08 / 08Think

Decide, then review

A job offer is accepted, a car is bought, a team picks one vendor over another. Six months later the result is in, good or bad, and the story of the decision quietly rewrites itself. A good outcome makes the choice look wise. A bad one makes the warning signs look obvious. Neither memory is quite what was known at the time. In a class this happens after an exam, when a study plan that worked gets credit it may not have earned. At work it happens in the review after a launch. This lesson writes the decision down before the result, imagines it failing while there is still time to change it, and reviews it later against the page.

In five minutesFor one decision you make this week, write three lines before you act: what you chose, what you expect to happen by a date, and the main reason. Put the date on your calendar.

Method · Decision journal and pre-mortem  /  Research psychology · Business

↺From last time · Write to find out

Answer these from memory before you start. Tap to check.

Why write without stopping or editing?

Stopping to fix sentences keeps the writing near the first thought. Writing straight through forces the next sentence to follow from the last, and the places where it can't are what you need to know.

Why put the notes away before writing?

With notes open, the writing copies them. Writing from memory makes you pull the ideas out and arrange them yourself, and in the study, tasks that required that kind of recall led to better scores.

What are the marked vague places for?

They show where the understanding runs out. Those few places are what to check in the source, which saves rereading everything.

01Where it comes from

Three findings sit under this method. In 1975 Baruch Fischhoff showed that knowing how events turned out changes how predictable people judge that outcome to have been [1]. In 1988 Jonathan Baron and John Hershey showed that people rate the same decision as better thought out when it happens to turn out well [2]. In 1989 Deborah Mitchell, J. Edward Russo and Nancy Pennington studied prospective hindsight: explaining a future event as if it had already happened [3]. The research psychologist Gary Klein built on that idea for the pre-mortem, described in Harvard Business Review in 2007. Before a project starts, the team imagines it has failed and writes down why [4]. Poker players call judging a decision by its result "resulting," a term Annie Duke brought to a wide audience in Thinking in Bets (2018) [5]. The decision journal itself has no single inventor we could trace. Its job is to keep a record that an honest review can check against.

Also called Pre-mortem, premortem, prospective hindsight, outcome bias, hindsight bias, resulting.

A good plan thwarted by chance

A well-laid plan is ever to my mind most profitable; for even though it be thwarted later, yet none the less has the plan been good, and it is but chance that has baffled the design; but he that has made a sorry plan has gotten, if fortune favour him, but a chance prize, and none the less has his plan been evil.

Artabanus, in Herodotus, Histories (The Persian Wars) Book VII, chapter 10 · tr. A. D. Godley

Xerxes has asked his council whether to invade Greece. His uncle Artabanus advises him to take time and think it over alone before deciding. Then he separates the plan from how it turns out. A good plan stays good if luck defeats it, and a poor one stays poor if luck rescues it. Herodotus wrote this in the fifth century BCE. Baron and Hershey measured the opposite habit in 1988.

Read it in full

02The method

Decision journal and pre-mortem

  1. Before deciding, write the decision, the date, and the options you turned down.
  2. Write what you expect to happen, with a number and a date where you can, and how sure you are.
  3. Write why: the reasons and facts you are relying on, and what you are unsure of.
  4. Pre-mortem: imagine it is a year from now and the decision failed. Write the likeliest reasons, quickly, without editing.
  5. Change the plan for any reason you can still prevent. Then decide.
  6. On a set review date, write what happened before rereading the page. Then compare, and sort what went wrong into reasoning, missing information, and luck.

03An example

What it looks like done

Worked on Accepting a job offer with higher pay and a longer commute

The decision, and the date

Take the analyst job at the logistics company. Turned down: staying in my current role. March 3.

What I expect to happen

  • Pay goes from $58,000 to $66,000.
  • Commute goes from 20 minutes to 45 each way.
  • After a year I'll still be glad I moved: 70% sure.

Why

  • The raise is $8,000 a year before tax. The extra driving costs about $2,000 a year in gas.
  • The work is closer to what I want to be doing in three years.
  • The manager answered every question directly in the interview.

What I'm unsure of

  • Whether the commute wears me down by winter.
  • Whether "some remote days" means one a week or none.

A year from now it failed. The likeliest reasons

  • The commute ate my evenings and I stopped exercising.
  • Remote days never happened.
  • The manager left and the role changed.
  • Fix now: get the remote policy in writing before I sign.

Review on (date). What happened

Set for March 3, next year. On that day, write down what happened first, then reread this page.

