Returning to the main graphical view, we notice that the diagram has changed a little.
Now, the two stochastic nodes, 窶廸umber of chocolate bars窶 and 窶弃roduction cost窶, just have a single circle in the 窶廨raphical view窶. This way one can easily see where the uncertainties are located in a model. This is also consistent with established influence diagram standards.
Taking a little closer look at our model, we can interpret it as follows: The number of chocolate bars produced are multiplied by the production cost per bar (hidden as a local value inside the 窶弃roduction cost窶 node) and the selling price per bar (hidden as a local value inside the 窶彜elling price窶 node). The total production cost and the total selling price is then added together to obtain the revenue. Clearly this is not correct. We should subtract the production cost from the selling price to get the revenue. We shall see how we can correct this error.
