Decision Tree Node Analysis

For any node in a decision tree, you can mouse over and see a tooltip with lots of metrics. Here is an example.

Mouse-over tooltip on the Call dealership decision tree node showing Expected Value -420, Total Payoff -720, and Marginal Probability 0.25 at that point.

Also, notice that the Expected Value metric is shown in light gray color on every node for a quick understanding of what's going on for each node.

Decision tree with Expected Value labels (x-bar values) shown in light gray text on every node and a callout reading 'Expected Value is displayed on all nodes.'

You can choose to hide these gray color metrics for all nodes if that looks noisy. From the context menu of the diagram, simply choose "Hide Metric from all nodes".

Diagram context menu with the Hide Metric from all nodes command selected, used to remove the gray metric labels from every node at once.

If you want to show Total Payoff instead of Expected Value for all nodes, you can do that too from the context menu.

Diagram context menu in Decision Tree Analyzer with the Show Total Payoff option selected to replace the expected value labels with total payoff on every node.

If you want to show or hide metric for a specific node, rather than all nodes, you can do that too. Just select a node and from the context menu of the SELECTED NODE, you will find options for showing or hiding metrics.

Per-node context menu offering Show Metric and Hide Metric commands that act only on the currently selected decision tree node.

In addition, you can get a complete risk profile for that node by expanding the node analysis tab. For example, in the following decision tree, we can see the node analysis of the selected node "Go by car".

Node Analysis tab expanded for the Go by car action, showing the full risk profile chart for that single node in the decision tree.

If you have a probability distribution in any selected node or children of a selected node, then a full Monte Carlo Simulation will be performed to generate the risk profile. Here is an example screenshot where we can see that the "Car break down" node has a Normal Distribution.

Node Analysis panel running a Monte Carlo simulation because the Car break down child node uses a normal probability distribution, producing a smooth risk profile.
Last updated on Jan 7, 2026