This paper seeks to study how computerized decision-making techniques compare to one another, and what accounts for the differences. Although different algorithms tend to prefer specific formulations of fairness preservations, many of these measures strongly correlate with one another. In addition, fairness interventions might be more brittle than previously thought.