You probably heard of the Six Degrees of Kevin Bacon game. The objective is to find a way to get from any actor to Kevin Bacon in six steps or less. For example, Sean Connery has a Bacon number of 2 (Sean Connery was in Wrong Is Right with Myron Natwick, who was in Cavedweller with Kevin Bacon). That's possible because Kevin Bacon stars in a whole lot of movies. But is he really the center of Hollywood?

I tried to find out if Kevin Bacon's network is indeed as dense as the Six Degrees game suggests. And of course, I used Google for that. Here's what I did, and you can try the same; I picked a list of 50 random famous actors, including Kevin Bacon, and searched Google trying to find out if any two of the actors on the list were in a movie together. Of course, this isn't statistical correct proof. But it's fun. Here's an example of a search query:

"Sean Connery and Julia Roberts" OR "Julia Roberts and Sean Connery" -degrees

This will return all pages with either the first or second phrase in them. (I exclude pages with the word "degrees" because I don't want to hit on pages where people played the Six Degrees game, as that would give Kevin Bacon an unfair advantage.) Whenever over 500 results have been found, I will count this as a "hit."

The following map shows all hits combined into a social network1. Some actors of the 50 I included in the game actually didn't make the list because they had no connection at all – like Humphrey Bogart.

all hits combined into a social network


What does the map show? For one thing, that Kevin Bacon is not the center of the Hollywood universe – at least not using this (non-representative) sample. Instead, Julia Roberts, Johnny Depp and Tom Cruise seem to be the most connected. On the other hand, you can also see that it's easy for almost everyone on the list to get to Kevin Bacon in six steps or less.

A Network of Everything

How well does this approach of visualizing a network fare with something other than actors? We can also use it to find connections between any two things. For example, we can create a network of connections between things and their categories. To create the following image, I used the words Britney Spears, apple, horse, speakers, piano, violin, carrot, and orange. As categories I used food, actor, movie, book, song, album, company, band, tool and a few more. I applied a threshold of 50 Google results to count something as connection, and I used glue phrases like "is a", "are an" and so on:

A Network of Everything

You can see Britney Spears is a celebrity singer. "Apple" is an ambiguous term, meaning both the company, and the fruit.

End Notes

1. The visuals are created using Sun's GraphLayout tool.