Finding

The candidates fall into two disjoint families, and no candidate is in both

The matrix produced one structural result that no individual candidate file could show. The candidates fall into two families and no candidate belongs to both. The first family is documented households with the wrong shape: C(ii) Berkhamsted, Hitcham, Ackworth and the eliminated Eythorne each have a father, sometimes a mother, and children with dates - what they lack is the right children. C(ii) has a John and a Richard in the right order but no Mary in the parish at any date; Hitcham has a Mary satisfying the hard bound and nothing else at all; Ackworth has the parents right and the children wrong. The second family is documented men with no household at all: Candidate B, Cheddington 1608, A(1603), Candidate D and Weston Turville are, respectively, the best trade-and-corridor fit, the only candidate reached by prediction from the naming custom, the exact age match, the best trade match and the best Weymouth-proximity fit - and every one of them scores zero on the four child attributes. The emigrant needs a man of the right age and his household, and every candidate supplies one half. That is the strongest argument in the file for the record-state rows: the man's own baptism and his household's baptisms are probably not in the same searchable place, one side being lost, unindexed, or in a class not yet opened.

Confidence: Moderate-high

Topic file location: Identity — the shape, the matrix, and the two families