Glossary

HookThe opening move of a piece of content whose only job is to make the reader feel a gap and reach to close it. Avoid: Attention-grabber, opener, first line (too vague)

The GapThe felt distance between what the reader already knows and what they suddenly want to know. Curiosity is the sensation of this gap. Every hook manufactures one. Avoid: Curiosity hook, open question, teaser

Sweet spot (of a gap)A gap small enough that the answer feels one swipe away (closeable) and big enough that the answer would change something the reader cares about (matters). Miss either side and the pull dies.

Failure 1: No gapA hook that hands over the answer, or asks about something the reader already knows or doesn't care about. Nothing is missing, so nothing pulls.

Failure 2: Gap too wideA hook whose crack is so big or vague the reader can't imagine a real answer, so it reads as clickbait or noise. The gap must feel scratchable.

Believable bridgeThe sense that a real, satisfying answer exists behind a surprising gap. A pattern-break with no bridge reads as random and produces confusion, not curiosity, a specific way Failure 2 happens even when the hook is concrete. Sweet spot = surprise AND plausibility together.

Hook-payoff matchThe hook must open the same gap the content actually closes. If the hook promises a bigger or different gap than the payoff delivers, even readers who bite feel let down.

ArousalHow much a feeling fires up the body. High-arousal emotions (awe, anger, anxiety, excitement, recognition-shock) stop the scroll and get shared; low-arousal ones (sadness, contentment, mild interest) get scrolled past. Arousal is the fuel poured into the gap. Avoid: Emotion, intensity, drama (too loose)

ValenceWhether a feeling is good or bad. Distinct from arousal, and NOT the lever that drives sharing. Sad (negative) and content (positive) are both low arousal, so both flop. Fire alarm vs lullaby: both are sounds (valence differs), only one activates you (arousal differs).

Open loopA question or tension deliberately left unresolved so the brain stays snagged on it and wants to close it. A mental browser tab you can't shut. The engine of retention (staying), where the gap is the engine of attention (stopping). Avoid: Cliffhanger (that's only one kind), teaser

Payoff delayThe space between opening a loop and closing it. The tension, and the attention, lives in the wait, not in the question or the answer.

Loop stacking (nesting)Opening a new loop before closing the current one, so there's always an outstanding "debt" pulling the reader forward. How threads and videos hold attention for minutes.

The iron ruleEvery open loop is a promise. Close too early → no pull. Never close → broken trust. Open loops are debts; always pay them.

The MaybeUncertainty about the outcome itself, not just a hidden answer, but a real doubt about which way it goes or whether it pays off. Sustains attention because the brain stays activated while the result is genuinely undecided. Avoid: Suspense, cliffhanger (those can be faked; the maybe must be real)

Reward prediction errorDopamine tracks the gap between expected and actual reward, not the reward itself. A fully predicted reward makes almost no dopamine; a surprise makes a spike; an uncertain (50/50) reward keeps dopamine ramped through the whole wait. Source: Schultz.

The certainty penaltyA bigger, surer promise pulls less, for two reasons: the brain disbelieves an overclaim (trust drops), and a certain payoff makes less dopamine than an uncertain one. Honest "I'm not sure this works, but..." out-pulls a bold promise.

Variable reward (ethics)Uncertain reward is also the engine of slot machines and doomscrolling. The honest line: promise only the uncertainty you actually have. A rigged tease wins once and loses the follower for good.

Title hook (on-screen text)The words pinned on screen in a video's first frame. Defined by delivery, not content: its one hard job is to work with the sound off (the muted majority). On demo/B-roll formats it often frames a visual payoff; on talking-head content it just carries the hook as text while the payoff is spoken. Glanceable, 3-8 words. The billboard.

Verbal hook (spoken)The first thing said aloud. Job: deepen the pull once they've stopped (or, with sound on, it IS the hook). Often carries an argument or story that lives in words (claim, paradox, confession). One breath, ~5-12 words. The voice in your ear.

The split ruleAim to divide the four engines across the two channels: title carries the gap (and stands alone muted); voice adds arousal, the maybe, or escalation. Some overlap of the core claim is fine, muted vs sound-on are different audiences, so restating on screen is coverage, not waste. The waste is a verbal that adds nothing over the title.

Show vs tell (a tendency, not a law)If the payoff is something you SHOW, a title-framed visual often leads. If it's something you ARGUE or NARRATE, the verbal often leads. But talking-head title hooks carry claims with no visual at all, so this is a lean, not a rule.

Self-relevance (the targeting lever)"Make it about them." Not a pull by itself, it aims the four engines at a specific person so they feel caught. A gap about "people" is weak; a gap about "you, the one who [does X]" is magnetic. Narrowing the audience increases the pull (the narrow-to-widen paradox). Avoid: Niche, targeting (too abstract)

Cocktail party effect(Cherry, 1953): In a noisy room we filter almost everything out, but our own name cuts through instantly. Self-relevant info gets automatic attention priority. Why an "if you're [X]" call-out stops the right scroller.

Self-reference effect(Rogers, Kuiper & Kirker, 1977): Information tied to the self is remembered better. Self-relevance both grabs attention and makes the hook stick.

Flavor (vs engine)An engine is a force in the reader's head (gap, arousal, open loop, the maybe). A flavor is the shape you pour that force into. Heat is one force; boiling, frying and grilling are shapes. Run all four engines at once, but usually one flavor per engine. Two gap flavors in one line reads as noise. Engines diagnose; flavors produce.

