Wednesday, May 7, 2014

Which Friend is the friendliest?

Slate coded Friends transcripts into a database and produced an interactive display of which of the friends shared scenes together.

According to Slate's results, each friend had about equal screen time. But I wanted to ask which friend appeared in the most different combinations, a.k.a., which friend was the "friendliest"?

To start with, Joey lost 1 point for being the friend who appeared in the most scenes alone. Ross also lost a point for being the friend most commonly excluded (that is, the Chandler-Joey-Monica-Phoebe-Rachel fivesome was the most common fivesome).

Next, I looked at the dyads. Given that Joey was in a dyad on screen, Chandler was his most common companion (so Joey gives one vote to Chandler). These are not always symmetric: given that Chandler was on screen, Monica was his most common companion (so Chandler gives one vote to Monica). There was only one tie: Phoebe was equally likely to pick Monica or Rachel.

Things get a bit more complicated with triads. I started with a given dyad. Given that Chandler and Joey were on screen in a triad, Ross was the most common third member, which makes sense. Given that Monica and Phoebe were on screen in a triad, Rachel was the most common third member.

Quads were more complicated still. Given that Monica, Phoebe, and Rachel were on screen in a quad, Ross was the most common fourth member. Likewise for the quints.

And the final results are:

How to square this with the fact that each friend had approximately equal screen time? The reason is that some friends tended to dominate a few groupings: Ross was Rachel's first pick for a dyad by a wide margin; Ross was Joey and Chandler's pick for a triad by a wide margin; and Ross was the most common guy picked to join Monica, Phoebe, and Rachel. He accumulated his screen time in just a few other groupings, but none as often as these.

Other friends spent time in a much more distributed way. Monica and Chandler appeared in all sorts of different groupings frequently. And in my book, that's a reasonable enough way to measure "friendliness."

No comments:

Post a Comment