The big percentage answers a specific question, and reading it as anything else causes most of the confusion. It is the share of simulated paths that end above the level you asked about, over the horizon you selected. Change the level or the horizon and the number changes with it.
The curve below it is the whole distribution, not decoration. Its middle is the median outcome, its width is how uncertain the model is, and the shaded area is where most paths land. A wide curve on a volatile name is the model being honest, not the model being useless.
The horizon selector matters more than people expect. A 72% at 30 days and a 72% at 365 days are different statements about the world: the second one has far more room for things to go wrong, and the band around it is wider for exactly that reason.
The confidence label describes how much history sits behind the estimate, not how sure we are of the outcome. High confidence on a liquid, long-listed company means the inputs are solid. It never means the forecast cannot be wrong.
One habit worth building: before reading the number, write down what you expected it to be. If your guess and the model disagree, that gap is the useful part of the screen, and it is where a good discussion in class starts.
Educational material. Not investment advice.