This is a link post to a recent talk by Julia Evans, popular programming blogger, called "So you want to be a wizard". I'm linking it on LW because I think it's one of the better takes I've seen on how to "apply rationality" to programming. A few interesting examples of things she discusses that I think will appeal to the LW audience:

• How she built a game to get better at Fermi estimate of how long different computer operations should take.
• How believing that she can achieve a gears-level understanding of things she doesn't yet know much about makes her a better programmer and debugger.
• How she uses "premortem"s to predict what will go wrong with projects.

While I suspect none of the above will be new to much of the LW audience, I think she does a good job of connecting each strategy she discusses to real-world examples that help you understand what it would actually look like to apply them.