A forecast is usually delivered as a number. Thirty percent. The number is the easy part, and it is almost useless on its own, because a month later nobody, including you, can reconstruct what the thirty percent was made of. Was it a base rate you trusted? A gut adjustment you have since forgotten? The number does not say. It is a conclusion with the argument thrown away.
A board keeps the argument. Instead of one figure, you get the whole chain: the base rate you started from, the sources you pulled, the adjustments you made, and the arithmetic that ties them together. The forecast is still there at the end. But now it is the last node in a diagram you can read, not a claim you have to take on faith.
Nodes are the nouns, wires are the verbs
The canvas has a small vocabulary, on purpose. A circle holds a value: a base rate, a figure, a guess. A square does something to values: it sums, multiplies, divides, or turns a series into a rate. A wire carries a number from one thing to the next. That is nearly the whole grammar, and it is enough to write most of what a forecast needs to say.
Because the pieces are visible, the shape of your reasoning is visible too. A forecast that leans entirely on one shaky input looks lopsided on the canvas, one lonely wire doing all the work. A forecast built from several independent estimates looks like what it is. You can see the structure before you even read the numbers.
A spreadsheet hides the logic in cells. A board puts it on the wall where you can argue with it.
The board recomputes, so it never goes stale
The other thing a number cannot do is change its mind. Wire a new figure into an input and everything downstream updates the moment you let go. The base rate moves, the pushes re-apply, the final probability lands somewhere new, and you watched it happen. Nothing is frozen at the value it had the day you typed it.
This is what makes the board a live model rather than a drawing of one. You can stress it: what if the base rate is half what I assumed? Drag it and see. The forecast that survives a few of those questions is the one worth trusting.
Why the receipt matters later
The real payoff comes when you reopen the board weeks on. The question has resolved, or new information has landed, and you want to know whether you were right for the right reasons. The board tells you. Every assumption is still sitting there, labeled, wired to what it fed. You can find the one adjustment that was wrong and see exactly how much it moved the answer.
That is the difference between keeping a forecast and keeping a score you cannot learn from. The number was never the artifact. The reasoning was, and the board is where it lives.
Try it
Sketch a real question you are chewing on. Drop a base rate, wire in two or three things that push it up or down, and end on a combine. When you are done you will have a probability, and next to it, the reason you believe it. Open a board.