Random story: I was in a team where we were tasked with creating a skills matrix containing everyone in the team for use by management to decide who would work on new projects. It started out simple and high level like having an "automated testing" skill, but then someone would decide to break that down further into "unit tests", "integration tests" etc. and rate themselves for each one as if the distinction was that important.
This then prompted people who had nothing to rate themselves for in those skills to create new skill categories and break those down into more and more little subcategories so they could show off all the skills they had. Pretty soon there was well over a hundred skills.
Your rating for each was something like it couldn't be over 5 out of 10 unless you'd use it in a production project (academic, side-project or open source use didn't count) and higher ratings required a specific number of years worth of use.
The skills matrix was never used for anything as far as I knew.
