A Blog by

Nick Cage Movies Vs. Drownings, and More Strange (but Spurious) Correlations

When Tyler Vigen was finishing law school a few years ago, he went on a search for identical twins—not in the real world but in data sets.

So he would find a study like this:

scan-sept1

And then keep searching until he bumped into a totally unrelated study, one that looked like this:

scan-sept

Then he’d put one plot line on top of the other, and—lo and behold!—at least wiggle-wise they looked close to identical.

A chart showing two separate line graphs, one showing spending on scientific research, the other suicides by strangulation
Chart courtesy of Tyler Vigen
Chart courtesy of Tyler Vigen

Seeing them coupled like this, it’s hard not to think, just for a second, that, Oh my god, science research causes hanging, strangulation, and suffocation.

I know, I know, that’s wrong. Correlations like these are total accidents. They don’t even imply causation, but when the match is compellingly close, you can’t help telling yourself a this-caused-that story. Tyler calls these “spurious correlations,” which is also the name of his blog. And his new book. They’re wonderfully silly and totally random. Unless …

What If … ?

… unless by some uncanny and particular set of circumstances, one data set did cause the other. I mean, it could happen, right? What looks absurd might, in some weird scenario, be true. But what would the story be?

Well, I looked very closely at some of Tyler’s graphs, closed my eyes, and told myself, Here’s what really (might have? could have?) happened …

A cheese pizza, an equals sign, and a bed with rumpled white sheets
Illustration Photographs by (L) Bon Appetit, Alamy (R) Dorling Kindersley, Getty
Illustration Photographs by (L) Bon Appetit, Alamy (R) Dorling Kindersley, Getty

 

A chart showing two line graphs, one charting per capita cheese consumption, the other people who died by becoming tangled in their bed sheets
Chart courtesy of Tyler Vigen
Chart courtesy of Tyler Vigen

So here’s what I imagine: Eating melted cheese in bed isn’t safe. There should be a warning label: “Do Not Bring Cheese Products Into Close Contact With Bed Linen” for the obvious reason that mac and cheese, a warm pizza, or any “soft” cheese like brie or camembert can slip out of one’s sleepy hands, drop onto the bedsheets and create an oily surface that is potentially dangerous—especially to a restless sleeper. Tossing and turning on “cheesed linen” creates extra knots and tangles, and as they say, “Tangles can kill.”

Had Will and Sandra Beckerman read such a warning, they might not have brought a Big Mamma Giant Sicilian pizza to their bed while binge-watching the History Channel’s seven-part series, Sunken Luggage of the North Atlantic. When they lost consciousness during episode seven, their sheets became so slick with mozzarella that it took only one left-to-right body flip to lock Sara’s torso into a knot so tight she lost circulation and ended up in the critical care unit at her local hospital. She survived, but as you can see from the rising black line on our graph, she was one of the lucky ones.

 

A portrait of Nicolas Cage, an equals sign, and a photo of a crowded swimming pool
Illustration Photographs by (L) Fotos International, Getty (R) Bernadett Szabo, Corbis
Illustration Photographs by (L) Fotos International, Getty (R) Bernadett Szabo, Corbis

 

chart showing the graphs of people who drowned by falling in pools and a separate graph of films Nicolas Cage appeared in
Chart courtesy of Tyler Vigen
Chart courtesy of Tyler Vigen

So here’s what I imagine: The Nicolas Cage Aquatic Society loves to meet, especially when there’s a new Nick Cage film to see and everybody’s all excited and can jump in the pool and organize themselves into rows, each person treading water, while Nick’s up on the big screen (which hangs above the diving board at the deep end).

Bobbing through a Nick Cage film is, as any Cage-aholic will tell you, the best way to watch our favorite actor at work. When Nick has a really busy year (he made four movies in both 2007 and 2009), so many fans show up poolside that—yes, it’s true—more people than usual fall into the pool and, well … drown. Forensic pathologists were surprised that in many years, most pool drownings took place at Nick Cage Aquatic Swim-ins, but hey, that’s understandable: In Moonstruck, when Nick first sees Cher all dolled up, standing in the moonlight, hair glistening, the night beckoning, he is so bewitched that he can barely breathe. Same for his fans. The ones who stopped breathing? They go up on this graph.

