Checkpoints¶
Partial results between questions - what each card can say, and what it costs to say it.
Contents¶
| File | Description |
|---|---|
| Axis closeness | Where the taker stands on one axis |
| Double axis puzzle | Guess which named position you are closest to |
| Event model | What fires a checkpoint, and when |
| Halfway through | Progress and time left, nothing personal |
| New trait | A trait unlocked at full agreement |
| Nolan chart path | The route travelled across the compass |
| Random copy | Varying the wording so cards do not repeat |
| Single axis puzzle | Guess which side of one axis you came out on |
| Stats chart | How rare the taker's answer is |
Context¶
A checkpoint is a card shown between two questions, telling the taker something true about themselves before the quiz is over. It is the payoff the gamification bet depends on: the reward has to feel closer with every question, and a progress bar cannot carry that alone.
What every card shares:
- One card, one finding - a headline, one sentence, and a visual borrowed from the result modules.
- Lead-in plus statement - a grey opener that sets the tone, then the finding in bold. "We already know this - your radicalism is high."
- The result screen's visuals - checkpoints render the same charts as the result, so mid-quiz and final look like one product.
- An opt-out on every card - turning checkpoints off is always one tap away, and the system has to degrade to showing nothing at all.
- Fired, not scheduled - the event model decides which card wins and whether pacing allows it; a card that cannot be shown now is dropped, not queued.
Two constraints apply to all of them. Confidence - a card only fires when the number behind it is stable, because being wrong mid-quiz costs more than saying nothing. Zero configuration - types work without an author writing any copy, because community quizzes ship without touching the settings.