THE CONCEPTS AND METHODS INVOLVED IN THE INVENTORY MODELING AND DECISION THEORY PERSPECTIVES ARE ILLUSTRATED IN TWO APPLICATIONS: (1) THE DECISIONAL PROBLEM OF DETERMINING THE OPTIMUM THRESHOLD PROBABILITY OF GUILT TO MERIT A CONVICTION IN CRIMINAL PROCEDURE; AND (2) THE STATISTICAL HYPOTHESIS THAT BLACK DEFENDANTS ARE TREATED THE SAME AS WHITE DEFENDANTS IN SENTENCING OR OTHER ASPECTS OF CRIMINAL PROCEDURE. INVENTORY MODELING INVOLVES MINIMIZING THE SUM OF HOLDING COSTS (TYPE 1 COSTS, E.G., THE COSTS OF CONVICTING THE INNOCENT IN THE CONVICTION CRITERION PROBLEM, AND THE COSTS OF DECIDING THAT THERE IS DISCRIMINATION WHEN THERE REALLY IS NOT IN THE RACIAL DISPARITY PROBLEM) AND OUTAGE COSTS (TYPE 2 COSTS, E.G., THE COSTS OF ACQUITTING THE GUILTY AND THE COSTS OF DECIDING THAT THERE IS NO DISCRIMINATION WHEN THERE REALLY IS). DECISION THEORY INVOLVES MAXIMIZING THE EXPECTED VALUE OF ACCEPTING OR REJECTING A HYPOTHESIS OR PRESUMPTION (E.G., THAT A DEFENDANT IS INNOCENT, OR THAT BLACK DEFENDANTS ARE TREATED THE SAME AS WHITE DEFENDANTS). BOTH PERSPECTIVES PROVE MEANINGFUL, BUT ONLY THE DECISION THEORY PERSPECTIVE SEEMS FEASIBLE IN THAT IT REQUIRES SIMPLE DATA TO BE APPLIED. FOUR GRAPHS SUMMARIZE THE BASIC IDEAS DISCUSSED IN THE ARTICLE. FORMULAS REPRESENTING BOTH APPROACHES ARE APPENDED. (AUTHOR ABSTRACT MODIFIED--LKM)