decision tree

One trigger at the root, then a branch per condition, fanning forward and never rejoining. Every leaf is one complete outcome, so you can hang a count on it: of the last 16 times, this is where nine of them ended. That makes the tree a way to choose and also a way to sort past cases, which is how you test whether the condition you blame is really the one splitting your outcomes.

One trigger at the root, then a branch per condition, fanning forward and never rejoining. Every leaf is one complete outcome, so you can hang a count on it: of the last 16 times, this is where nine of them ended. That makes the tree a way to choose and also a way to sort past cases, which is how you test whether the condition you blame is really the one splitting your outcomes.

Example

Best when you suspect one condition decides most outcomes, and you have past cases you can sort onto the leaves.