Pattern Recognition

Or: Statistics, with Commentary

The Ramp Rules: Statistics

Fun with numbers & patterns

How did the statistician die on his camping trip? He tried fording a river that was three feet deep on average.

I suspect most of us know that statistics can be misleading, but they're still one of the most important tools we have for understanding the world around us. Just ask any baseball fan. Or meteorologist. Or political analyst. (Actually, maybe not political analysts. Best to just slowly back away and not make eye contact with them.)

One of the biggest problems with statistical data actually comes from one of our greatest strengths: our pattern recognition ability. No computer in the world comes close to our ability to spot patterns, but sometimes that very ability backfires on us.

We often start looking for patterns in the data that aren't actually there. An absurd example is the correlation between pirates and global warming—as the number of pirates goes down, global temperatures go up. Sports superstitions are another good example: my favorite team won when I wore my lucky hat, so they'll lose if I don't wear it while they're playing.

Even if you entirely avoid the false pattern recognition pitfall, you still have a lot to consider.

McCoy Fields: Pattern Recognition

McCoy: Searching for Patterns

Humans also tend to remember the negative more often than not: when the weatherman's wrong, you remember that better than all the (many more) times he’s been right. Commercial weather channels often artificially inflate low chances of rain—say, from 5-10% up to 20%—so that if it doesn't rain it seems like a nice bonus. And if it does rain, well, 20% isn't nearly as low risk as 5%. We like it when errors fall in our favor.

So you're keeping your eyes open for suspect statistics. Great. Now you just have to keep on the lookout for the possibility that the people presenting the statistics are deliberately misleading you. Assuming they haven't just made up the data entirely (less common than you'd expect, since it's so easy to mislead with real data), you've still got to worry about them changing the scale of one of the axes of a graph, switching the axis themselves, or using one of a thousand different tricks within the math behind the data.

A quick example of another trick people can play with data: I recently saw a chart that was supposed to show Victorian England as a safer place to live than the modern day by comparing crimes per capita.

The problem with that? Society's definition of a crime. Crimes that we legislate against and prosecute tend to change over time. We currently have more laws today than at any other point in our history, so it's not surprising that we have more criminal charges per capita. In my example, what historians and criminologists actually use for comparison is murders per capita, because that's always a crime. By that metric, we're a lot safer than Victorian England. (Among other things, muggers today are less fond of garroting their victims.)

It's easy to play tricks like this—using categories of data that seem comparable but really aren't.

How can you look out for statistical shenanigans, if they're so easy to pull off? First, and easiest, try to stick with reliable, well known publications with solid reputations. (Of course, if you're far to one side or the other on the political spectrum, you're probably convinced they're a tool of the other side.)

Second, and much more difficult: Educate yourself. Learn to read data in a discerning matter and familiarize yourself with the subject matter. A good book to start with is Nate Silver's The Signal and the Noise.

Of course, you could just look for data and news sources that agree with what you already think. Confirmation bias is fun!

_________

I kid a lot with my pal Jeff at The Yard Ramp Guy. And now, also this: I'm seriously proud of my friends and colleagues at that fine company on their receiving the Blue Star award from Bluff Manufacturing, which has recognized them as a Gold Dealer of Excellence. Absolutely fitting, and absolutely right.

_________

Quotable

Okay, enough mush. Dear Yard Ramp Guystatistically I stay quotable each week more than you (but who’s counting?).

“There are lies, damned lies, and statistics."

— Mark Twain

Ig Nobel: Sometimes Nobel-Worthy

Or: Why Stinky Cheese Helps

telescope2Some science results in vital, society-altering work. Think about our creating vaccines, researching earthquakes, and discovering new construction materials.

Other science makes valuable contributions to our understanding of the universe around us—theoretical physics or marine biology, for example—and these very often produce society-altering work.

Then, of course, there is the science that doesn't quite have the same sort of impact, like magnetically levitating a frog or creating a chemical recipe to partially unboil an egg.

The Ig Nobel Prize is given out each September, around the time of the real Nobel prizes. It's broadcast every year on the Friday after Thanksgiving on NPR's program “Science Friday,” which makes for a much better use of my time than hunting Black Friday deals.

The awards either criticize (awarding a prize in “science education” to the Kansas State Board of Education for their stance on teaching evolution in schools) or to gently rib scientists for absurd research (testing the “five second rule” for eating food off the ground, which they proved to be false).

A few of my favorites:Discovering that the word “huh” or an equivalent seems to exist in every language.

  • doctorRealizing that if you attach a weighted stick to the rear end of a chicken, it then walks in the same way that dinosaurs were thought to have walked.
  • Measuring the friction between a banana skin and a shoe, and a banana skin and the floor, when we step on the banana.
  • Investigating whether cat ownership is mentally hazardous (I could have answered that).
  • Testing how reindeer react to seeing humans disguised as polar bears.
  • Finding that some people could run on water, if that water were in the Moon's gravity instead of Earth's.
  • Discovering the biochemical process by which onions make people cry.
  • Figuring out that leaning to the left makes the Eiffel Tower seem smaller.
  • Calculating the balance of physical forces inside a human ponytail
  • Discovering a way to extract vanillin (the active component of vanilla) from cow dung.

