Street Math and the Zeitgeist

Finding out where movies are filmed is easy. IMdB and others have comprehensive databases. But a huge proportion of movies are filmed in cities like Los Angeles, Vancouver, and Albuquerque and made to look like they’re set somewhere else. A database of movie plot locations doesn’t exist yet.

Here’s my estimate of how likely each major U.S. city is to be featured in a movie plot:

Washington, D.C., for example, is 20.3 times as likely the setting for a movie plot as the average movie (per capita).

The Process

1. For every non-TV movie, I parsed review sites, Wikipedia and IMdB listings, and tried to dynamically assess which cities were referenced or featured in the movie’s plot. Some parsing was easing: Wikipedia references cities like Omaha,_Nebraska and IMdB has a robust list of user-contributed and site-monitored keywords. Other implied references were made using crude natural language processing, trained using primary Wikipedia city and state articles (landmarks were common triggers). I ignored movies that didn’t have any reviews or plot descriptions: about 10,000 made the cut.

2. For each U.S. city/state that occurred in the text (verbatim or implied), I came up with a score for the movie-city or movie-state combo loosely defined by:

I wanted better, more popular movies to weigh more, but not overwhelm more quintessential classics. Rain Man was as critically and commercially successful as any movie, but didn’t top the Las Vegas movies because it took place in other cities too.

3. I added up the score for all U.S. cities with a population of 400,000 or more and sorted by score per capita. Foreign cities were harder to parse because of inconsistent language.

4. All the parsing and scoring was automated and no manual edits were made, so there’s plenty of fuzz in the estimations. The Pivotal Movies were the highest scoring movies associated with each city.

By State

State scores were more abundant, but less precise. A visual interpretation:

Likelihood of US states being the subject/location of a movie plot (per capita).

According to the news, Charlie Sheen’s career is in shambles. He’s facing some unhealthy distractions that may cause him to miss some work, but he’s having fun too. I don’t doubt his short-term work will be limited, but I don’t buy that his long-term prospects are gloomy.

Is Charlie Sheen’s decline the worst in Hollywood? Let’s put the fact that Charlie’s still the face of the biggest sitcom on TV aside and just look at movies to find out.

Using some math, here’s a list of the the actors whose movies are sucking the fastest. Keep in mind that an actor’s “decline” implies they have somewhere to fall from — the better they originally did, the further they have to drop.

