A form of YouAintGonnaNeedIt, where an unjustifiable amount of time and effort (from a ReturnOnInvestment point-of-view) is spent thinking about and coding for extremely unlikely situations where failure would not be a major problem anyway.
Also a frequent mistake made by students in any science, where they assume that all the digits of precision provided by their calculator are significant, even though it is vastly more precise than the measurements the calculation was based on.