Gap flavorscuriosity gap (missing what) · utility gap (missing how) · surprising-origin gap (missing the link) · emotional-contradiction gap (missing how two feelings both fit) · process gap (missing the road, show the destination, hide the route; the engine under "transformation" hooks). Two laws govern all five: a gap needs prior knowledge to stand on, and small gaps pull harder than large ones (Loewenstein 1994). Kang et al. (2009) measured the peak: curiosity is highest when confidence of knowing sits around 0.50. The half-knowing point. Give them half the answer.

Arousal flavors (the ladder)awe · anger/indignation · anxiety/fear · excitement/hope · recognition ("this is so me") · validation ("finally someone said it", the relief of hearing a quietly-held belief said out loud; needs the belief to have been unspoken). Awe, anger and anxiety are the three with direct evidence (Berger & Milkman 2012). Validation rests on Hart et al. (2009): a moderate preference for agreeable information, d = 0.36, and it flips when the disagreeable info is useful. Recognition is our own build (self-reference effect + a privacy-violation jolt), not a named emotion in the research.

Open loop flavorswithheld noun · numbered promise · the turn ("but wait") · unit cliffhanger (built for carousels) · loop stacking.

Maybe flavorshonest uncertainty · the bet (a public stake you could really lose) · the skeptic's turn.

The uncertainty rampThe physical shape of a maybe. Fiorillo, Tobler & Schultz (2003) found two dopamine signals, not one: the quick prediction-error burst, and a sustained climb that builds through the whole wait and tracks uncertainty, at its maximum when the odds are P = 0.5, shrinking toward certainty either way. So a maybe is not a spike at the end; it's a rising line through the middle. Why "the bet" must be one you could genuinely lose.

Credibility (the bridge, made deliberate)Trust in the source, the move that makes a reader believe a real answer sits behind your gap. Not a pull; it's the believable bridge built on purpose. Flavors: the receipt (a number) · the tenure (time served) · the scar (an admitted mistake, trust from failure) · the witness (someone else's words). Distinct from The Maybe, which is honesty about the outcome. Credibility without a maybe slides into overclaim and eats the certainty penalty. Source: Hovland & Weiss (1951), same message, credible source, more opinion change. Their sleeper effect is the expiry date: weeks later the advantage faded as people forgot who said it, so credibility buys the first moment, exactly what a hook needs and no more.

The pratfall condition(Aronson, Willerman & Floyd, 1966): A quiz contestant who had answered 92% correctly became more likeable after spilling coffee on himself, but a mediocre one (30% correct) became less likeable for the identical blunder. So: a scar only builds trust if competence is already visible. A brand built on vulnerability still has to show the competence somewhere, or the admission just reads as someone who cannot do the thing. Governs both the scar and honest uncertainty.

Evidence tags (used in Lesson 07)research = a named study tests something close · inferred = the mechanism is established but the hook use is untested · craft = observed practice and reasoning only, treat as a hypothesis. The blanket caveat: none of these studies looked at a person scrolling a feed. The machinery travels; the sizes and rankings do not.

The 7 field strategies(videodatabase.org): A field catalogue of visible hook patterns: H.01 Curiosity & Open Loops → gap + open loop · H.02 Relatability & Contrast → self-relevance + recognition · H.03 Transformation & Payoff → process gap + hope · H.04 Credibility → the bridge · H.05 Secrets & Shortcuts → utility gap dressed as access · H.06 Opinions & Polarization → anger + validation · H.07 Identity & Specificity → self-relevance. Useful as a menu of shapes, not as a model: the seven sit at four different levels of the machine, their counts measure popularity, not performance, and the list is blind to The Maybe (which has no phrase tell). Rule: diagnose with the engines, unstick with the seven.

The build order(Lesson 08): The eight steps that turn the system into a hook you can produce on demand. 1. Write the payoff in one sentence · 2. Name one person · 3. Find the true feeling · 4. Pick the gap flavor your payoff can pay · 5. Set it at half-knowing · 6. Build the bridge if the gap is surprising · 7. Choose the debt and the doubt · 8. Split the channels.

The inversion(step 1, the important one): Write the payoff before the hook. Hook-payoff match is a constraint, and a constraint has to exist before you can satisfy it. Writing the hook first is how you end up promising more than you have, then inflating the content to catch up.

The complete artifact(the boundary of the system): Content where the payoff is the hook, the whole reward arrives in the first frame, so there is nothing to wait for. The four engines optimise attention over time; a complete artifact doesn't want time, it wants a reaction. There, validation + self-relevance is the whole machine and a gap actively hurts, because it delays an agreement people want to give immediately. Verified example: @wallylaflair's "Hot take: all relationships can survive mistakes, but they cannot survive patterns", 5.47M views, 17.0% engagement, with no gap, no loop and no maybe. Also skip the system for pure utility and for anything heavy (grief, bad news), where engineering a maybe is manipulation.

The like-to-comment ratio(a way to read a hook's real effect): Likes ÷ comments. A high ratio means agreement; a low one means argument. The @wallylaflair hot take runs ~620:1, roughly twice as like-dominated as comparable videos (~314:1 and ~333:1), so despite being labelled "opinions-polarization," it was almost pure validation. Rule: when a label and a ratio disagree, trust the ratio. Caveat: engagement rate is not watch time; it says what people did, not how long they stayed.