What I want here is narrower. Three things go on this site.

Papers, with their reproducibility artifacts. Every published or submitted manuscript gets an entry with the analysis code, the variable construction log, and an explicit note on which observations were excluded and why. If a result depends on a sample restriction, that restriction should be visible without emailing me.

Method notes. Short pieces on decisions I had to make and could not find cleanly documented anywhere: when a configurational analysis adds something a regression does not, how to handle a collapsed response category, what a robustness battery should actually cover.

Reading from the climate-venture literature. I am translating the MIT climate and energy entrepreneurship framework into Korean, which means reading a great deal on the subject. The useful parts land here.

Nothing about my week. Nothing that would not survive being read by a reviewer.