I was reading a magazine in the break room. I think it was a current issue of Smithsonian. It was an article about confusing similar trends with correlation. I'm not a statistician, but the idea intrigues me. There was an example comparing Nicolas Cage movies to shark bites. I suppose I remember that one because it's hard to say which one of those things is worse, but apparently if you put the data on a graph, the numbers match almost perfectly.
I don't remember if it was the amount of money the movie made or when it was made, but in this age of metadata in which there are so many numbers to compare, there was a distinct similarity between trends in completely unrelated fields. I also don't remember if the shark bites were in the U.S. or the world. But Nicolas Cage movies and shark bites aren't generally as bad as people think they are, so maybe there is a reason behind the matching statistical graphs.
We must remember that sequence doesn't determine causality and that if we have enough numbers we can get some of them to match purely by chance.
I often think that if I have two thoughts that connect I can easily add a random third that seems to fit right in there, and I wonder what value this skill has in my life other than to make me distractable.
I think if I randomly pulled out a word, an image and any one other thing, or two, I could write something interesting about their connection. But why do I want to do this, and for what end?
Since I wrote that, I've found the article again, in the break room, and it's not about shark attacks. It's about accidental drownings. And I think it was in Scientific American, if not Smithsonian. I'm going to resist the urge to check on this.
There's a book about connecting things that aren't connected, and this was one of the examples.
One way to connect things that aren't connected is to make stuff up.
No comments:
Post a Comment