04Try it · 25 minutes

For a class

Before your next exam, write a one-page decision: how you will study, for how many hours, and the grade you expect, with how sure you are. Imagine the exam went badly and list three reasons. Fix what you can. After the grade comes back, write what happened before rereading the page, then compare. Note which parts of the plan helped, which didn't, and which came down to luck.

For work and life

Pick a decision due this month: a purchase, a hire, a schedule change. Fifteen minutes. Write the decision, your expectation with a number and a date, and your reasons. Then spend five minutes on a pre-mortem: it failed, so why? Adjust the plan. Put the review date on your calendar, and on that day write the outcome before rereading.

The blank page for this is in the Think workbook.

05Where it goes wrong

  • Judging by the result alone. A good decision can turn out badly, and a careless one can work. Review the reasons and the information first, then the result.
  • Writing the page after the fact. A record made once the outcome is known is shaped by it. Write it before.
  • A pre-mortem no one speaks up in. On a team, people hold back doubts about a plan the boss likes. Have each person write reasons alone first, then read them out in turn.
  • Rereading before recalling. Seeing the old page first anchors the memory to it. Write what happened, then compare.
  • Skipping the review. A journal never reread is a diary. Put the review date on the calendar when you write the entry.

06Research

What research has measured

Imagining the outcome has happened

One study found that treating a future outcome as certain changed how people explained it, while setting it in the past made little difference.

The numbers and sources
Effect
In two experiments, whether an event was set in the past or the future had little influence on the explanations people gave for it. Whether its outcome was certain had a strong influence. Explanations for certain events were longer, gave more reasons tied to specific episodes, and were written in the past tense [3].
Population
Two experiments. The abstract does not give a participant count [3].
Caveats
The study measured what kind of explanations people gave, not whether they were correct. It did not test the pre-mortem itself [3].

Judging a decision by its result

Early studies suggest people rate the same decision as better reasoned when it turns out well, even when they say outcomes shouldn't count.

The numbers and sources
Effect
With the same information the decision maker had, people rated the thinking as better, rated the decision maker as more competent, and were more willing to let that person decide for them when the outcome was good. People who were asked said outcomes should not matter, and used them anyway [2].
Population
Undergraduates in five studies of medical decisions and money gambles [2].
Caveats
Hypothetical cases rated by students, in one paper [2].

Hindsight bias

Shown that knowing how something turned out shifts people's sense of how predictable it was.

The numbers and sources
Effect
Across 95 studies and 252 effect sizes, the mean effect was .39. It was larger with almanac questions than with real events or case histories. Manipulations meant to reduce hindsight bias did not produce smaller effects [6].
Population
95 studies, 83 published and 12 unpublished [6].
Caveats
The size varied with how it was measured. Attempts to reduce it inside the studies did not work, so the best-tested defense is a record made before the outcome [6].

07Check yourself

Answer from memory first. The next lesson asks these again.

Why write the expectation down before the result is known?

Once the outcome is known, memory of what you expected shifts toward it. A page written beforehand keeps the real forecast, so the review compares the result with what you actually thought.

Why imagine that the decision has already failed?

Picturing failure as a fact makes people explain it, and explanations come out longer and more specific. On a team, it also gives everyone a reason to name doubts they might otherwise keep quiet.

Why sort what went wrong into reasoning, information, and luck?

Only bad reasoning calls for changing how you decide. Missing information calls for a better search next time. Luck calls for neither, and blaming the decision for it teaches the wrong lesson.

Sources
  1. Fischhoff B. Hindsight is not equal to foresight: The effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance. 1975;1(3):288-299. https://doi.org/10.1037/0096-1523.1.3.288
  2. Baron J, Hershey JC. Outcome bias in decision evaluation. Journal of Personality and Social Psychology. 1988;54(4):569-579. https://doi.org/10.1037/0022-3514.54.4.569
  3. Mitchell DJ, Russo JE, Pennington N. Back to the future: Temporal perspective in the explanation of events. Journal of Behavioral Decision Making. 1989;2(1):25-38. https://doi.org/10.1002/bdm.3960020103
  4. Klein G. Performing a project premortem. Harvard Business Review. 2007;85(9):18-19. https://hbr.org/2007/09/performing-a-project-premortem
  5. Duke A. Thinking in Bets. New York: Portfolio; 2018. Portfolio/Penguin, 2018
  6. Guilbault RL, Bryant FB, Brockway JH, Posavac EJ. A meta-analysis of research on hindsight bias. Basic and Applied Social Psychology. 2004;26(2-3):103-117. https://doi.org/10.1080/01973533.2004.9646399

MVIII puts the day in front of you each morning. Get MVIII