Dallas Cowboys fans don't wait until Monday morning to find out if their team won. They sit on the couch Sunday night watching the game with money down on the outcome, eyes locked on the television, because they have made a prediction and now the world has to confirm or refute it. Matt Dicks names this the universal human attention-mechanism: Human beings are prediction machines. That's why gambling is so difficult for people. It's not really the money as much as it is — I think the Cowboys are going to win this week and I'm going to put money down on it to prove that I'm right. Those people don't, like, wait till Monday and open the newspaper to see if they won, right? They're watching it at the moment the game's playing because we're prediction machines.1
The crystal-ball technique is the storyteller's deployment of this mechanism. When I say to you, "it's okay because I'm going to get that spoon in a little while" — right? I'm making a prediction about the future, and now you're going to want to hear as an audience whether that prediction is going to come true.2 The storyteller out-loud predicts something. The audience's prediction-machinery activates. They now want to stay through the piece just to find out whether the storyteller's prediction was right. The technique is small. The effect is durable.
A crystal ball has three properties:
1. It is stated out loud, not implied. Unlike breadcrumbs which work through understatement, crystal balls are explicit predictions. The storyteller says: I'm going to get that spoon, or this is going to end up being the worst date of my life, or I didn't know it then but I was about to make the biggest mistake of my career. The audience hears the prediction clearly.
2. It activates prediction-machine cognition. Once the prediction is in the audience's mind, they cannot help tracking whether it comes true. We're prediction machines. The mechanism is involuntary. The audience does not consciously decide to care about the prediction's accuracy; their cognitive architecture cares automatically.
3. It can resolve confirmed, refuted, or split. A skilled storyteller can use any of three outcomes. Confirmed predictions produce satisfaction (the storyteller was right). Refuted predictions produce surprise (the storyteller was wrong in an interesting way). Split predictions — partially right, partially wrong — produce both. All three are usable; none is structurally preferred over the others. The technique's effectiveness comes from the audience's tracking of the prediction, not from any specific resolution shape.
The crystal ball is technically a sub-variant of the stakes engine but operates through a different cognitive door than typical stakes. Normal stakes engage rooting-for-protagonist psychology; crystal balls engage prediction-validation psychology. Both produce attention; the door is different.
The reason crystal balls work so reliably is that prediction-validation is one of the deepest cognitive habits humans have. From early survival contexts (predicting predator behavior, weather, food availability), brains evolved to reward prediction-tracking with attention-investment. Once a prediction is in the cognitive space, the brain will not release it until the prediction is resolved. Either way I'm winning. Because if you made a prediction that it's a knife or a gun, now you want to know if your prediction is going to come true.3
This is also why the opening of a story — the moment when the audience is first deciding whether to stay — is such a good place to install a crystal ball. I'm standing in front of this enormous pile of fall leaves and they're quivering. There's a little boy inside the leaves and his hand emerges and he's got a metal object in his hand.4 The phrase metal object is implicit-crystal-ball: the audience immediately predicts (knife, gun, something dangerous), and now they're locked into staying because they want to find out whether their prediction was correct.
Dicks's earnings-call example shows the technique at a different scale. Corporate guidance is essentially a crystal ball. When they give guidance, those people are going to come back at the next earnings call to see if that guidance played out, or they're going to pay attention over the course of the quarter to see what the results are. Guidance is nothing but a crystal ball saying "here's what we think is going to happen" and then you wait three months to find out if it actually happened.5 The mechanism is identical: an out-loud prediction activates the prediction-machine, the audience tracks for resolution. This is why earnings-guidance is treated as load-bearing communication despite being technically optional information disclosure.
The crystal ball is the prediction-engine in the engagement-technique taxonomy. It complements backpacks but operates differently:
The two often deploy together. The storyteller can share a plan (backpack) and predict its outcome (crystal ball). When the prediction resolves, both techniques fire simultaneously.
Crystal balls also pair cleanly with hourglasses: after a crystal ball is installed, the audience wants prediction-resolution; the hourglass delays the resolution; the delay amplifies the resolution-pleasure. The orange-chair moment in the spoon story works on top of multiple earlier crystal balls (will Matt get the spoon? will Jamie keep it?).
The technique reinforces the no-one-wants-to-hear mindset by giving the audience a specific, concrete reason to stay through the next sentence — they want to find out if the prediction came true. Crystal balls are one of the most reliable opening-and-middle techniques for keeping audience attention from drifting.
