A variety of programming accidents, i.e., models and methods, including extreme programming, are examined to determine that each has a step that programmers find painful enough that they habitually avoid or postpone the step. This pain is generally where the programming accident meets requirements, the essence of software, and their relentless volatility.