Table 1: Distribution of duplicates identified between 2015 and 2020 by linkage criteria.

Mapping criteria	Matching pairs
High probable	N	Within category (%)	Overall (%)
exact RSA ID, *simscore(surname) >=0.85, simscore(date of birth) >= 0.95	10 070	86.21	3.47
exact RSA ID, simscore(firstnames) >= 0.85	748	6.40	0.26
exact RSA ID, simscore(date of birth) >= 0.95	640	5.48	0.22
exact RSA ID, switch(exact surname, exact firstnames)	206	1.76	0.07
exact RSA ID, simscore(surname) >= 0.85	17	0.15	0.01
sub totals	11 681	100.00	4.02
Probable
exact firstnames, exact surname, exact date of birth (incl RSA ID check)	80 386	77.86	27.70
switch (exact surname, exact firstnames), exact date of birth	22 861	22.14	7.88
sub totals	103 247	100.00	35.57
High possible
exact date of birth, exact firstnames, simscore(surname) >= 0.85 (incl RSA ID check)	67 102	38.27	23.12
exact surname, exact date of birth, simscore(firstnames) >= 0.85 (incl RSA ID check)	59 622	34.01	20.54
exact surname, exact firstnames, simscore(date of birth) >= 0.95 (incl RSA ID check)	42 324	24.14	14.58
exact firstnames, simscore(surname) >= 0.85, simscore(date of birth) >= 0.95 (incl RSA ID check)	6 273	3.58	2.16
sub totals	175 321	100.00	60.40
Grand Total	290 249		100.00