Predictive policing seeks to allocate scarce resources where and when they are most needed. Yet analysis is often based on recorded crime data that typically understate the concentration of crime at the same places or against the same people. This study outlines a means of developing more accurate estimates, termed the Recorded Repeats Adjustment Calculator (RRAC), and applies it to burglary data for Metro Vancouver. Whereas repeat burglaries constituted 20 percent of recorded burglaries, after adjustment they were half of burglaries. Moreover, households with five or more burglaries accounted for less than 1 percent of recorded but, after adjustment, one in five actual burglaries (21 percent). These results are closer to those found by crime victim surveys but still likely to be conservative. The study aspires to produce a tool for analysts that produces more accurate information on crime's concentration and, thereby, more informed crime control efforts. (Published Abstract)