(Warning: nerdy, but skippable, portions found in [brackets])

The other night I was thinking about a friend who shares an email account with his wife. I know there are a lot of efficiencies that come from sharing – but does that apply online, when things are free and no real resources are consumed? Sharing an email account seemed more like a bother than anything. Do I write to both of them? Then who responds? Do they have to confer?

Couples share blogs, too, and this is pretty common. But again, rather than having separate authoring accounts, every post comes from Jim & Kathy, Susan & Richard, Bramlet & Chartreuse. That doesn’t make sense. Let’s be honest – it’s the wife writing it.

I started seeing shared profiles on facebook recently, too. Facebook even provides a way to link profiles through a marriage indicator. I didn’t go to high school with both of you – that just doesn’t make sense. Again.

So I decided to see if my friend with the shared email account had a blog. I took a pretty good guess that it would be his_nameandher_name.blogspot.com. The site existed, but it wasn’t them. Surprise. Frankly, it might as well have been them. I can’t tell the difference between most personal blogs anyway – more on that later.

But I wondered – how interconnected are all these silly blogs? People tend to link to the blogs of their friends. But friends seem to form circles. So, the big question: is there enough overlap, that I could navigate from a random blog back to my own?

There’s only one way to find out, really. So I developed a heuristic [from computing: proceeding by trial and error or by rules that are only loosely defined] for how to traverse the blog-space:

• Prefer last names that I know
• Prefer geographical locations where I have a connection
• Prefer other commonalities
• Prefer anything familiar
• Don’t follow links already followed (obvious) [cycles don't help anyone here]

[nerdy: If you consider blogs a tree, I was doing a depth-first pre-order traversal, with a self-imposed depth when I felt I wasn't getting anywhere. But if you consider blogs a tree, you're crazy. Obviously a tree isn't the right topology for how the blogs are connected. They're more like a disconnected graph (and therefore I was doing a Prim-Jarnik traversal?). I suggest they're disconnected because it's certainly not a complete graph and I don't believe every blog is connected - unless you count google as some kind of super-node, but google isn't a blog. Really, I was building a minimum spanning tree representing my paths. And I think anyone with even the simplest of advanced degrees in computer science would recognize that. ;) And that's an allusion to a line from Simpsons, to take this nerdy post up another notch.]

So I started clicking on links. I obviously had no idea how long it would take. [nerdy: Worst-case big-O(time) would look like the number of all connected nodes to that particular blog, which is not just unknown but potentially huge.] At first I had no hope that it would actually work. But I stuck to my rules, and kept clicking.

I was so shocked that only 37 blogs later I had connected back to my own blog. Now, that was 37 trials [there were 37 nodes on my constructed tree]. But if you follow the direct path, I was only 15 steps away from my own blog [15 nodes on the tree before reducing it].

Would you have guessed that? That is, from a random blog, I was only 15 clicks away. It may actually be fewer clicks if I took different paths and found a shorter one [a reductive step].

What did I learn along my journey?

• Blogs look the same.
• For some reason, when people decide to have music on their blog, it’s usually country – so, doubly obnoxious.
• People blog about really mundane things. I wasn’t reading blogs as I went, but I could tell it was nothing exciting. But while neither I nor the internet at large cares about what people are doing – whoever is reading that blog does. That is, I care about what my sister is doing, but I don’t care what your sister is doing – unless you’re my other sister.
• People have strange last names.

[Edit: Ok, I realized that it's not necessarily a Prim-Jarnik algorithm. That would be used on a weighted graph, which this isn't. All paths here have equal weight. And this exercise was more about searching than about finding minimum-paths.]

Tags: 7 Comments