How to Decide: Simple Tools for Making Better Choices


A guide to making better decisions.


Annie Duke of poker fame wrote a book about decision making.


  • quality of decision is different from quality of outcome, good decisions do not always lead to good outcomes and vice versa (judging a decision based on outcome is known as Resulting)
  • keep a decision journal to assess accuracy of decision and avoid Hindisght Bias and Memory Creep
  • three p's of judging a decision: preference, payoff and probability
  • take an outside view/opinion to get beyond own filters
  • to overcome analysis paralyss, figure out if you are in a Free Roll or Sheep in Wolf Clothing Decision situation
  • anticipate what might go wrong, do post a pre-mortem as well as a Backpath
  • avoid telling people your opininon to collect untainted opinions


Work backwards to figure out how you succeeded

what we believe and what we do

Making a decision tree of possible outcomes for a decision

Looks like a Free Roll but will destroy you over time

Something with low to no downsides

believe event was predictable after it happened

when items after outcome makes you feel you knew things all along

  • use the outcome of a decision to judge the quality of the decision
  • 4 types of outcomes
    • just desert: bad outcome, deserved it
    • worthy decision: good outcome, deserved it
    • bad luck: bad outcome, did not deserve it
    • dumb luck: good outcome, did not deserve it

All options are close. If this were the case, then it doesn't matter what option you go with

heighten or lower barriers to do something



  • the things that determine your life
    • your decisions and your luck
    • you can only control one of them
  • exercise: think of best and worst decisions, how do you judge them?
    • if your like most people, you judge quality of decision based on the outcome #star

1 - resulting outcome

  • called outcome bias or Resulting
    • use outcome to shortcut evaluating whether decision was good
  • four types of outcomes #ent.model #star
    • just desert: bad outcome, deserved it
    • worthy decision: good outcome, deserved it
    • bad luck: bad outcome, did not deserve it
    • dumb luck: good outcome, did not deserve it
  • watch out for bad luck and dumb luck, makes good decision seem bad and vice versa #star
    • even good decisions are not necessarly the best decision, always study decisions to learn from them
  • anecdote: hollywood pass on star wars, was it a bad decision or a good one with bad luck?

2 - hindsight is not 20/20

  • Hindisght Bias: believe event was predictable after it happened #star
  • Memory Creep: when items after outcome makes you feel you knew things all along
  • use knowledge tracker
    • hindsight bias causes us to lack compassion for others and ourselves,leads to we should have known it syndrome
  • keep a decision journal to counter this #star

3 - the decision multiverse

  • the paradox of experience

    • lots of experience is useful, one event is not helpful #star
    • eg. quit and try something, it fails, will weight on you
  • put individual experience in context of all other experiences

  • Decision Forestry #star

    • forest massacre, before making decision, tree of possibility, afterwards, you take a cognitive chainsaw and wipe everything out, making the one thing seem inevitable
    • glue branches back on to tree to make outcome more like the twig it is then the bow it became (create decision tree)
    • explore counterfactuals
  • many futures but one past, why it feels inevitable #star

    • we can counteract this by making a decision tree and exploring what could have been
    • important to also apply this to good outcomes

4 - the three ps, preference, payoff, probability

  • Step 1: Identify the reasonable set of possible outcomes

    • These outcomes can be general scenarios or be focused on particular aspects of the outcomes that you especially care about.

  • Step 2: Identify your preference for each outcometo what degree do you like or dislike each outcome, given your values?

    • These preferences will be driven by the payoffs associated with each outcome.
    • Gains comprise the upside and losses comprise the downside.
    • Include this information in your decision trees.
  • Step 3: Estimate the likelihood of each outcome unfolding.

    • As a start, use common terms that express probabilities. Don’t be afraid to guess.
  • Step 4: Assess the relative likelihood of outcomes you like and dislike for the option under consideration.

  • Step 5: Repeat Steps 1–4 for other options under consideration.

  • Step 6: Compare the options to one another.

  • The above is from the PDF accompaniment of How to Decide

  • preference

    • two people can have same outcomes but different preferences
      • eg. maximize career vs family
  • payoff

  • probability

    • you never go in with no information, don’t overlook territory between right and wrong
    • guess like archer, you might not hit the bulls eye but your aiming for it #word.anecdote
      • contrast this with pin the tail on the donkey where the kid has a blindfold, its just luck
    • take aim: what do you know, what can you learn to move forward #star
  • pro/con list not useful because no payoff, preference or probability

5 - taking aim at future: power of precision

  • people have different percentages in mind with words (eg. real possibility is can mean between 20 to 80 percent) #star
    • get better with precision
    • aiming at target vs bulls eye
  • give a range of uncertainty for estimate #star
    • make sure range isn’t too wide or narrow
    • the shock test: would i be shocked if it’s outside of range test
    • do the shock test exercise in the book #todo (Private)
    • if i were wrong, why?

6 - turning decisions outside in

  • it’s easier to solve other peoples problems #star

    • we are over confident in our own beliefs
    • our own circumstances filter everything
  • some filters

    • confirmation bias
    • disconfirmation bias
    • recency bias
    • illusion of control
    • availability bias
  • everything strengthens your own believes

    • you aim to protect your own beliefs
  • inside view vs outside view, outsider helpful, diff conclusion with same set of facts #star

    • eg
      • friend that always dates jerks, maybes it’s them
      • 90% professors think they’re above average, most people think they are good drivers
      • 50% end in divorce, 5% have prenuptial agreement
  • harder to see outside view when you are smart

  • what is true in general, outside of anyone's views? this is base rate

    • see to actively find out what other people know
    • people avoid disagreeing because they don't want to hurt your feelings but this actively harms you
      • eg. spinach in your teeth, people are unkind to not tell you
  • start with outside view and then inside view

    • marry the two narratives

7 - analysis paralysis

  • base decision on payoff
  • if decision repeats, easier to go fast #star
  • Free Roll: low downside
  • watch out for Free Donut: lottery ticket
  • Sheep in Wolf Clothing Decision: all options are close
    • test: if this were the only option, would you go with it? if so, then do it
  • some other factors
    • consider cost of quitting
      • sometimes its worth it to quit
    • exercise options in parallel
      • more expensive but you get to sample multiple outcomes
    • if you want to do research, do you have information that you can get that would change your mind? #star

8 - power of negative thinking

  • Behavior Gap: what we believe and what we do

  • techniques

    • mental contrasting: figure out what might go wrong and what to do
    • Pre Mortem
    • Backpath: how to succeed
    • pre-commitment, Ulysses Contrac: heighten or lower barriers to do something
    • dr evil: death by a thousand cuts, give a rational sounding reason to not get closer to your goal
    • categorical decision: make a decision that rules out other decisions (eg. i will avoid meat vs i'm vegan) #star
    • tilting: making bad worse, plan for this
    • hedging: control bad luck

9 - decision hygiene

  • if you tell people your opinion, they tend to agree #star
    • avoid doing this
  • study with visual test of card with three lines, two of which which are clearly not the same length
    • in one room, planted confederates who said all lines were equal
    • over 30% of test participants went with confederates

  1. Concepts


  1. area.decision-making