Why this book, and not that one.
Adult science fiction and speculative fiction published in the US, judged on whether a real speculative mechanism drives the book rather than on which shelf it is sold from. What we look for is a premise that can carry a segment, backed by credible evidence that the book is worth your time. A book we cannot document is a book we cannot publish, however good it is.
These are Sci-Fi Books Weekly's criteria, set by its editor — not a Cur8r-wide policy. Every other channel writes its own.
What has to be true before we score anything.
A brilliant book in the wrong category is still the wrong book. So before any of the scoring below happens, a candidate has to clear a small set of gates — and a book that fails one is out no matter how well it would have scored.
These are not quality judgements. They are questions about whether a book is the kind of thing this channel exists to cover, and they are answered from primary sources rather than from a description. A release date that cannot be confirmed by a publisher or an ISBN record is not a release date we will publish.
- 01Adult
Marketed for adults, not young adult, middle grade or children's.
- 02New work
A genuinely new work, or enough new material to earn a new-release slot; reissues, abridgments and repackaged omnibuses normally fail.
- 03Verified release
The US publication date falls inside the target month and is confirmed by a primary bibliographic source, not by a retailer listing.
- 04SF centrality
A speculative mechanism, system, discovery, technology, alien ecology or future society materially governs the premise and its consequences.
All of them, before the rubric is run at all
The last gate asks about mechanism, not shelf label.
This is the claim on this page you are most likely to disagree with, so here is the reasoning. An earlier version of this gate asked whether a book was “science fiction enough,” and in practice that got applied as a genre-label filter. It cost the channel books that belonged in it — titles shelved as horror, as crime, as literary fiction, whose speculative machinery was doing real work.
So the question is now about what the book runs on. If a mechanism, a system, a discovery, an ecology or a future society governs the premise and its consequences, the book is in scope, and the horror and thriller release calendars get searched alongside the SF ones. What the gate is actually there to catch is decorative science fiction: a conventional story wearing SF scenery. That is a judgement about what the mechanism does in the book, and it needs a review or an excerpt — a shelf label cannot answer it. Where the evidence is thin, the book is held rather than quietly dropped.
How a candidate is scored.
Everything that clears the gates is scored out of 100 across 6 dimensions. The weights are the honest part: they say what this channel actually cares about, and they permit tradeoffs on purpose. A well-known author can partly compensate for a premise sitting on a genre border. A striking concept can partly compensate for the risk of a small press. Neither can compensate for a gate.
- Speculative idea
- 30
- Execution evidence
- 20
- Author and audience signal
- 20
- Science and SF centrality
- 15
- Episode usefulness
- 10
- Verification quality
- 5
A top score means: Distinctive, central and consequential, and expressible as one question a segment can actually explore.
A top score means: Multiple credible reviews or excerpts support disciplined prose, plotting and characterization.
A top score means: Major prior success, named awards, demonstrated readership, or exceptional prepublication attention.
A top score means: The speculative mechanism materially governs the world and the characters' choices.
A top score means: Adds variety, comparison value, accessibility or a strong discussion hook to the edition.
A top score means: Edition, date, format and every claim confirmed by primary sources.
The score ranks. It does not decide. The score ranks candidates so the editor knows what to read first, and it is neither a publication gate nor the thing that decides what makes the edition.
Cur8r has a second scoring rubric, and it is a different instrument. The platform rubric ranks the candidate items inside a single production run, for every channel on Cur8r, and the platform set it; the one above ranks books competing for Sci-Fi Books Weekly's slots, and this channel's editor set it. Both are weighted out of 100, which is exactly why they are worth telling apart.
The platform rubricWhat the score actually buys.
A score does not buy a place in the edition. It buys attention, in a specific order — what gets read first, what gets held for more evidence, and what gets set aside for the month with the reason written down.
Modifiers land after the base score, and they are mostly there to make a research shortcut expensive. The largest of them is the quietest: a gap that was searched for and recorded as missing costs almost nothing, while a field left silently blank costs real points. A blank is indistinguishable from never having looked.
- Select
- Goes to the editor as a proposed selection, unless a scope gate fails.
- Shortlist
- Compared against the rest of the edition for theme duplication and balance.
- Bubble
- Held back for a strong review, excerpt or audience signal that would move it.
- Reject
- Set aside for the month, with the reason recorded rather than left implicit.
- automatic rejectYoung adult, middle grade or children's. Outside this channel's audience, whatever the book's quality.
- −25Reissue or mostly-old omnibus. Unless the edition has an announced canon or reissue segment.
- −15Romance-first or comedy-first. The speculative element supplies decor rather than driving the story.
- −15No verifiable primary bibliographic source. And nothing publishes until it is resolved.
- −10Poor entry point to its series. Unless the series carries major audience value on its own.
- −10A required field left silently blank. A gap that was searched for and recorded costs almost nothing; a blank is indistinguishable from never having looked.
- +10Exceptional known-author or award signal. Never enough to override a scope gate.
A person makes the call, and says why.
Everything above describes a scoring system, and it would be easy to finish this page believing a number picks the books. It does not. The score sorts a month's releases so an editor knows what to read first. The editor reads them, decides, and writes down the reason in their own words — and that reason is kept as written, not smoothed into a summary later.
- 01Discover
Science fiction, horror, thriller, crime and literary release calendars, not just the SF ones.
Research - 02Gate
Four scope gates, all of them. A failure is recorded with the gate it failed.
Research - 03Document
Publisher page, release copy, reviews, the people involved, and an image with a rights basis.
Research - 04Rank
Weighted 0 to 100, then modifiers. It answers what to read first.
Ranking - 05Decide
The editor's call, with the reason written in their own words and kept as written.
The editor
Steps 01 to 03 gather evidence and record every gap they could not close. Step 04 ranks — it answers what to read first, and it never decides. Step 05 decides, and the reason is kept in the editor's own words.
The rules that hold whoever is deciding
- Judge the book's published positioning and execution, never the author's identity or assumed beliefs.
- Politics in a book is not disqualifying. The test is whether the politics emerges from a substantial imagined system, and it is applied the same way across every viewpoint.
- Never invent a rating, a sales figure, an award, a review consensus or a controversy.
- Preserve the counterevidence. A book we passed on may be excellent and still be a mismatch for this channel.
- Never present a publisher's own copy as independent evidence of how good a book is.
- Mark an inferred reason as inferred, and let the editor decide the unclear cases.
- A missing fact is recorded as missing, never guessed.
Where these criteria are not settled yet.
These are real tensions in the criteria above, not rhetorical ones. They are published because a reader deserves to know which parts of this are firm and which are still a judgement call being made edition by edition.
- How hard does the science have to be?
- Concept and credibility have consistently outweighed hard-science plausibility, which makes rigorous science a bonus rather than a gate. Where exactly that line sits is not settled.
- How much horror is acceptable?
- Gate 04 puts a horror-first title with a concrete speculative cause in scope. How much author signal is needed when that cause is thin is still being calibrated.
- Are discoveries penalised for thin evidence?
- The evidence requirement falls hardest on small-press and translated titles, which have the fewest reviews and the thinnest publisher web presence. That runs against this channel's stated interest in discoveries, and the fix is undecided.
- Where does genre-first fiction sit?
- A thriller or a literary novel with a real speculative mechanism is in scope, but positioning still influences the call in practice, and the channel does not yet have a clean rule separating the two.
Now go and see whether it worked.
Every briefing shows the books these criteria selected, and the reasoning behind each one.