Table 6 shows the information presented in Table 5 with a calibration. The purpose of this exercise is to observe what happens when the composition of the groups is placed on an even footing; the exercise is experimental, aimed at considering whether this might be a helpful way to view chronological trends. The procedure for calibrating the groups adopted here is straight-forward. The year AD 80 is taken as a somewhat arbitrary watershed in the relationship between the Drag. 29 and the Drag. 37; it is a date around the time when the 37 is becoming more common than the 29. Years prior to AD 80 are counted back and given sequential minus scores; hence 79 is -1, 60 is -20, while years later than AD 80 are scored sequentially with plus numbers, so the year 111 has a score of 31; the year AD 80 is also scored as +1. With each site group the mid year of the date range of the group is calculated and the number score for that year established. So for the range 100-118 the mid year is 109, with a score of +29; the number of Drag. 37s present in the group is used to multiple +29 giving an aggregate number, so 8 examples gives a figure of +232; however, the number of Drag. 29s in the group (say 20) is also multiplied by the year score of +29 giving a figure of 580, but since this is after AD 80 this score for Drag. 29 is treated as a minus number. The difference between the two aggregates is the calibrated figure: -348. The scale of any minus calibrated scores in a year after AD 80 is a potential index of how old the samian might be in the group. When there are no Drag. 37s present in a group this is simply identified as a minus score, and equally when no Drag. 29s are present the group is allocated a + score: calibrations cannot be calculated in these cases since the number of examples of either Drag. 29 or Drag. 37 may, with such groups, simply be a function of sample size. The extent to which a group dated to after AD 80 has a minus score is a guide to the degree of residuality in a group, degree of curation of older vessels, or may suggest that the group is actually younger in composition than its ascribed dating suggests. Equally plus scores prior to c. AD 80 may indicate a precocious occurrence of the Drag. 37 or suggests the ascribed date of the group might be reviewed.
The method seems a reasonably reliable guide. It highlights the group from Hayton fort as having a precocious emphasis in favour of Drag. 37; this seems an unusual imbalance at this date, even for a newly established military site. Table 6 also highlights the group from Period 5a at Segontium as having a rather odd composition, with a ratio in favour of Drag. 29 rather than Drag. 37, despite its early second century date; this is commented upon by King and Millett (1993, 242; cf. Section 7.3.5, this report). The calibration also suggests that groups from major civil centres can have a high level of Drag. 29 vessels represented in contexts dating to the late first and earlier second centuries AD, as at Southwark, indicating continuing use of older decorated samian vessels and/or high levels of residual items.