Thread: Six Degrees

Ever seen the 6 degrees of Kevin Bacon game?

http://www.thekevinbacongame.com/

Actually, experiments done on that theory show it's really more like 7 degrees, but the idea is the same.

Actually, simulations and well-conducted experiments (read: not Milgram's experiments) point to the average number of steps being around 3.

For example, a really cool online computer simulation called The Oracle of Bacon, using data from the IMDb, shows that the average number of steps between Kevin Bacon and any other actor/actress who has played in a television film or major motion picture is 2.952. Interestingly, Bacon ranks only 507th on the ordinal scale of Hollywood connectedness. The most connected actor? Dennis Hopper at 2.743404 average steps. The actor with the worst average degree of connectedness -- as in the most obsure actor out of all the obscure actors ever to set foot in Hollywood -- has an average of 9.001 steps to any other actor.

Probably depends on what someone you know is defined as.

Originally Posted by DuB

Actually, simulations and well-conducted experiments (read: not Milgram's experiments) point to the average number of steps being around 3.

For example, a really cool online computer simulation called The Oracle of Bacon, using data from the IMDb, shows that the average number of steps between Kevin Bacon and any other actor/actress who has played in a television film or major motion picture is 2.952. Interestingly, Bacon ranks only 507th on the ordinal scale of Hollywood connectedness. The most connected actor? Dennis Hopper at 2.743404 average steps. The actor with the worst average degree of connectedness -- as in the most obsure actor out of all the obscure actors ever to set foot in Hollywood -- has an average of 9.001 steps to any other actor.

Yeah, but the movie star world is a much smaller and more inbred group. Movie stars hang with other movie stars.

In the experiment I was referring to, the goal was to deliver a letter by hand, from person to person until it reached the intended target. The average number of handoffs was 7, even when the recipient was on another continent.

If you're referring to the Milgram experiments that I mentioned earlier, the average was actually about 6. But I don't doubt that if the experiment was repeated over entire continents that they might have found an average of 7, as you say.

The main problem with these letter passing experiments is that the letters aren't passed in an optimal way. The people who receive a letter are supposed to pass it to the person they know who they think would be closest to the ultimate recipient of the letter. But how on Earth could they know who that person is? It's hardly possible at all. So what we see is that the letters meander around and eventually find their way to the recipient in 5-7 steps, but this doesn't actually tell us much about how people are actually connected -- it tells us a lot more about how little people know about their extended social networks. It's for this reason that computer simulations are much more accurate in this area. They can tell what the shortest path to any other member really is, which is what we are interested in finding out. In other words, while letter passing experiments give us the true number plus a lot of human error, computer simulations just give us the pure number. And simulations run on many sets of data -- not just the IMDb -- consistently give us an average number of steps that is around 3.

Originally Posted by DuB

The people who receive a letter are supposed to pass it to the person they know who they think would be closest to the ultimate recipient of the letter. But how on Earth could they know who that person is? It's hardly possible at all.

Good point.