Joe Tharakan | We are all idiots

Cup O'Joe | So let’s stop pretending we can predict the future

By Joe Tharakan 02/20/17 11:06pm

In high school, I used to participate in extemporaneous speaking. We were given a question on current events; something like, “How will Brexit impact the UK economy?” might be asked today. We were expected to make some argument or prediction, back it up with relevant articles and evidence and present it to a judge in a 7 minute speech.

Of course, when I was in high school, I wasn’t asked about Brexit, because it hadn’t happened yet. I was asked to make all sorts of predictions on events in the news at the time. For example, I was asked about the United States debt-ceiling crisis, the Syrian Civil War and the Arab Spring. I would prepare by reading publications like The New York Times or The Economist to get a better understanding of the issues, so I could make an accurate prediction.

I saved some of the outlines I wrote, and I recently looked at them with the benefit of knowing the outcome. And I was so wrong about everything.

I predicted that the debt-ceiling crisis of 2013 would be resolved without a government shutdown. I was wrong.

I predicted that the United States would take some military action against Syria after the country's president, Bashar al-Assad, used chemical weapons on his own people. I was wrong about that too.

I predicted that the Arab Spring would be a wave of successful revolutions in the Middle East that brought democracy and stability to the nations in the area. Now, Libya is a failing state, Syria is in the midst of an ugly civil war that is spilling over into Iraq and Yemen has devolved into sectarian conflict. I was extremely wrong.

If I had to guess how many of the predictions I made ended up being correct, I’d estimate a dismal 30 percent. So much for expertise.

“Hey Joe,” you might say, “we know you’re an idiot, but that doesn’t mean the rest of us are!”

Well, I’d argue this trend isn’t just confined to me. How many of you predicted that the New England Patriots would comeback from a 28-3 deficit to win the Superbowl? Or that the Cleveland Cavaliers would come back from a 3-1 deficit? Or that the Chicago Cubs would do the same thing? (We’ve had a lot of sports comebacks lately, huh?)

And if you did predict any of those comebacks, was it based on any evidence? Or was it blind faith and fandom? This question is important, because any good prediction model needs to have some reliable process, or it likely won’t reach the correct outcome again.

Anyone who has ever read "Moneyball" by Michael Lewis knows that baseball scouts, who are supposed to predict which players should be stars and which should be busts, often fail miserably. It’s not that they aren’t trying or that they aren’t experienced — these are people who know as much about baseball as anyone can.

After the success of "Moneyball," Michael Lewis wrote a book called "The Undoing Project," which expands on ideas originally formulated by psychologists Daniel Kahneman and Amos Tversky. To sum them up in a sentence: We are all fundamentally flawed and imperfect, and everything we feel certain of is wrong. Yes, experts do predict results more accurately than the average person, but not by nearly as much as you’d expect.

When I was making my bold predictions in high school extemp, I was confident in my arguments. I cited articles and made my conclusions. In retrospect, I realize I could have easily found articles for the opposite argument. So why did I pick one side over another? More than anything, I had a gut feeling that I was right. Then, once I decided I was right, based on instinct, I backed up my conclusions with whatever biased and selectively sampled evidence I could find.

As it turns out, our gut feelings on literally anything, from predicting complicated current events to sports outcomes to who is “Most Likely to Succeed,” are super wrong and usually based on next to nothing. Yet somehow, we feel like they are completely correct.

The next time some expert predicts some random outcome that looks ridiculous in hindsight, perhaps now we can understand how they got it so wrong. More importantly, we should recognize that nobody is good at making predictions, least of all ourselves, and we should cast a healthy amount of doubt on the things we “feel” are certain.

JOE THARAKAN is a College senior from The Bronx, N.Y., studying Biological Basis of Behavior. His email address is jthara@sas.upenn.edu. “Cup O’Joe” usually appears every other Tuesday.