A gap in scientific reporting

It is common to report optimality gap values in the computational results section of papers related to the solution of optimization problems. Several years of refereeing experience have taught us that these gaps are often improperly defined or incorrectly computed. In this note, we offer some comments on this topic.