Observed 2026-09-01 ยท aggregate corpus audit
Queer Romance Reader-Fit Evidence Index 2026
Prepared by: Queer Romance Books editorial data process, 2026-09-01. Reviewed by: independent automated adversarial release review, 2026-09-01, covering calculation, limitations, claims and the reproducibility boundary. This is not external subject-matter-expert endorsement.
101 of 223 profiles (45.3%) supported a same-lane next-book comparison after requiring every selected match to share at least two meaningful non-structural fit cues. The remaining 122 profiles now show an explicit insufficient-evidence state instead of a weaker recommendation.
The result by romance lane
| Lane | Profiles | Match ready | Share |
|---|---|---|---|
| Sapphic | 103 | 40 | 38.8% |
| M/M | 120 | 61 | 50.8% |
| All profiles | 223 | 101 | 45.3% |
The lane difference describes this frozen site corpus only. It is not evidence that one readership, genre, or set of books is easier to recommend in general.
A second completeness check: heat evidence
Heat was confirmed for 198 of 223 profiles (88.8%). The other 25 remain explicitly unknown; unknown never means low heat.
Method
- Freeze all published-profile records on 2026-09-01 and record the source JSON SHA-256.
- Normalize trait labels and remove structural labels such as Sapphic romance, M/M romance, F/F, M/M and Couple.
- Keep only same-lane candidates sharing at least two remaining traits.
- Mark a profile match ready only when at least three qualified candidates exist; otherwise publish no comparison list.
- Recalculate the aggregate rows from the frozen JSON and independently run the release validator.
Frozen source SHA-256: 489ed40cf714293137d86522ff337d01e11cb86bc6d807d51360b5cef3ed78af
Version: 2026.1. Recommended citation: Queer Romance Books (2026). Queer Romance Reader-Fit Evidence Index 2026, version 2026.1. Observed 2026-09-01.
Downloads
Published aggregates are licensed CC BY 4.0. The underlying books, third-party facts and source materials retain their respective rights.
Limitations
- This is one site’s selected publication corpus, not a representative sample of all queer romance or all readers.
- Evidence availability is uneven. Missing or unknown information is not treated as absence or poor fit.
- The analysis measures evidence readiness, not book quality, popularity, sales, reader identity or universal preference.
- The public files contain aggregates, not the rights-restricted row-level evidence snapshot. They document the calculation and permit checksum verification, but an outside party cannot independently reproduce the 101-profile classification from the downloads alone.
- No external subject-matter expert endorsement is claimed.
Corrections and next update
Send a correction with the affected field and supporting source to [email protected]. The next planned review is September 2027, or earlier after a material evidence-graph expansion.
Return to the research hub or read the recommendation methodology.