SearchSeen

A name with possibilities.

Compare the supplied lists before explaining the difference

Suppose two fictional captures use the query blue ceramic mug. Capture A lists item IDs P, Q, R, S and T in that order. Capture B lists Q, P, U, S and V. Each supplied window contains five distinct items. The example provides no real search data and no claim that the captures used matching devices, accounts or times. It illustrates a comparison method, not a performance test.

SearchSeen.com
Two bounded windows in the SearchSeen result comparisonCapture A: P · Q · R · S · TCapture B: Q · P · U · S · VShared: P, Q, S · union: seven IDs3 / 5 = 60% · 3 / 7 ≈ 43%Different denominators, different questions

01

Separate shared items from window-only items

The shared IDs are P, Q and S, giving three shared items out of five in each window. The overlap share for either supplied list is therefore three fifths, or sixty percent. R and T appear only in A; U and V appear only in B. Across the two lists there are seven distinct IDs. If using intersection divided by union instead, the value is three sevenths, about forty-three percent. Name the denominator whenever reporting a percentage because these two calculations answer different questions about the same pair of lists.

Shared-item positions in the SearchSeen result comparisonP: position 1 → 2Q: position 2 → 1S: position 4 → 4Other positions: not supplied, not zeroOutside a window does not mean absent

02

Describe position changes only for shared IDs

P moves from position one in A to position two in B, while Q moves from two to one. S remains at position four. R and T have no supplied position in B; their absence from this five-item window does not establish where they appear elsewhere. Similarly, U and V have no supplied position in A. Keep those positions blank with an explanatory status rather than assigning zero. A zero would introduce a position that the example never observed and could distort a later calculation of movement.

Comparison record in the SearchSeen result comparisonQuery · source · capture referencesTime · window size · ordered IDsKnown conditions · missing conditionsVisible changes ≠ established causeKeep the original observation boundary

03

Save the conditions needed for a useful comparison

A reusable comparison record includes the exact query, source or catalogue, capture references, timestamps, window size, item identifiers and ordered lists. Add viewing conditions when known and label missing conditions explicitly. Before attributing differences to an update, establish which conditions were held steady and what independent evidence supports that explanation. The worksheet can report the visible changes immediately while leaving their cause unresolved. If a later capture expands the window, retain the original five-item observation and create a separate comparison with the new boundary.

searchseen.com

Explore SearchSeen

Held for potential development. Acquisition inquiries may be considered.

Acquisition ↗