All of my favorites are in the real-but-absurd science category: the awards given out to people doing homeopathy or predicting the end of the world are grin-worthy…but not all that interesting.

experienceThe absurd science category sometimes becomes relevant: Sir Andre Geim won both the Ig Nobel Prize (he was the one to levitate the small frog with magnets) and the Nobel Prize for his work on graphene.

And the Ig Nobel award-winning study showing that malaria mosquitoes are attracted to Limburger cheese has been used to create large-scale mosquito traps in strategic locations in Africa.

Mostly, though, I just find that the real science is funnier.

_________

Quotable

Yes, Yard Ramp Guy—I remain blog relevant. You?

“The most exciting phrase to hear in science, the one that heralds new discoveries, is not ‘Eureka!’ (I found it!) but ‘That's funny…’"

— Isaac Asimov

The World is Still Round

But Kansas is Still Flatter Than a Pancake

 

Quick quiz: When did humanity discover that the Earth is round?

hand holding the earthIf you answered Copernicus or Galileo, you're off by a millennia or two. The Ancient Greeks actually discovered this.

To prove the theory, they created an experiment using two dry wells, hundreds of miles apart. Then, at noon on the same day, the Greeks measured the shadows at the bottom of the well to see if they were at the same angle. When they weren't, they used the difference in angle to figure out the actual size of the Earth…with remarkable accuracy.

(Trigonometry is important, kids. If you don't believe me, just ask Jeff over at The Yard Ramp Guy.)

Okay, another quick quiz: When did humanity forget that the world was round?

I bet the most specific thing you can think of is the Dark Ages or the Fall of Rome, right? Well, both are incorrect: We never forgot. To quote the brilliant Stephen Jay Gould, "There never was a period of 'flat Earth darkness' among scholars (regardless of how the public at large may have conceptualized our planet both then and now). Greek knowledge of sphericity never faded, and all major medieval scholars accepted the Earth's roundness as an established fact of cosmology."

FlammarionIn fact, we can trace the popularization of the idea that people thought the Earth was flat back to specific historians in the 1800s.

Sadly, this is pretty common stuff when we think about history. People are swift to assume that their distant ancestors (or at least other people's distant ancestors) were dumber than people today.

It really doesn't take that much effort to realize that idea is wrong, but most people won't even try to put in the effort.

(Really don't believe me? Then let's see you design an aqueduct with pen and paper.)

_________

Quotable

Oh, Yard Ramp Guy: what goes around comes around:

"’I'll follow him to the ends of the earth,' she sobbed. Yes, darling. But the earth doesn't have any ends. Columbus fixed that."

— Tom Robbins, Still Life with Woodpecker

Surrounded by Skeuomorphs

Fun with Tchotchke Design

Why do some molded plastic objects have molded plastic screw heads in them?

They're not real screws, and they serve no purpose in holding the object together. As it turns out, that's not just an aesthetic choice but something more.

They're called skeuomorphs—“derivative objects that retain ornamental design cues from structures that were necessary in the original.” These are things like fake woodgrain on a station wagon, a volume control panel on a computer that looks like an old school sound system with dials and knobs, or plastic chairs imitating wooden ones.

Ancient Greek stone architecture had small features called triglyphs and guttae that didn't serve any purpose in stone, yet had important roles when the ancient Greeks primarily worked in wood. (They were the carved ends of beams along with the pegs craftsmen used to secure them.) Also, potters mimicked expensive silver goblets using pottery and often inlaid clay pellets to resemble the rivets in the metal ones.

Though skeuomorphism has obvious applications in design, architecture, and art history, we’re seeing extensive use in the digital realms. Digital skeuomorphs are actually a fairly important design element: when you can make the user interface resemble whatever you want, it can be helpful in giving the user a point of reference.

The most prevalent digital skeuomorph is probably the shutter-click that camera phones and digital cameras make when you take a photo. In film cameras, the sound was caused by a physical function. In digital devices, the sound exists solely to let you know the photo was taken in digital ones.

Of course, not everyone supports the use of skeuomorphs in this way. Apple, for instance, has begun moving away from them entirely in favor of a more simplified, streamlined design style.

Personally, I'd rather be surrounded by skeuomorphs. I think they're charming. They can really make objects seem classy and not ostentatious. I have been accused—and rather fairly—of being resistant to change…once or twice, though. Maybe more. Not counting Maggie saying it several times a week. Or all the times my doctor's said it.

I can adjust to change if I really want to. I usually just don't want to.

_________

Quotable

Oh, Yard Ramp Guy: be the change you wish to see in, er, your yard ramps:

“Yesterday I was clever, so I wanted to change the world. Today I am wise, so I am changing myself.

— Rumi