Decision tree basics
A decision tree lays out a choice the way you actually face it: first the decision, then the things you can't control, then the results. Reading left to right, you pass through three kinds of points — decision nodes (options you choose between), chance nodes (events that happen to you, each with some likelihood), and outcomes (where you end up). Drawing the tree forces the vague swirl of "should we or shouldn't we?" into an explicit list of paths, which is usually where the real insight happens.
How to build one here
- Write the decision as the first line — phrase it as a question.
- Indent one line per option, with the label in brackets:
[yes] Launch now,[no] Wait 6 months. - Under each option, indent the possible events and results. Keep nesting until every path ends in a concrete outcome.
- Export as PNG or PDF, or share the editable link with whoever you're deciding with.
From picture to decision: expected value
The tree becomes an analysis tool when you attach numbers. Estimate a probability for each chance branch and a payoff for each outcome, then work backwards: the value of a chance node is the sum of each outcome weighted by its probability.
Take the example above. Suppose there's a 60% chance of high demand (worth +$120,000) and a 40% chance of low demand (−$30,000). The expected value of launching now is 0.6 × $120,000 + 0.4 × (−$30,000) = $72,000 − $12,000 = $60,000. Do the same for the other branch and pick the option with the higher expected value — or notice that the answer depends on an estimate you should go test. Switch this tool to probability mode and the path probabilities are computed for you.
When a decision tree is the right picture
- One clear decision with a handful of options — not twenty.
- Uncertainty you can at least roughly quantify.
- Outcomes you can compare on one scale (money, time, points).
- A need to show the reasoning — trees are as much for the discussion as for the answer.
For plain hierarchies without choices — org charts, category breakdowns — use the standard tree diagram maker, and see the examples gallery for ready-made starting points.
Frequently asked questions
How do I add yes/no labels to the branches?
Start a line with the label in square brackets, like [yes] Launch now. The label is drawn on the connecting line, and yes/no labels get green/red styling automatically so the two paths are easy to follow.
What do the different node shapes mean in a decision tree?
In classic decision analysis, squares are decisions you control, circles are chance events, and the end points are outcomes. In this tool, branch points use squared corners while outcome leaves are styled with a filled accent so results stand out at a glance.
Can I calculate expected values with this tool?
Switch to probability mode and put probabilities in parentheses on the chance branches — the tool multiplies them along each path. Multiply each outcome's payoff by its path probability and add them up to get the expected value of a choice.