Anchor → resolver score-delta harness — openaddresses-de-sample.jsonl
Offline early-signal for the DEFERRED postcode-anchor resolver re-ranker (task #59, #240). For each row we query the locality lookup with no country (the honest multi-locale baseline), then soft re-rank the candidates by the postcode anchor's country posterior, and log what changes. The shipped resolver is untouched.
- anchor weight: 2 · candidates/query: 10 · rows: 3000
- eligible (locality + candidates + anchor): 1197 (skipped: 941 no-locality, 862 no-candidate, 0 no-anchor)
| metric | value |
|---|---|
| anchor changed the top-1 pick | 3.3% (40/1197) |
| of those, wrong-country → anchor-country corrected | 33 |
| gold locality match — anchor-OFF | 82.5% (988/1197) |
| gold locality match — anchor-ON | 82.5% (988/1197) |
| net gold-match delta (name) | +0.0% (0 improved, 0 regressed) |
| mean score margin the new winner overcame | 4.493 |
| median coord error — anchor-OFF | 10.3 km |
| median coord error — anchor-ON | 10.3 km |
| coord error improved >100 km / worsened >100 km | 33 / 7 (of 1197 placed) |
Read
The name-surface gold-match metric is blind to country confusion — a US "Berlin" name-matches the German gold "Berlin" while sitting an ocean away. Coordinate error to the OA gold point is the non-gameable signal, so weigh the coord deltas over the name deltas here.
Feeding the anchor's country posterior corrects 33 wrong-country picks and pulls 33 rows >100 km closer to the gold point (median 10.3 km → 10.3 km, 117492 km saved total). That value is invisible to name-match (+0.0%) — exactly the artifact the coordinate-first resolver direction flagged. The re-ranker is worth prototyping; the mean margin (4.493) is the score gap a soft boost must clear.