Tuesday, March 22, 2016

Bumper Sticker Thinking

I'm going to be discussing "correspondences" in the next few weeks. To introduce the topic I'm reposting part of a blog post by Scott Adams, creator of Dilbert, titled "Bumper Sticker Thinking."

http://blog.dilbert.com/post/141310196656/bumper-sticker-thinking

Perhaps you are familiar with this famous quote:

“Those who cannot remember the past are condemned to repeat it.” – George Santayana

The idea here is that patterns in history repeat. That might be true. Or it might be false. I have no idea. But I’ll tell you one thing I know with 100% certainty: 
People see patterns where there are none.
Oh, and people also fail to see patterns when they exist.
I don’t know whether or not historical patterns repeat themselves. All I know for sure is that the stuff we think are important patterns are mostly confirmation bias and cognitive dissonance. We see the patterns we want to see.
Don’t believe me? How about an example.
On Twitter, people have been hammering me over the fact that Clinton has higher poll numbers than Trump in a general election matchup. Therefore, say the helpful strangers on Twitter, Clinton will probably win, because polls usually do a good job of predicting the future. People believe polls have predicted well in the past – albeit imperfectly – and we should expect them to predict fairly well in the future. History repeats itself, right?
But here’s the thing. I have publicly and accurately predicted Trump’s rise since last summer. And I ignored polls to do it. Doesn’t my track record count as history that will repeat itself?
If polls say Trump will lose, and I say Trump will win, you have two histories that predict opposite outcomes. Which one do you pick?
Answer: The one that agrees with your existing opinion.
Looking at history is [bogus] if you don’t know which part of history matters. And you don’t. Because people are dumb like that.

No comments:

Post a Comment