~philosophicalhacker

This blog post clearly lays out a few approaches to systematically exploring the impact of confounding variables in your data. — Mai

How I'm improving my ability to properly estimate how long it'll take to complete a programming task. — philosophicalhacker

Because abstractions aren’t free, sometimes we’re better off duplicating code instead of creating them. Here I consider different views on this trade-off and argue for a new, brier score based approach. — philosophicalhacker

A fascinating analysis of whether or not diffs in a large repository are serially correlated across months and quarters. — nicholasbs