NovelForge DocsReference

Reader Engagement · The positive quality model after prose clears the floor

TIER C · CONTRACT  ATTENTION  REWARD  FORWARD PULL

TIER C · CONTRACT ATTENTION REWARD FORWARD PULL

Surface Fundamentals ask whether prose is malformed, mechanically AI-like, causally thin, or falsely polished. Reader Engagement asks the positive question: what makes a reader want to continue, feel rewarded, care about the people, update their mental model, and experience meaningful movement?

Boundary ✦ Reader Engagement is a generic quality model. It is not a single score, not a mandate for constant maximum tension, and not a substitute for genre/platform/project taste. Profiles change weighting and thresholds; they do not erase the dimensions.

01 · Quality floor and positive quality are different

A chapter can pass every Surface hard rule and still fail because it is:

safe but procedural;
technically coherent but predictable;
populated by functional characters;
emotionally inert;
full of correct actions with little cost or choice;
structurally complete but weakly propulsive.

The relationship is:

Surface Safety       = minimum realization floor
Reader Engagement    = positive reading-value model
Genre / Platform     = weighting + threshold
Project / User       = calibrated emphasis

Passing Reader Engagement still does not grant Canon authority or replace continuity/independent gates.

02 · RG-01 · ACTIVE READER QUESTION

Maintain at least one live near-term question appropriate to the form.

Examples:

Can a concrete objective be achieved?
What will another person choose?
Which hidden fact changes the current interpretation?
What cost will be paid?
Will permission, status, trust, or access change?
What will the focal character do as options narrow?

The question should evolve, sharpen, branch, resolve, or be replaced by a more consequential one. Repeating the same countdown or vague threat is not evolution.

Quiet/literary profiles may use subtler reader questions. High-grip serial profiles usually require stronger near-term pull.

03 · RG-02 · PRESSURE LADDER

Pressure matters when it changes the available option landscape.

A meaningful beat may:

remove or weaken a viable option;
reveal or increase cost;
introduce credible resistance;
expose an omission or mistake;
force an earlier decision;
resolve one pressure while opening a more consequential one;
change the meaning of earlier information;
alter relationship permission or trust;
make inaction costly.

Do not add arbitrary obstacles merely to “raise stakes.” Useful pressure grows from character, institution, resource, environment, prior choice, information limit, or already planted state.

04 · RG-03 · REWARD CADENCE

Readers need returns on attention, not only delayed correctness.

Possible rewards include:

competence payoff;
useful information or reveal;
emotional intimacy;
social reversal;
humor;
recognition/callback;
status or permission gain;
a satisfying objection answered;
a planted detail becoming functional;
surprising but credible character action;
sensory or rhetorical pleasure when the profile values it.

There is no universal reward quota. Ask whether the reader receives something worth the time before attention debt becomes too large for the chosen form.

05 · RG-04 · FRICTION OVER PROCEDURE

When a scene contains routine process, expand the parts where process becomes story:

people disagree;
information is incomplete;
a mistake is discovered;
responsibility or risk changes hands;
time/resources become binding;
an imperfect choice must be made;
relationship state changes what can be asked or done;
expertise reveals an unexpected consequence.

Compress routine operations unless the procedure itself is part of the promised pleasure.

Procedural fiction is not exempt: the procedure must still supply competence, uncertainty, discovery, conflict, consequence, or domain-specific satisfaction.

06 · RG-05 · CHARACTER-OWNED ENERGY

Important characters should contribute scene energy through their own:

agenda;
competence;
bias;
humor;
fear;
appetite;
social strategy;
information;
initiative;
unexpected but credible choice.

A scene loses energy when everyone except the protagonist becomes a delivery mechanism.

“Character-owned” also means the energy must belong to the person’s knowledge, relationship, social position, and voice—not to narrator hype.

07 · RG-06 · CHARACTER SURPRISE WITHIN CONSISTENCY

Memorable characterization often comes from a choice that is surprising enough to feel alive but consistent enough to feel earned.