1		Vin Diesel Decline / Year: 4.88		Early Hits: 91 – Saving Private Ryan (1998) 97 – The Iron Giant (1999) 67 – Boiler Room (2000) Recent Misses: 36 – F&F: Tokyo Drift (2006) 7 – Babylon A.D. (2008) 28 – Fast and Furious (2009)
2		Damon Wayans Decline / Year: 4.23		Early Hits: 84 – Beverley Hills Cop (1984) 88 – Roxanne (1987) 87 – Hollywood Shuffle (1987) Recent Misses: 48 – Bamboozled (2000) 9 – Marci X (2003) 9 – Marmaduke (2010)
3		Amanda Seyfried Decline / Year: 3.67		Early Hits: 75 – Nine Lives (2004) 83 – Mean Girls (2004) 53 – Mamma Mia! (2008) Recent Misses: 33 – Boogie Woogie (2009 28 – Dear John (2010) 41 – Letters to Juliet (2010)
4		Lindsay Lohan Decline / Year: 3.22		Early Hits: 79 – The Parent Trap (1998) 88 – Freaky Friday (2003) 83 – Mean Girls (2004) Recent Misses: 13 – Just My Luck (2006) 17 – Georgia Rule (2007) 7 – I Know Who Killed Me (2007)
5		Barbra Streisand Decline / Year: 3.18		Early Hits: 92 – Funny Girl (1968) 100 – On a Clear Day… (1970) 91 – What’s Up, Doc? (1972) Recent Misses: 54 – Mirror Has Two Faces (1996) 38 – Meet the Fockers (2004) 10 – Little Fockers (2010)
6		Orlando Bloom Decline / Year: 3.10		Early Hits: 92 – Lord of the Rings I (2001) 77 – Black Hawk Down (2001) 96 – Lord of the Rings II (2002) Recent Misses: 20 – Love / Other Disasters (2006) 45 – Pirates of the Carr. 3 (2007) 38 – New York, I Love You (2010)
7		Billy Burke Decline / Year: 2.95		Early Hits: 86 – Under the Skin (1997) 78 – Without Limits (1998) 60 – The Independent (2000) Recent Misses: 15 – Untraceable (2008) 27 – Twilight 2 (2009) 50 – Twilight 3 (2010)
8		Macaulay Culkin Decline / Year: 2.88		Early Hits: 56 – Uncle Buck (1989) 71 – Jacob’s Ladder (1990) 47 – Home Alone (1990) Recent Misses: 8 – Getting Even with Dad (1994) 27 – Party Monster (2003) 9 – Sex and Breakfast (2007)
9		Julie Andrews Decline / Year: 2.70		Early Hits: 100 – The Amer. of Emily (1964) 100 – Mary Poppins (1964) 100 – Thor. Modern Millie (1967) Recent Misses: 41 – Shrek the Third (2007) 58 – Shrek 4 (2010) 17 – The Tooth Fairy (2010)
10		America Ferrera Decline / Year: 2.64		Early Hits: 83 – Real Wom. Have Curves (2002) 86 – Steel City (2005) 79 – Sis of the Traveling Pants (2005) Recent Misses: 11 – Towards Darkness (2008) 61 – The Dry Land (2010) 14 – Our Family Wedding (2010)
11		Derek Luke Decline / Year: 2.62		Early Hits: 79 – Antwone Fisher (2002) 84 – Pieces of April (2003) 81 – Friday Night Lights (2004) Recent Misses: 36 – Miracle at St. Anna (2008) 26 – Madea Goes to Jail (2009) 50 – Notorious (2009)
12		David Spade Decline / Year: 2.61		Early Hits: 94 – Light Sleeper (1992) 65 – Reality Bites (1994) 1994 – PCU (1994) Recent Misses: 12 – The Benchwarmers (2006) 18 – Grandma’s Boy (2006) 10 – Grown Ups (2010)
13		Chris Klein Decline / Year: 2.38		Early Hits: 92 – Election (1999) 59 – American Pie (1999) 63 – We Were Soliders (2002) Recent Misses: 20 – The Good Life (2007) 8 – New York City Serenade (2007) 10 – Stright Fighter (2009)
14		Diane Keaton Decline / Year: 2.10		Early Hits: 100 – The Godfather (1972) 100 – Sleeper (1973) 100 – Love and Death (1975) Recent Misses: 5 – Because I Said So (2006) 8 – Mama’s Boy (2007) 22 – Mad Money (2008)
15		Katherine Heigl Decline / Year: 2.07		Early Hits: 80 – That Night (1992) 96 – King of the Hill (1993) 60 – 100 Girls (2000) Recent Misses: 14 – The Ugly Truth (2009) 29 – Life as We Know It (2010) 11 – Killers (2010)
	Check it out, Carlos is only 35!
35		Charlie Sheen Decline / Year: 1.25		Early Hits: 81 – Ferris Bueller’s Day Off (1986) 86 – Platoon (1986) 78 – Wall Street ( 1987) Recent Misses: 36 – Scary Movie 3 (2003) 37 – Scary Movie 4 (2006) 40 – Due Date (2010)

The Process

1. Rotten Tomatoes derives a score out of 100 for every movie based on the percentage of favorable reviews. They’re used for this exercise, unedited.

2. The decline per year and graphs were auto-generated. The decline is the slope coefficient for a standard linear regression, the graphs use a moving average for the trend line.

3. Only the top 500 actors are included (the 500 whose movies have earned the most at the box office), and actors must have starred in 10 movies.

Caveat

Since rate is a function of time, actors with a shorter career will have a much more volatile number. That means actresses like Julie Andrews and Diane Keaton have to work extra hard to sully all those 100′s from the beginning of their career.

The Extended List

The sequel has been around forever. It’s natural for an audience to want the stories they like to continue. When storytelling became marketable through print in the 18th century, sequels became a reliable stream of income. When movies emerged a century ago, the sequel seemed like a logical fit, both as a storytelling device and as a money-maker.

So how many movies made are actually sequels? Here’s my raw estimate:

Here’s how that compares to the number of all movies:

Sequel-making is on the rise, but it’s not keeping up with overall movie releases. The cost of production is dropping, so new movies are being released all the time. More and more aren’t sequels. Yet — people are watching sequels more than ever.

Originality in 1984

The Top Grossing Movies of 1984 (Domestic) 1. Beverly Hills Cop Original $235M 2. Ghostbusters Original $229M 3. Indiana Jones and the Temple of Doom Sequel (2) $180M 4. Gremlins Original $148M 5. The Karate Kid Original $91M 6. Police Academy Original $81M 7. Footloose Original $80M 8. Romancing the Stone Original $77M 9. Star Trek III: The Search for Spock Sequel (3) $76M 10. Splash Original $70M In 1984, original screenplays dominated the box office. Even Indiana Jones and the Temple of Doom and Star Trek III were based on original screenplays despite being sequels. These are stories that aren’t based on anything except the imagination of the screenwriters’ and producers’ imagination. Not books, not comics, not video games, not old TV shows (except Star Trek). The top 10 isn’t an anomaly either. Other hugely successful originals in the top 25 included Purple Rain, Amadeus, Tightrope, Revenge of the Nerds, Breakin’, Bachelor Party, Red Dawn, the Terminator, the Killing Fields, and Places in the Heart (among others).

