Enrichment continues: 651 of 13,750 public race pages processed, 49 more races deep-researched — not finished

Asked (summary)

Comprehensively enrich the ~18,000 races in the RaceRove database at 50-race-finder.vercel.app, working through existing races to find missing or incomplete information.

Asked (follow-up)

"finish enriching"

This run processed 200 additional previously unprocessed RaceRove race pages and ran deeper external research on 49 priority races, yielding 327 source-level fact rows from 95 independent hosts. Cumulatively 651 of the 13,750 publicly exposed race pages are processed (4.73%) — 3.60% of the 18,075 active listings RaceRove reports; the public sitemap does not expose the remaining 4,325 listings. Enriching all 18,075 races cannot fit in one run, so this is a progress checkpoint, not a finish line.

Coverage so far, and what this run added

Inventoried means the RaceRove page was extracted; externally researched means outside sources were verified; fully enriched (every dimension sourced or marked not found) applies to none of the 651 yet.

All 49 deeply researched races got at least one external source; 48 of 49 now have a registration cycle or pricing on record — e.g. the 1/2 Sauer 1/2 Kraut marathon's full tier pricing and deferral policy from uberendurancesports.com.

Old editions are a live risk: 35 of 49 races surfaced at least one historical date in sources against 31 with a confirmed future date — a 2023 edition of the 100 on 100 Relay still ranks in search via mountaintimes.info.

Directories dominate the evidence (163 of 327 rows); official organizers and registration platforms, which we prefer for currency, supplied 125 rows — such as runsignup.com for Lakad for Rights.

Discrepancies persist: 28 of the 200 newly inventoried pages carry a stated date or distance mismatch, and external research found conflicts for 12 of the 49 deep races — e.g. an EverOut listing dating Lakad for Rights to the wrong weekday on everout.com.

The confirmed-duplicate rule stands: the Hampton duplicate action from the previous report is kept, and no new duplicates are inferred without locality plus organizer, registration, venue or date evidence from pages like racerove.vercel.app.

What external research filled in for the 49 races

Left: rows of evidence by source type (327 rows). Right: how many of the 49 races now have each dimension covered by at least one source.

Deep-research explorer — 49 races, filterable

Filter by evidence dimension and source type; click a race to open its sources. Green "current date" means a source confirms a future edition; "historical dates in sources" flags rows that must not be presented as current.

Source type

Date currency

Registration

Logistics

Course / elevation

Discrepancy review

The 200 newly inventoried RaceRove pages

One row per race page processed this run. Dots mark what the RaceRove page itself already holds; "deep" marks the 49 that also received external research above. Blank fields remain unresolved.

Race	Location	Date	Distances	Reg link	Official site	Discrepancy	Deep	Source

Method: progressive enrichment of the RaceRove database (racerove.vercel.app). This run extracted 200 previously unprocessed canonical race pages (captured 2026-09-25) and verified 327 external fact rows across 95 hosts for 49 exact-matched races (verified 2026-09-25). Cumulative coverage: 651 of 13,750 sitemap-exposed pages (4.73%); RaceRove reports 18,075 active listings, so 4,325 are not publicly exposed. A dimension counts as covered when at least one source supplies it; none of the 651 races meets the full-enrichment bar (canonical metadata, registration, five-year weather, logistics, sourced food, RaceRaves synthesis, course/elevation, discrepancy review, field-level provenance all sourced or marked not found). Long text fields and per-source detail beyond host links were cut for space.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.