A group of motorcycle riders in action, an equals sign, and a photo of sour cream containers
Illustration Photographs by (L) Mirco Lazzari, Getty (R) Fred Tanneau, Getty
Illustration Photographs by (L) Mirco Lazzari, Getty (R) Fred Tanneau, Getty

 

Chart showing two separate line graphs, one featuring sour cream consumption, the other noncollision deaths on motorcycles
Chart courtesy of Tyler Vigen
Chart courtesy of Tyler Vigen

So here’s what I imagine: It all began when the medical journal Bovine Studies found that a daily serving of ordinary sour cream (four ounces a day) can “significantly” reduce the onset of male pattern baldness.
All over America, thin-haired men wondered how effective this remedy really was, so when Wilber Orlander, 41, a dental hygienist from Lorraine, Ohio, showed off his luxuriant hair on Oprah Winfrey’s television show and explained how, for seven years, he brought a cup of 2 percent-fat sour cream on his daily morning commute, ate half, then threw away the rest, his Orlander System became a must-do sensation. Which produced this unfortunate result: So many commuters threw so many half-finished, still creamy sour cream containers from so many motor vehicles on so many American highways after the year 2000 that sour cream became a traffic hazard.

Trucks and automobiles weren’t affected, but even a small splat could create what the graph calls a “noncollision transport accident.” “Avoid the Cream” signs appeared on many interstates, and in 2006, several highway departments dumped baked potatoes onto heavily trafficked areas to add a “clump factor.” (This technique was especially popular in 2006 and dramatically reduced the accident rate, as you can see from the graph, but potatoes-by-the-ton are expensive, and many highway departments had to switch back to salt in 2007. )

Sour cream, as the graph indicates, continues to be a significant traffic safety problem. A ban on driving with unlidded sour cream is now being considered in seven states.


See more of Tyler Vigen’s Spurious Correlations on his website.

7 thoughts on “Nick Cage Movies Vs. Drownings, and More Strange (but Spurious) Correlations

  1. NICE! You should turn this into a contest. Readers go to the spurious correlations site and come up with a theory for one of the seemingly unrelated matches. Winning entry’s author gets to appear on Radio Lab to discuss his/her findings.

  2. My first instinct would be whether a third factor might relate to both. It’s remarkable the weird correlations that arise that way. In my country sausage consumption tracks negatively with dementia rates. Not just over time but per capita in different cities. This turns out to be because poor people eat sausages here and poor people tend to die before they get old enough to get dementia. Could there be connections here? Maybe, maybe…

  3. Look at the past 30 years of movie releases and news events, including vhs and dvd releases after the original releases. All types, all channels.

  4. My first statistic curse began with the teacher comparing two graphs : one was the number of TV sets and the second the number of beds in psychiatric hospitals. And they coincide. TV turns us mad.
    Correlation is a proof of nothing, so many forget it !

  5. Kudzu- the type of factor to which you refer is a major subject in the field of statistical analysis. The term for such a factor is, appropriately, a “confounding variable.”
    Eric- I suppose the confounding variable related to the TV set-psychiatic bed correlation could simply be affluence. TVs are of course luxury items; wealthier regions tend to have more TVs than poorer regions because wealthy people obviously have the means to purchase this luxury item in greater quantities. “Psychiatric hospitals” other than drug rehabilitation centers no longer exist in modernized nations; ‘regular’ hospitals include psychiatric wings. The number of beds in a psychiatric ward at a hospital is less likely to relate to the number of psychiatric patients admitted to hospitals and more likely to correspond with well-funded hospitals. Hospitals with larger quantities of beds in their psychiatric wards probably have large quantities of beds in all of their wards, along with other surplus accommodations. Hospitals in wealthy areas tend to be able to afford more supplies, being wealthier than hospitals in poorer areas.
    Or the confounding variable could be even simpler: population density. In metropolitan areas, there are more households per acre. Even if the average household has only one TV, a city apartment building would have a lot of TVs in a small area compared to rural regions. As more people live in metropolitan areas, naturally, area hospitals must be equipped to accommodate the patients in densely populated areas (more people inevitably results in greater numbers of people being admitted to hospitals, though the rate does not need to be disproportionately high compared to sparsely populated areas).

Leave a Reply

Your email address will not be published. Required fields are marked *