By Eikna Editorial Team · Editorial-method draft · 2 October 2026
What kind of review are you reading?
| Label | What it supports | Its limits |
|---|---|---|
| Researched editorial assessment | A synthesis of inspected descriptions, criticism and attributed reader responses. | No claim that our team read the novel. Craft and emotional-impact judgments must remain attributed or explicitly interpretive. |
| Firsthand review | A documented reading, with the edition and reviewer’s notes recorded. | One reading is still a perspective, not proof of every reader’s experience. No current review draft carries this label. |
| Reading guide or selection hub | Comparison of supported premises and routes to explore. | A checked synopsis does not verify every detail of a book record. |
When a numerical score is withheld
Our current researched reviews have not yet met every evidence requirement for numerical editorial ratings. Their conclusions use evidence and reader-fit notes rather than a formula built around another platform’s average. An internal research worksheet is not a validated quality model and does not become a public ranking.
Where a scorecard uses Strength, the available evidence supports a particular appeal. Mixed marks a meaningful counterweight or differing responses. Weakness identifies a supported concern, with its source and scope explained. Unassessed means there is not enough inspected material to judge. None of these verdicts means all readers agree.
We consider the relationship, characters, writing, emotional effect and overall fit when evidence permits. Heat and ending receive descriptive treatment. Reread value needs reading evidence; it is not inferred from sales, popularity or a happy ending. Findings should include their limitations rather than meet an arbitrary bullet count.
Which source answers which question?
There is no single source that settles every field. The edition’s publisher can establish its ISBN and format. A documented reading can establish a scene. An author’s warning list can report selected content. A critic can support an attributed judgment about execution. A retailer’s category or an individual reader’s tag is a lead to investigate, not automatic confirmation.
We retain the source URL, the field it supports, the retrieval date and a short evidence summary. Two pages repeating the same publisher blurb are one underlying account. A prior Eikna article is a route to its sources, not an independent witness.
When sources disagree
Keep the disagreement visible in the record. Page counts and audiobook durations may differ by edition or listing; match identifiers before deciding that a source is wrong. If the discrepancy remains unresolved, report the conflicting values or omit a single definitive claim. Do not average incompatible metadata.
Unknown does not mean absent
| State | Meaning |
|---|---|
| Supported | Inspected evidence supports this specific claim, with its source named. |
| Reported | A named source states it; we have not independently established the underlying detail. |
| Editorial interpretation | Our reading of inspected evidence, distinct from a factual field. |
| Unknown | Relevant sources were checked but did not settle the question. |
| Unassessed | The required check has not been completed. |
| Conflicting | Inspected accounts disagree and resolution remains open. |
A missing warning list does not establish that a book contains no difficult material. A synopsis that omits sex does not establish closed-door romance. Silence about a sequel does not establish standalone closure.
Heat, burn and pace answer different questions
Heat describes sexual content: whether it appears on page, its detail and its frequency. Burn describes the development of attraction, trust or romantic commitment; the milestone should be specified. Pace describes the movement of the story. A book can be explicit and slow burn, or closed door and fast paced.
Our current drafts do not infer numerical heat from genre, cover or synopsis. If a specialist source supplies a band, identify that source and its definition rather than silently convert it into an EIKNA score. Where only particular sexual content is reported, say that and leave frequency or overall heat unassessed. Heat is never a reward or penalty in quality assessment.
Three separate ending checks
We separate the relationship outcome (HEA, HFN or another supported outcome), unresolved plot or cliffhanger, and series dependency. A happy couple can coexist with an unfinished external conflict. A continuing series does not by itself prove a cliffhanger. See our HEA and HFN guide for spoiler-light distinctions.
Exact ending details belong behind a clear spoiler notice and need scene-level evidence. We do not fill missing spoiler sections from genre expectations.
What reader responses can tell us
An individual response illustrates that reader’s experience. Several independently inspected responses can show a recurring theme within the material examined; they do not establish a majority. Different reviewers quoting the same account are not independent corroboration.
Any published count must specify the platform, collection date, selection rule, sample size and coding rule, with traceable sources. Our supplied 30-review worksheets concern the first 30 displayed reviews, not a random sample or a verified most-liked sample. Internal arithmetic checks do not verify the underlying review texts. Until those texts and links are independently inspected, their counts cannot substantiate public consensus claims.
Platform averages, if later included, need a dated source and count. They are platform statistics, not our score. Combining an average, star distribution and happy-reader percentage can count the same underlying ratings repeatedly.
How we classify books for discovery
Subgenre, pairing, trope, setting and reader audience remain separate. A rivalry can have serious stakes; institutional hostility does not automatically establish personal enmity. Our enemy-intensity framework is a provisional local editorial tool, not an approved EIKNA score or a measure of darkness, consent or quality.
Selections should explain the supported obstacle and who may enjoy it. The contemporary hub currently has 18 premise checks, not 18 complete book records. The enemies hub is clearly an 11-book starter. We do not relabel either as a fully verified catalogue.
How editorial research reaches EIKNA
The article and the structured book record are separate layers. Interpretive prose does not automatically become database truth. Each proposed factual field needs its evidence and origin: EIKNA_ORIGINATED, EIKNA_ROMANCE_RESEARCHED, EXTERNAL_SOURCE or EDITORIAL_JUDGMENT. Information received from EIKNA and returned through this site cannot strengthen confidence as independent confirmation.
Canonical identity, taxonomy IDs and physical field names require the live schema and identity checks. Unknown values stay unknown. Our draft supplements are not approved exports; no current draft has been submitted to or written into EIKNA.
Commercial disclosures
The editorial policy owns commercial independence and disclosure requirements. Current draft purchase buttons remain inactive.
Corrections
See the correction policy for change records and the pending public correction channel.
Explore the method in practice
Compare the researched assessments of Book Lovers, Delilah Green Doesn’t Care and The Kiss Quotient, or use the contemporary reading routes. For the remaining checks, see the publication checklist.