Lewis (Newel of Knowledge) describes his own crystal-ball deployment in a video titled whatever scares you go do it.6 The 5-second moment of the piece is Lewis at age 18 turning down a public-speaking opportunity at Toastmasters because he was nervous. The crystal ball Lewis installs at the opening is himself talking up his confidence:
I just came back from traveling. I felt like I could talk to people. Started going to this public speaking workshop called Toastmasters and I was, you know, feeling really confident in myself, giving like one-minute speeches.
The audience hears the confidence-talk. Their prediction-machinery activates: is this going to hold up? Is he going to be tested? They start tracking the prediction. The piece continues until Lewis gets asked to do a full speech and lead the evening — and he says no, because he was nervous. The crystal ball refutes. The audience experiences the moment of refutation hard, because they had been tracking the confidence-prediction since the opening.
Lewis: The crystal ball, you could argue, was planted there to make the audience ponder, "Ah, he's giving it the big about his charisma and confidence. Is that going to hold up to be true in the future?" I set something up, I set an assumption, and then I derailed it by saying no to the public speaking opportunity.7
The case study reveals: crystal balls can be either self-assertions (Lewis says he's confident — refuted later) or external predictions (Matt says he'll get the spoon — eventually subverted). Both forms work. What unites them is the explicit out-loud-statement that activates the audience's prediction-machinery.
A second crystal ball type Dicks uses regularly: the narrator-from-the-future prediction. I didn't know it then, but I was about to meet the woman I'd marry. The narrator-self predicts the protagonist-self's future. The audience now tracks the prediction across the piece. This is essentially the same mechanism deployed from the storyteller's retrospective vantage rather than from a character's current vantage.
You have a draft. You're going to install one crystal ball in the opening to keep the audience from leaving in the first 90 seconds.
Read your opening aloud. Identify what the audience will be wondering after the first 30 seconds. What is the metal object? Will the speech go well? Why is Sarah looking at him that way? These wonderings are implicit crystal balls — the audience is already making predictions, you just haven't made them explicit yet.
Now ask: would making one of these predictions explicit — having the storyteller voice it out loud — strengthen the audience's tracking? If yes, draft the crystal-ball sentence. I knew when I saw that look on her face that the night was going to end badly. That's a crystal ball. I knew when I walked into the room that I was about to embarrass myself. Same. I had no idea then, but in twenty minutes I'd be hugging a stranger who'd just saved my life. Same.
Place the crystal-ball sentence early — usually in the first or second paragraph. Then proceed with the story. Resolve the prediction at the planned moment. The audience will track from installation to resolution without dropping the thread.
Working storytellers report that 5-minute pieces typically deploy one crystal ball; longer pieces can deploy two or three. More than three usually feels engineered. The technique is potent enough that one well-placed crystal ball usually suffices.
Your prediction is too vague to track. Things were going to get interesting is not a crystal ball; it's a mood. A crystal ball is I was going to get fired by Friday — something specific the audience can hold and check.
You over-explain the prediction. And I had this feeling, this premonition, this sense of doom that I would later realize was actually pointing toward... The prediction is buried under qualifications. Make it clean: one sentence.
You install the crystal ball but don't resolve it. The audience tracks the prediction throughout the piece. If the piece ends without addressing whether the prediction was right or wrong, the audience feels unresolved. Always resolve installed crystal balls — confirmed, refuted, or split.
You install too many. Three crystal balls in a 5-minute piece is the working ceiling. More creates tracking-fatigue.
Your crystal ball is so obviously going to be refuted that no real tracking happens. I thought it was going to be the best day of my life. It wasn't. If the refutation is telegraphed too obviously, the audience doesn't really track — they coast on the obvious eventual reveal. The best crystal balls have some genuine uncertainty in their outcome.
The crystal-ball technique is named explicitly in the Knowledge Project interview as part of Dicks's engagement-technique taxonomy.8 It is treated as a foundational mid-story technique in Lewis's masterclass synthesis.9 The earnings-guidance and gambling analogues confirm the technique operates on universal cognitive architecture, not just on storytelling-specific psychology.
A real tension exists with breadcrumbs: both techniques manage audience-expectation about future moments, but breadcrumbs work understated (some audience members miss them) while crystal balls work explicit (no audience member misses them). The distinction matters in revision: when you want a moment to be felt by all listeners, use a crystal ball; when you want some listeners to feel it and others not to, use a breadcrumb.
