Here’s a Venn diagram displaying the overlap of my Facebook friends, my Twitter friends, and the percentages of each that I’ve met in person:

After a bit more work, I put together this rather difficult to decipher chart that shows overlap between all of the various categories.

Each bar in the chart represents the % of group A that are in group B as well.  So, for example, on the left of the chart you can see that out of everyone on Facebook, I’ve seen about 52% of them in the last year.  However, if you look under the “Seen in last year” column, you can see that of all the people that I’ve seen in the last year, 90% of them are on Facebook.

Note: the sets of “Met” and “Seen in last year” are not comprehensive, but only include people who are also on Facebook, Twitter, Livejournal, Flickr, or in my phonebook.

In any case, I mention this because I think it makes a good argument that when you have 665 friends on Facebook that it represents a group of people that are not in any way analog to your friends in real life.  Jonah Lehrer argued that we may even have less than typical empathy with our online friends.

I think that this project, for me, is revealing that my “online friends” simply represent a broader slice of people that I know, including my baker, my lawyer, my conference friends, my highschool friends, etc.  It’s not a qualitatively different circle than the normal circle of friends, it is simply the first time we have a way of visualizing the size of all the different people we run into during life and have brief (or strong) relationships with.

My Friendsheet project is a part of my general self-tracking goal.  Over the last couple months, I’ve been working on collecting all of my friends on Facebook, Twitter, and Flickr, deduping them for duplicates, and then slowly going through and flagging them for various different shared qualities.

The distribution is pretty much believable, considering that I had friended about 635 people on Facebook at the time that this was begun.  It makes sense that I’ve eaten with 31% of the people but have had only about 13% of the people over to my house.

I’m going to continue adding new categories to this until I feel like I’ve covered enough to move on to the next step in the project.  A few axes I’m considering: talked to on the phone with, IMed with, met parents of, and given a present to.

The next step is going to be to come up with a couple different ways to visualize this, more as a distribution of similarities, and less as info-graphic.  I love taking data like this and abstracting it out so that the information is gone, but the patterns remain.