The 37 signals/Basecamp team once shared a post about using thick sharpies when sketching early designs. The beauty about sharpies is that they force you to focus on the core elements of what you are trying to build. That’s because you can’t obsess about the details the same way you can with thinner markers.
I find a similar approach to be useful when attempting to make long term predictions or sizing markets. Complex models look good and seem more accurate – but they often end up focusing us on details that don’t matter. Since the point of long term predictions is getting a sense of magnitude, there typically are a handful of drivers that actually move the needle.
So, the fewer the variables and simpler the assumptions, the better the prediction generally is.
Or, put differently, if you’re not able to share your model on a whiteboard using sharpies, there’s likely work to be done.