Seven degrees of separation and the dark knight ...

Just saw 'The Dark Knight'. I had fully expected it to have failed to live up to its hype. It didn't fail. It is every bit as good as the hype suggests.

The difference between 'The Dark Knight' and other movies of this genre is the touch of gritty realism. The Joker has a painted face and a permanent smile, but in a Hannibal Lecter way rather than the farce of most movie versions of comic book villains. There is little effort made to soften or explain the Joker's villainy. Also, the story has twists and turns that are more reminiscent of a well thought out whodunit than a popcorn thriller.

(If you have concerns about spoilers, you might want to skip the rest of this post.)

The one criticism I had about the movie is that when Rachel Dawes dies, Bruce Wayne (Christian Bale) seemed less disturbed about the loss than one might have expected. In fact, Bruce Wayne's reaction was so muted that I had to watch it again to confirm that the Joker had in fact tricked Batman into saving Harvey Dent instead of the Rachel Dawes.

The only other criticism was that all the protagonists, Batman and the Joker included, execute fascinatingly complex plans that, to paraphrase Jeeves, rely on too many imponderables to have a hope of success. All the plans have too many people who need to play their parts exactly, and too many aspects that need to go just right at exactly the right time for it all to work. Like a magic trick, it seems to fall into place as you are watching it, yet, if you look closely, it doesn't quite fit. However, this in itself is a paen to the film's quality. How often does one get round to seriously discussing the plausibility of the plot points of a film based on a comic book?

On a completely different note, I read a fascinating article that suggests that there is evidence to support Stanley Milgram's theory of Six Degrees of Separation, although its more like seven degrees of separation.

Microsoft's research suggests that 78% of people on IM could be connected to one another in 7 or fewer steps. This of course doesn't cover everyone. For one, unless those not on Microsft's IM are not human, it clearly covers Microsoft IM users only. Also, even the research itself suggests that there are at least 22% of IM users who are not as closely networked into the rest of the world. However, 78% is a pretty good coverage.

The other interesting part is that to connect the vast majority of the world in 7 or fewer steps, each person would need to bring an average of ~25 unique connections. I realize that Linked In and Facebook have made this seem more possible than ever before, yet, it does seem rather high. After all, how many people do we know that no one else in our group knows? That many unique connections for everyone seems very unlikely.

Instead, I wonder whether this level of interconnectedness is actually made possible by certain nodes, i.e. highly connected people who connected several relatively independent groups together. I remember having read something to this effect in some marketing research. If that's true, the number of degrees of separation really depends on who we know.