Why do friendships matter?

Our society tends to place an emphasis on romantic relationships, and we think that just finding that right person will make us happy and fulfilled. But research shows that friends are actually even more important to our psychological well-being, especially over the course of our lifes. Friendships have a huge impact on our mental health and happiness. Good friends relieve stress, provide comfort and joy, and prevent loneliness and isolation. Developing close friendships can also have a powerful impact on your physical health. Friends are even tied to longevity, since a study from Harvard, conducted over 80 years, found that maintaining a rich network of friends can significantly increase your life. They even state that “Loneliness is as powerful as smoking or alcoholism”, meaning that a lack of social connections may pose a severe a risk to your health. But you can’t make friendships out of thin air. Many of us struggle to meet people and develop quality connections. Whatever your age or circumstances, it’s never too late to make new friends and greatly improve your social life, emotional health, and overall well-being.

What would your approach to making new friends be?

There are countless psychological guides on how to make friends. And undoubtedly this is an essential aspect when approaching new people. But what if it’s not just your social skills that determine whether you make friends? What if you could multiply your chances of making friends and vastly increase your friendship network by moving to a certain place? At a certain time of the year?

Our research is concerned with exactly these external factors that influence our social network. We have looked at two location-based social network datasets. Location what? Location-based social networks are social networks that use GPS data to locate you and that let you broadcast your location and other content from your mobile device. So to give an example: Imagine you travel to London for business. You know people there, but you haven’t had the chance to send them a message. However, you are connected with them on Foursquare, one of the biggest location-based social networks. All you have to do when your train arrives is to check-in to London’s train station. Your contacts will immediately know you have arrived and suddenly, they begin contacting you saying “Hey there, I didn’t know you were in town. Let’s meet up!”.

There are some more location-based social networks, like Gowalla and Brightkite. Unfortunately those two no longer appear to be in operation, however Gowalla announced a potential comeback for 2021. Nevertheless, from the time when they were active, Gowalla and Brightkite collected their user’s data and made it publicly available. These users have checked in to various locations all over the world between April 2008 and October 2010. For each check-in, we know the geographical location, the exact time, the user and all the friends of the user. Based on these check-ins, we could compute the home location of each user and whether or not a check-in was a visit to a friend or not. We will use this data as an unconventional approach to investigate friendship across time and space, providing insights about the external factors that determine someone’s social circle, like the place of residence and the time of the year.

There is nothing more to say in this introduction than: buckle up, sit back, and get ready to get answers to questions like “Where are people best connected? At what time do I make the most friends? Where do I find the most intense friendships?”.

Let’s start inspecting the home locations of our users. To calculate it, we discretised the world in 25x25 km cells, took all the check-ins of each user and calculated the home location as the average of all check-ins in the most frequently visited cell. This gives us a 85% accuracy of the true user’s home location.