A tension also exists with end with heart discipline: stories that end with explicit prediction-resolution (the crystal ball was right / wrong) can feel mechanically constructed if the resolution overshadows the heart-landing. Working storytellers learn to resolve the crystal ball before the heart-landing so the prediction-tracking doesn't compete with the emotional close.
Open question: do crystal balls work in first-person-present-tense storytelling? The technique relies on a narrator-vantage that can predict the future of the story. Pure present-tense storytelling (we're in the moment, no narrative remove) doesn't have access to this vantage. Whether crystal balls can work without narrative-remove or whether the technique requires retrospective framing is open.
Dicks and Bilton converge on the prediction-installation discipline. Bilton's cliffhanger architecture for narrative non-fiction often deploys chapter-end predictions that fire one or two chapters later. Both writers treat explicit future-claims as load-bearing audience-attention tools. Where they diverge: Bilton's medium (print) allows the reader to put the book down and resume later, so cliffhangers need to be strong enough to survive the gap; Dicks's medium (oral) keeps the audience in the room, so weaker prediction-installation suffices. Same architectural principle calibrated for different consumption-modes.
Dicks and Whyte split on this technique. Whyte's prolonged attention as poetic method explicitly avoids prediction-framing — Whyte's poetic discipline is to stay with the present moment without imposing prediction on it. The split reveals: prediction-machinery is one attention-mechanism but not the only one. Contemplative aesthetic traditions deliberately bypass prediction-engagement and produce a different kind of audience-attention through sustained present-moment focus. Both modes can be right; they engage different audience-cognition systems.
The crystal ball is fundamentally an explicit-prediction-installation technique — a way of activating universal prediction-machinery to lock audience attention. That move has direct analogues in adjacent disciplines that depend on engagement-design.
Behavioral mechanics: Six-Minute X-Ray (Hughes) — Hughes's elicitation work uses anchored expectations — installing a stated expectation early in a conversation that the target's later behavior will confirm or refute. The mechanism is structurally identical to the crystal ball: explicit future-claim activates the target's prediction-tracking, target's subsequent behavior either confirms or reveals the gap. What neither domain produces alone: Dicks teaches the storyteller to install crystal balls for audience attention-locking; Hughes teaches the elicitor to install expectations for target-disclosure diagnostic. The cross-orientation parallel reveals that prediction-machinery activation is a universal cognitive hook that crosses craft-orientations. Whether the practitioner's goal is cooperative (storytelling) or adversarial (elicitation), the same explicit-future-claim mechanism produces the engagement-lock that the practitioner needs. The technique transfers cleanly between contexts with different intent.
Eastern spirituality: Dhvani — Suggestion as Soul of Content — Sanskrit aesthetic theory holds that explicit statement (vacya) is one of three communication modes alongside implication (lakshana) and suggestion (vyanjana). The crystal ball operates squarely in vacya mode — explicit out-loud statement of future intent. Classical aesthetics treats vacya as the lowest-status of the three modes (suggestion is highest), but recognizes it as the most reliable for engagement-installation — the audience cannot miss what was stated explicitly. The crystal ball is Dicks's working storyteller's recognition of the same trade-off: when you need the audience to definitely catch a prediction, use explicit statement (vacya / crystal ball); when you want only some audience members to catch it, use suggestion (vyanjana / breadcrumb). What neither domain produces alone: classical Sanskrit theory supplies the full three-mode taxonomy of communication; Dicks supplies the specific working technique (out-loud prediction) optimized for one of those modes. The cross-tradition handshake reveals that crystal balls are not new — they're the working storyteller's recognition of a communication-mode that classical aesthetics named centuries ago, with the explicit understanding that the mode is appropriate for some moments and not others.
The Sharpest Implication The crystal ball implies that most amateur storytellers are leaving the most reliable attention-tool unused. Out-loud prediction is structurally simple — one sentence — and reliably activates the deepest attention-mechanism humans have. Yet most amateurs avoid it because it feels too direct, too obvious, too unsubtle. The avoidance is misguided: the audience does not experience explicit prediction as heavy-handed; they experience it as being included in the storyteller's vantage. The technique that feels too direct is the one the audience welcomes most. The misallocation is real and structural — beginners refuse the most effective tool because it doesn't feel sophisticated, and their work suffers for it.
Generative Questions