Tuesday, February 12, 2008

Tuesday Statisticz

I've always thought that statistics should be spelled with a z, like in the title, to make it appeal to young people more. Sort of like how on the OC the cool teenagers always drink Mountain Dew, because it's the edgiest soft drink.

Anyway, here's my newest section: the Tuesday Statisticz. This is my new weekly segment where I take totally irreverent numbers and try to make sense of them, in order to show that numbers are silly (in a good way). This week I've taken crime data from the FBI's website and contrasted it with the favorite movies of different regions, accessed from the networks page on facebook. The idea was the immortal question of whether a preference for violent movies corresponds with how much violence there is in the real-life city in which that person presides. I tallied up the data for ten cities, and here's what I got:

Each of those data points corresponds to one city in the United States. The x axis refers to the number of violent crimes committed in the city limits, while the y axis refers to the percentage of movies that I considered "violent" of the city's 10 favorite. One of the problems with the data is that there isn't much variation of favorite movies. The percentages range from .2 to .5, with Fight Club and Pulp Fiction (both violent ones) appearing on many top 10s. Speaking of which, here are the movies that I had to deliberate on for a little bit as "violent" or not, with explanations.

Top Gun -- Not violent. The volleyball scene on the beach clearly precludes it.
Lord of the Rings trilogy -- Not violent. Sorry fanboys, but if anything, based on some of the talk between those hobbits, I'd call it an erotic thriller.
Braveheart -- Violent. Torture scenes, sweet battle scenes, and the bare ass Scottish mooning scene all push it over the edge.
Gladiator -- Violent. I hesitated because the story is so central and there's too many icky love scenes, but then I thought of Russell saying "I will have my vengeance" before stabbing Joaquin Pheonix, and I had to give it the nod as violent.
Star Wars -- Not violent. My parents wouldn't be very good if they had let me watch the trilogy 500,000 times when I was 8 if it were violent, would they? Chewbacca is way too cuddly to be a major character in a violent movie.

As you can see, there's very weak correlation (0.044) between violence per capita in a city and a preference for violent movies in the top 10 favorites of people who identify themselves in that city on facebook. The cities I chose were San Francisco and Boston, and then I went down the list alphabetically, so most of the rest of the cities are in Alabama, Alaska, and California.

Aren't you proud of me for getting a non-significant result on the very first edition of Tuesday Statisticz? No data hunting here. Check back next week for another installment.