The key ideas

Risk Savvy: How to Make Good Decisions

Gerd Gigerenzer, 2014

A pill “doubles” the risk, a screening “cuts deaths by twenty percent”, and nobody tells you what to do with that. Gerd Gigerenzer shows that the problem isn't that people are stupid. It's how the numbers are put to them. You'll learn to tell a risk you can calculate from uncertainty, to turn percentages into people, and to ask the questions that make a frightening number make sense.

Who it is for

For anyone who reads news about risks and can't tell whether to be scared. And for anyone who makes decisions about health, money and safety and wants to understand the numbers instead of taking them on trust.

Knowledge is the antidote to fear.

Thirty percent of what?

An evening weather forecast on TV: tomorrow there's a thirty percent chance of rain. You nod and decide whether to take an umbrella. Now a question: thirty percent of what? Of the time, of the area, of the days? Every listener answers differently, and every one of them is sure they understood the forecast correctly. The forecaster gave an exact number but didn't say what it referred to, and the precision turned into fog.

This is the kind of question Gerd Gigerenzer opens Risk Savvy with. He has spent many years studying how people make decisions. His research center in Berlin is called the Harding Center for Risk Literacy. The author's point is simple and uncomfortable: we're surrounded by probabilities, but almost nobody taught us how to read them. That goes for patients, for journalists and, as it turns out, for many doctors too.

And yet the book doesn't blame the reader. Gigerenzer argues against the view that people are simply stupid and can't cope with risk. Often the culprit is the form the numbers come in. Change the form, and a problem that left people stumped almost solves itself. Which means understanding risk can be learned, just as we once learned to read.

Ahead are a turkey that was sure of tomorrow, a pill that frightened a whole country, and America's roads after September 11. And at the end, a few questions that help you make sense of anyone else's number.

Roulette and life are different games

In a casino, everything is fair in one sense. A roulette wheel has a known number of slots, and the odds of any bet can be worked out in advance. Gigerenzer calls this kind of situation risk: all the outcomes and their probabilities are known.

Now a different situation. You're deciding whether to take a new job, whether to invest money, whether to move to another city. Here you can't list all the outcomes, let alone give them exact odds. This is what the author calls uncertainty. Almost the whole world outside the casino works this way.

Picture Anton, choosing between two job offers. He opens a spreadsheet, gives each option scores and probabilities, and gets a neat answer correct to one decimal place. The spreadsheet looks convincing. But Anton made up the numbers himself: nobody knows whether the new company will grow or whether the team will gel. This is Livria's example, not a case from the book, but it shows what the author is warning about.

The mistake is confusing one world with the other. When methods built for roulette are applied to uncertainty, the calculation looks rigorous but rests on assumptions nobody has checked. As the author sees it, the world of risk calls for logic and statistics, while the world of uncertainty also calls for experience and simple rules.

Gigerenzer explored how such rules work in detail in his book Gut Feelings. What matters here is the first step: before you calculate, ask yourself which of the two worlds you're in right now.

Read in Livria

About the author

Gerd Gigerenzer is a German researcher who studies how people make decisions, and director of the Harding Center for Risk Literacy in Berlin. For many years he has taught doctors and the general public to understand statistics.

All books by this author: Gerd Gigerenzer

Terms from the book

  • Thanksgiving
  • PSA
  • Uncertainty
  • DNA
  • Harding Center: The Harding Center for Risk Literacy is a research center in Berlin, founded in 2009. Gigerenzer was running it when he wrote the book.
  • Turkey illusion: An image from the book, not a scientific term: the mistake of someone who takes uncertainty for a calculated risk and is sure they know every possible outcome.
  • Illusion of certainty: The author's phrase for the belief that there is an answer with no “maybe” in it, for example a test that is one hundred percent reliable.
  • HIV: Human immunodeficiency virus. The test for it is very accurate, but like any test it occasionally gives a wrong result.
  • Committee on Safety of Medicines: A British government body that oversaw the safety of medicines in the United Kingdom until 2005.
  • Thrombosis: A blood clot forming inside a blood vessel. The clot can block the flow of blood, which is why thrombosis is dangerous.
  • Relative risk: How much a risk has changed compared with where it started, as a multiple or a percentage. It says nothing about how many people are affected.
  • Absolute risk: How many people out of a given number will actually experience an event, for example one in seven thousand.
  • Mammography: An X-ray examination of the breast, used to find breast cancer early.
  • Gynecologists: Doctors who look after the health of the female reproductive system.
  • Natural frequencies: The method the author argues for: showing probabilities as a number of people out of a clear group, for example nine out of a thousand, instead of percentages and conditional probabilities.
  • Giuliani: Rudy Giuliani, mayor of New York City from 1994 to 2001. In 2007 he ran for the Republican nomination for president of the United States.
  • Five-year survival: The share of people who are still alive five years after being diagnosed. It goes up when a disease is found earlier, even if people don't live any longer.
  • Mortality: How many people in the whole population die of a particular cause over a particular period.
  • Dread risk: The author's term for an event in which many people die at once, such as a plane crash or a terrorist attack. Such events frighten us more than the same number of deaths spread out over time.
  • Defensive medicine: Ordering tests and procedures the patient doesn't need, mainly so the doctor is protected from a possible lawsuit.
  • Checklist: A list of required steps worked through in order so that nothing is missed. It came to medicine from aviation.
  • Error culture: A phrase from the book: how an organization treats mistakes. In a positive culture they are reported and the system is changed. In a negative one they are hidden for fear of punishment.
  • Nudging: An approach that quietly steers people toward the right choice, for example through a default option. Gigerenzer argues against it and proposes teaching people to understand risks instead.

All terms

Similar books