Machine learning and big data are becoming increasingly important, and some say they can make predictions more accurate than ever before. But as risk expert Gerd Gigerenzer explained in his book Risk Savvy, the surprising truth is that we often get better results by using simple rules and considering less information in the real world.

In this episode, Gerd reveals concepts behind Risk Savvy and how misunderstanding statistics leaves us misinformed and vulnerable to exploitation. He explains how to make better decisions for our health, finances, and family, without the aid of an expert or supercomputer. Join him and Greg as they tackle the Turkey Illusion, Illusion of Certainty, and his other books Simple Heuristic and Ecological Rationality.

Episode Quotes:

Why do you believe that cultures with a positive attitude toward failure will have much better ways of making decisions than those who have a negative view of failure?

Positive error culture assumes that errors can happen. If an error happens, it is taken as a piece of information to find out what's going wrong. A negative error culture assumes that errors must never happen. If an error happens, the idea is to cover it up. Or, if that doesn't work, blame someone to find who is guilty.

Will processed data help doctors share more accurate information with the public?

It would work if doctors understood the result. But the problem is, if you don't understand the number that it spits out, you're lost. We have now COVID-19, and then we learned that a particular vaccine has 90% efficacy. Another one has 70%; what does it mean? Nobody knows what that means. In my observation, most people misunderstand that. And the misunderstanding started a long time ago, last year. So, this was before the vaccine. Dr. Fauci still hoped that it might be 50% efficient. And then there was an NPR program where someone explained what that means. He said it means that out of a hundred people who get vaccinated, 50% won't get the disease. By implication, the others will. That's exactly what it doesn't mean. It's a reduction of those who are not vaccinated and get the disease to those who are vaccinated and get the disease.

What are the incentives for news media to provide valuable information? Or does a desire for better ratings override the need for accurate information?

Half the primary goal is to increase attention, which is the case for many. Then they will try to not just provide bad news or good news, alarmists, less alarmist, but also use numbers that frightened people. One example is the use of additive risk increases as opposed to absolute risk increases.

How can you teach young people statistics in a fun and playful way, in ways that can be applied in real-life?

For instance, teach them easy, simple things like what a 30% chance of rain tomorrow means. They will find out that the parents probably don't know what the reference class is. Is it 30% of the time tomorrow, or the region, or maybe three meteorologists think it will rain and seven not? And so, that's something where young people can get their expertise and learn. They can do something better than maybe many adults.

Time code Guide:

00:01:42 Judgement and decision-making

00:02:58 Benefits and complexities of machine learning

00:06:54 Thoughts on lack of computational resources and cognitive limitations

00:12:26 Human heuristics and machine learning application

00:14:00 How the environment impacts decision-making

00:15:00 The Turkey Illusion

00:20:11 Investing and some concepts from the book Ecological Rationality

00:22:51 Positive error culture

00:25:57 Illusion of Certainty

00:28:54 Basing inference, conditional probabilities, understanding confusion, matrices,

false positives and false negatives

00:32:29 How can experts like doctors process data better

00:34:22 Additive risk increases and absolute risk increases

00:36:28 Striking a balance between people thinking for themselves and trusting experts

00:42:37 How can young people learn statistics in fun and applicable ways

Show Links

Guest Profile

Gerd Gigerenzer Profile at Max Planck Institute for Human Development
Gerd Gigerenzer on LinkedIn

His Work

Gerd Gigerenzer on Google Scholar
Risk Savvy: How to Make Good Decisions
Ecological Rationality: Intelligence in the World
Rationality for Mortals: How People Cope with Uncertainty
Gut Feelings: The Intelligence of the Unconscious
Calculated Risks: How to Know When Numbers Deceive You
Adaptive Thinking: Rationality in the Real World (Evolution and Cognition)
Simple Heuristics that Make Us Smart (Evolution and Cognition)
The Empire of Chance: How Probability Changed Science and Everyday Life
Cognition as Intuitive Statistics

Risks and Uncertainty: Understanding Data and Making Rational Decisions feat. Gerd Gigerenzer

unSILOed with Greg LaBlanc