DecisionBranch
Build a tree

Decision trees for life's big choices

Take the job, move city, buy or rent: lay it out, put your own numbers in, and see which option holds up when you are unsure.

Build a decision tree – freeSee the example

The problem

  • The pros-and-cons list never adds up

    Twelve pros and nine cons tells you nothing about which matter. A tree makes you say how likely each outcome is and how much it is worth to you.

  • You already know what you want, and you are not sure it is right

    Put your gut's numbers in. If the recommendation matches, good. If it flips at a probability close to yours, that is the thing to find out before deciding.

  • Regret

    The switch point tells you how sure you would have to be. A decision made with that number in view is one you can stand behind if it goes badly.

Worked example

Relocate for the job?

Open in the editor Describe your own

The offer is good but it means moving. You can take it and move, ask to start remote and see, or stay. The payoffs are your own scale for how the next three years go, career and life together; the costs are the move and the goodwill spent asking.

RecommendedTake it and moveExpected value 52,000 · 2,500 ahead of Ask to start remote
Relocate for the job?−12k60%40%−2,00050%50%−12kRelocate for the job?EV 52,000→ Take it and moveTake it and moveEV 64,000Settle in well90,000Do not settle25,000Ask to start remoteEV 51,500They agree60,000They insist on the moveEV 43,000→ Move after allMove after all55,000Decline the offer30,000Stay where you are30,000

Switch point. If the chance of Settle in well under Take it and move falls below 56.2% (it is 60% now), Ask to start remote wins instead.

0%56.2%100%

Risk profile. Under the recommended policy the chance of ending up with a loss is 0%; the worst case is 13,000 and the best is 78,000.

40%13,00060%78,000

The recommendation is a narrow one, and the switch point says why: the answer depends on how sure you are that you would settle in. That is a question you can actually work on, with a visit and a few conversations, before you sign.

The outline behind this example
# Analytical: relocate for the job offer?
Relocate for the job?
  Take it and move (cost 12000) # moving, deposits, two flights home
    Settle in well (p 0.6): 90000
    Do not settle (p 0.4): 25000
  Ask to start remote (cost 2000) # a slower start, some goodwill spent
    They agree (p 0.5): 60000
    They insist on the move (p 0.5)
      Move after all (cost 12000): 55000
      Decline the offer: 30000
  Stay where you are: 30000

What you get

  • Your numbers, your scale

    Payoffs can be money, or points out of 100 for how life goes. The maths is the same.

  • The recommendation and the margin

    Which option is best on average and by how much, so you know if it is a clear call.

  • The switch point

    How sure you would need to be for the answer to change. Regret-proofing in one number.

  • Risk on the page

    The chance of the bad outcome under each option, not hidden in an average.

  • Flow trees for the messy questions

    Some decisions are questions, not gambles. Walk mode handles those too.

  • Private by default

    Trees are private unless you publish them. Five private trees free.

How people use it

  1. 01

    Job offer

    Take it or stay, with the chance the new role works out and what each outcome is worth to you. See the switch point before you negotiate.

  2. 02

    Move city

    Move now, try it remote first, or stay. Put a cost on the move and a probability on settling in; the tree does the rest.

  3. 03

    Buy or rent

    Buy, rent another year, or rent and invest the deposit, with your own view on prices and rates. Sweep the price growth and watch where the answer flips.

Templates for this work

All templates · How to make a decision tree · Expected value and decision trees

Questions

What if I do not know the probabilities?
Guess, then sweep. The switch point tells you whether the guess matters: if the answer only flips at 20% and you think it is 60%, you are done.
Do the payoffs have to be money?
No. Use any scale that is consistent within one tree, such as points out of 100 for how the next three years go. Expected value works the same way.
Can I keep it private?
Yes. Trees are private unless you make them public, and you do not need an account at all if you are happy to keep the link.

Start with the example, or your own decision.

The editor is free and needs no account. Describe the decision in a sentence and the AI builder drafts the tree.

Open the exampleDescribe your decision