Examples of mechanism, not templates:

a cautious person cuts through ritual when risk becomes concrete;
a joking person becomes precise when consequences turn real;
a junior catches a senior’s mistake;
an apparent obstacle offers the cleanest solution for self-interested reasons.

Do not manufacture quirks. Surprise should emerge from goals, competence, history, relationship, information, pressure, or a previously underused part of the character.

08 · RG-07 · TONAL CONTRAST / TENSION WAVE

Continuous maximum tension becomes flat. Lower-pressure beats can make later pressure land harder when they provide:

relationship texture;
humor;
aftermath;
information;
false relief;
setup;
wonder;
quiet moral or emotional recognition.

Contrast must still do work. Do not insert filler merely to satisfy a waveform.

09 · RG-08 · CHARACTER-OWNED HUMOR / HUMANITY

When humor is appropriate, prefer humor that emerges from:

relationship familiarity;
incompatible agendas;
status mismatch;
professional or period language in an awkward human context;
misunderstanding with real consequences;
self-interest;
dark humor under pressure;
an earlier confident statement acquiring a new meaning.

Humor is strongest when it also reveals voice/relationship, changes tension or willingness, buys/loses time, plants information, or sharpens a later reversal.

A profile may set humor to low or none. The generic invariant is that human texture should belong to people and situations rather than detachable narrator stand-up.

10 · RG-09 · OPENING ORIENTATION + GRIP

Early prose should establish enough orientation to read confidently while giving a reason to care.

Possible grip mechanisms:

an active problem;
contradiction;
desire + cost;
social friction;
deadline;
an unusual concrete fact;
distinctive voice or humor;
emotionally charged relationship state;
mystery grounded in real information limits.

Do not confuse obscurity with intrigue. A reader who does not know what to pay attention to is not necessarily curious.

11 · RG-10 · FORWARD-PULL ENDING

A scene/chapter ending should usually leave changed state + a reason for mental continuation appropriate to the form.

Valid mechanisms include:

irreversible consequence;
a concrete unresolved near-term question;
return of a planted issue;
recontextualization;
an incoming person/message/object;
a next-state sensory or voice intrusion;
an earned callback;
a consequential decision whose cost is deferred;
relationship permission or obligation that changes what comes next.

Reject abstract narrator advertising. Do not require a twist every time.

12 · RG-11 · INFORMATION VALUE

Information earns attention when its arrival changes:

interpretation;
choice;
anticipation;
emotional meaning;
relationship position;
perceived risk;
the reader’s model of what is possible.

Avoid exposition that merely fills a database field, ordinary facts withheld for cheap mystery, or reveals scheduled independently of current action.

Good information has consequences in the reader’s model or a character’s next move.

13 · RG-12 · CAUSAL MOMENTUM

A strong scene feels like meaningful beats create the next meaningful beats.

cause
→ response
→ changed state
→ new decision / problem

Chronology alone is not momentum. “Then this happened, then this happened” can be perfectly clear and still feel inert.

14 · RG-13 · RELATIONAL MOVEMENT

Important interactions should often leave a trace:

trust changes;
permission changes;
obligation appears;
attraction/aversion shifts;
status is renegotiated;
a private fact becomes shared;
one person revises their model of another;
future expectation changes.

Not every conversation must transform a relationship. Long-form fiction becomes inert when major relationships repeatedly reset to zero despite meaningful events.

Relationship movement remains subject to Canon/state evidence rules: a Review Draft can depict a delta, but current authoritative relationship state changes only after acceptance/settlement.

15 · RG-14 · SPECIFICITY OF STAKES

Pressure works better when cost is concrete enough to change a decision.

Prefer operational consequences such as:

money that can or cannot be paid;
a deadline with procedural or social consequence;
permission/access at risk;
an injury that removes options;
evidence that can disappear;
a public promise that must now be honored;
a relationship cost that alters future cooperation.

Generic “everything is at stake” language is weaker than a specific constraint the character must act around.

