Project graveyards as guideposts

Jul 26, 2020 20:43 · 422 words · 2 minute read


Look for patterns in your past projects as guideposts for important decisions.


I was recently listening to the Capital Allocators Podcast, and someone finally put into words an idea that I’ve struggled to articulate - research graveyards. In a discussion about hiring quants, one of the things they look for is a “research graveyard”. The idea is that they want to see someone who has tried a ton of things, and can lean on that experience. In this case, they’re looking for approaches to finding alpha or modeling interesting phenomena in the stock market.

I think this idea can be applied to those of us who are not quants. I started wondering if I have my own “research graveyard”, and the answer is: kind of. Rather than research, I have a project graveyard.

Project graveyards

A project graveyard is the mysterious pile of semi-coherently connected wires on an electrical engineer’s desk, the list of abandoned repositories on a software engineer’s github, the forgotten folder of GBs of unfinished music on a producer’s hard drive.

It is easy for us to forget about all of these projects because, at least in my case, most of them are terrible. However, taken in aggregate, they form useful patterns. The trajectory of your projects likely provide some insight into what it is that you’re interested in doing, and sometimes life has a way of convincing you to not do those things. But the patterns can be tricky to find unless you have a large enough graveyard to draw from.


The question that I found myself asking was: what are the things that I just can’t help doing? Some of the things that net out of this are going to be things that you already know, but some of the more subtle patterns may provide an interesting perspective. For example, I greatly enjoy making music. One might suggest that I should follow this thread, and try to become a great musician - I don’t think I’ll ever give up hope that some day I’ll fill a stadium with my as-yet-unaquired musical talents. However, I don’t think this is a particularly interesting pattern.

A more interesting pattern might be the specific ways that I have combined my interest for music with my interest for software, or how I’ve combined those two interests with other interests. These combinations become increasingly complex, and the signal gets weaker as the combinations grow. But the signal that is there can provide useful guideposts to important decisions regarding your career.