Thursday, April 28, 2022

Hubris on Predicting the future

 If you've read my random thoughts you will see that statistics and predictions about the future are incredibly difficult. Even the prognisticators who claim to be predicting whats in store for the future tend to have no idea. That's even when they have all the stats and data in front of them. If they did then why would Tom Brady have been picked 199 in the NFL draft instead of number 1?

The New York Times newsletter discusses the NFL Draft and our hubris for predicting the future. The piece from the is below: 

Five-dimensional chess

What is the broader lesson here? The world is frequently messier and harder to understand than people acknowledge. We tell ourselves artificially tidy stories about why something happened and what will happen next.

The stock market rises or falls, and analysts proclaim a cause; in truth, they are often just guessing, as Paul Krugman, the economist and Times columnist, likes to point out.

On the subject of Covid, both experts and journalists have imagined it to be more predictable than it is. When schools reopened or certain states lifted mask mandates, you heard confident predictions that cases would rise. Often, they didn’t. The invisible, mysterious ebbs and flows of virus transmission overwhelmed every other factor.

In her latest column, The Times’s Zeynep Tufekci argues that public health officials have given flawed Covid guidance based on a paternalistic belief that they could see into the future. Zeynep’s main example is the F.D.A.’s refusal to allow young children to be vaccinated, based on what she calls a “five-dimensional chess” prediction that allowing childhood vaccinations will undermine vaccine confidence.

The most direct analogy to the N.F.L. draft is the hiring process elsewhere. Most employers still put a lot of weight on job interviews, believing that managers can accurately predict a candidate’s performance from a brief conversation. Research suggests otherwise.

Interviews can help people figure out whether they will like another person — which has some value — but not how effective that person will be at a job. If you think you’re a clairvoyant exception, you are probably making the same mistake the Jets did.

To be clear, the implication is not that nobody knows anything. Structured job interviews, which mimic the tasks that a job involves, can be helpful. And at the draft tonight, N.F.L. teams won’t be totally clueless: Higher draft picks have historically performed better than lower picks, but only somewhat.

The trouble is that human beings tend to overstate their ability to predict events. People who can resist that hubris — who can mix knowledge with humility — are often at a competitive advantage.

Official website