16 · RG-15 · SAFE-BUT-FLAT

Fail the positive quality gate when prose is technically clean but many of the following are true:

the chapter reports correct actions chronologically;
obstacles are solved on the first attempt without tradeoff;
dialogue mostly transfers necessary information;
voices are distinguishable but not characterful;
pressure is announced but rarely changes choices;
procedure occupies more space than conflict/choice/consequence;
supporting characters remain functionally static;
the climax arrives because the outline says it is time;
state changes, but reader desire does not;
there is little useful tonal contrast, humanity, or reward.

Repair upstream at Reader Pressure, Scene Simulation, Character Simulation, or Story/Plan. Do not patch flatness with random one-liners, extra fragments, adjectives, fake twists, or narrator hype.

17 · Profile weighting

Profiles tune emphasis rather than redefine the model.

reader_grip: low|medium|high|very_high
serial_forward_pull: low|medium|high
reward_density: low|medium|high
humor: none|low|medium|high
mystery: low|medium|high
relationship_motion: low|medium|high
procedural_pleasure: low|medium|high
rhetorical_pleasure: low|medium|high
contemplative_space: low|medium|high

Examples:

high-grip serialized commercial fiction heavily weights RG-01/02/03/04/10/12;
cozy fiction may lower threat pressure while increasing relationship, reward, voice, and comfort dimensions;
literary fiction may accept slower near-term propulsion while demanding high interpretive/emotional specificity;
procedural fiction may expand process when process itself supplies competence, uncertainty, conflict, discovery, or domain pleasure.

Profiles alter thresholds. They do not delete quality dimensions simply because the work is quiet or specialized.

18 · Reader diagnostics versus Reader Engagement contract

NovelForge may use a bounded Reader Simulation Panel to gather diagnostic evidence. That panel can report signals such as continue desire, pacing, confusion, emotional response, favorite/stumble beats, forward pull, character investment, reward, and drop-off points.

The panel is one evidence source for this quality model. It is not the model itself.

Important distinctions:

simulated-reader personas describe reading behavior, not demographic stereotypes;
panel results are diagnostic evidence, not Canon;
an absolute score alone does not decide keep/discard;
A/B order swapping can expose position bias;
disagreement between reader personas is useful evidence rather than noise to average away;
Reader Panel does not automatically satisfy a mandatory independent semantic gate.

See Quality Evolution.

19 · Engagement audit artifact

A user-visible or machine-readable audit should record conclusions and observable evidence rather than private reasoning.

artifact_type: reader_engagement_audit
candidate_fingerprint:
active_reader_question:
question_evolution: []
pressure_ladder: []
reader_rewards: []
character_energy: []
tonal_contrast: []
information_value: []
causal_momentum:
relationship_movement: []
procedure_compression:
ending_forward_pull:
profile_weights:
flatness_risk:
result: pass|rewrite

The audit should be specific enough that repair can route to an owning mechanism.

20 · Failure routing

weak active question / pressure → Reader Pressure + Scene/Chapter planning;
no meaningful reward → payoff/information/relationship/competence design;
procedure dominates → compress routine and restore friction/choice/consequence;
character energy missing → Character Simulation;
relationship repeatedly resets → relationship state + scene design;
flat but surface-clean prose → upstream regeneration, not cosmetic polish;
reader confusion from missing context → context/story information design;
reader confusion from sentence realization → Surface layer;
reader diagnostics disagree strongly → inspect the tradeoff; do not blindly average them into one number.

21 · Invariants

Reader Engagement is positive quality, not a defect blacklist.
Tension is not the only engagement mechanism.
Reward is not the same thing as victory.
Forward pull should come from changed story state, not narrator advertising.
Important characters should contribute their own energy.
Profile weighting changes thresholds, not the existence of dimensions.
Reader diagnostics are evidence, not Canon or automatic independent review.
SAFE-BUT-FLAT routes upstream.

Surface Fundamentals — the prose-quality floor.
Story System — active questions, pressure, payoff, causal planning.
Character & Relationship System — character-owned energy, surprise, and relational movement.
Quality Evolution — Reader Panel evidence, typed findings, revision routing, and candidate evolution.
Production Pipeline — where Reader Pressure and Reader Engagement sit in the run.

A strong chapter does more than avoid bad prose. It creates attention, feeling, inference, reward, and desire for what comes next in the way its chosen form promises.