This webinar will discuss the definitions of each type of failures along with their key differences: random failures are included in probabilistic failure analysis and systematic failures require changes in operating procedures. It will look at the results of a survey where respondents classified real failures, showing that many failures classified as systematic which would result in an unrealistically low failure rate and possibly dangerous designs. We will then discuss why field failure collection systems must count all real failures until it is shown by new failure records that a change in a procedure is effective in eliminating a type of failure no matter if what it is called.