Originality in 2007

The Top Grossing Movies of 2007 (Domestic) 1. Spider-Man 3 Sequel (3) $337M 2. Shrek the Third Sequel (3) $323M 3. Transformers Based on TV $319M 4. Pirates of the Caribbean: At World’s End Sequel (3) $309M 5. Harry Potter and the Order of the Phoenix Sequel (5) $292M 6. I Am Legend Based on Book $256M 7. The Bourne Ultimatum Sequel (3) $227M 8. National Treasure: Book of Secrets Sequel (2) $220M 9. Alvin and the Chipmunks Based on TV $217M 10. 300 Based on Comic $211M Fast forward 23 years. Not a single movie in the top 10 is based on an original idea. Every one is the continuation of a movie franchise, or an adaptation of a book, an amusement park ride, or a reboot of a TV show.

Looking at Sequel Money for All Years

For lots of reasons — built-in audience, existing marketing streams — sequels make more money than non-sequels. Even if they take up a smaller percentage of overall movies made, they’re a much safer financial bet for studios than original movies.

The shift between 1984 and 2007 is dramatic, but is it a continuing trend? To see exactly how often these studios are going the “safe” route, I looked at revenue for the top 50 grossing (domestic) movies of each year since 1980. (The top 50 seemed like a good cut-off since 65% of 2010′s movie revenue has come from the top 50 movies. That number has been as high as 80% in recent years).

The Process – Instead of manually tabulating what movies are sequels and what aren’t (that would take forever), I set up semantic structures based loosely on IMDb‘s and Wikipedia‘s movie pages, then analyzed a combination of keywords, similar web links, and related metatags (Wikipedia, for instance, allows entries to have a “next” or “previous” in the series). A spot-check revealed a very accurate categorization — even with the subtleties — although I’m sure it’s not perfect. An example: Batman Forever was categorized a sequel, but Batman Begins a remake.

Broken down by percentage:

For another time: the yellow original section in the last five years is more than 70% book, comic, TV, and video game adaptations.

What’s Wrong with the Creep?

Sequels are profitable. Embracing this new incrementalism makes sense for studios, and you can’t fault them for sticking with what works. The ubiquity of this logic was even the basis for the satire of Hamlet 2.

There are more movies than ever to see. So what’s the problem? If I want to want some schlocky big-budget sequel, I should be able to do so without derision, right? There’s nothing wrong with going to see whatever I want!

Wrong. The internet is making it easier to stream obscure movies that weren’t available a few years ago. But there are only a limited number of full-sized theaters. The movie-going experience is something unique that hasn’t been fully replicated at home yet. Until it is, the movies that make money will flood the theaters first, and it will be more and more difficult for you to find original movies to see on the big screen and be seen in their intended original art form. Will there ultimately be a backlash and shift? Not until the market dictates it. And you can bet that Disney will continue to pump out hundreds of millions in marketing to guarantee their expensive investments like Pirates of the Caribbean are seen before anything else. Some say that sequels can be original but I think they’re recognized with a lot of awards because there’s a lack of mainstream alternatives.

One scary scenario: studios start behaving like pharmaceutical giants. 10-15 years ago, Big Pharma started making super profits from its blockbusters in a very similar way to movies do today. Pharma in one of the most profitable industries in the U.S. has spawned two trends: more money is being spent on marketing than R&D and government lobbying is growing like crazy. (I’d argue that, although government involvement is to blame for some of this; it’s a result of lobbying and not proactive anti-trust measures).

Liken Pharma R&D to movie originality, and lobbying to media promotion and studio PR. If in 10 years, the movie industry mirrored the pharmaceutical industry of today, we’ll all be slaves to monster sequels and 3D eye-rapings.

The size of home TVs and entertainment systems is going up and the cost is going down (whether that’s a good thing is worth another post). So at least the original stuff will find an audience eventually, even if it’s not on a 50 foot screen. Social media could play a role in helping indie gems bubble up to popularity. I personally think the most likely catalysts to derail the sequel gravy train are activist studios with less financial focus like Participant Productions.

But until then, there are unofficially 86 big budget movie sequels in production and no end in sight.

At least there’s a good chance I’ll get to see There Will Be Blood 2: The Legend of H.W.’s Black Gold.