behaviors spread better in groups where people are connected randomly, without many close-knit circles.
To figure this out, researchers created a model. They adjusted how likely people were to copy others and how much friends influenced them. Then, they compared how behaviors spread in two types of groups: one where people were tightly connected (like a close-knit community) and another where connections were random (like strangers meeting at a party).
The results were surprising. Most of the time, behaviors spread just as far—or even farther—in random groups, even when friends did influence each other. There were some cases where close-knit groups helped behaviors spread farther, but this wasn't the usual pattern. Close-knit groups were even less helpful when people stayed influential for longer, had more friends, or needed more friends to start copying a behavior. In fact, close-knit groups only did better than random groups in a small part of the cases studied—just 22% of the time under the best conditions.
This shows a trade-off: Random connections help behaviors reach more people, while close-knit connections make it more likely that friends will copy each other. So, if you want a behavior to spread widely, having some random connections might be more important than having a tight-knit group of friends.
continue reading...
questions
If behaviors were pizza toppings, would random networks be the ones trying every topping or the ones sticking to pepperoni?
Could the preference for random networks over clustered networks be a deliberate strategy by certain groups to control the spread of information?
Is the observed trade-off between reach and reinforcement a result of hidden agendas within social network structures?
actions
flag content