214 lines
8.2 KiB
Markdown
214 lines
8.2 KiB
Markdown
# Question bank gamification
|
|
|
|
## Objective
|
|
|
|
Make studying easy to begin, satisfying to continue, and valuable to repeat while
|
|
optimizing for durable learning rather than compulsive use. The core engagement
|
|
loop is:
|
|
|
|
```text
|
|
Answer -> immediate insight -> visible progress -> useful reward -> clear next goal
|
|
```
|
|
|
|
Success means learners return consistently, improve weak areas, complete spaced
|
|
reviews, and become more exam-ready. Raw time in the application is not itself a
|
|
success metric.
|
|
|
|
## Product principles
|
|
|
|
- Reward demonstrated learning, recovery from errors, and breadth of coverage.
|
|
- Make reward rules understandable and resistant to farming.
|
|
- Give every study session a user-selected finish line and satisfying conclusion.
|
|
- Keep stopping controls as prominent as continuation controls.
|
|
- Do not lock core questions, explanations, or remediation behind RZ Coin.
|
|
- Prefer weekly consistency over an indefinitely escalating daily streak.
|
|
- Keep social participation optional and private by default.
|
|
- Never use paid randomized rewards, near-miss effects, disguised odds, or
|
|
loss-heavy messaging.
|
|
|
|
## Mastery paths
|
|
|
|
Show mastery for each system, topic, and tested concept. Mastery should combine:
|
|
|
|
- First-attempt correctness
|
|
- Question difficulty
|
|
- Recency and spaced-review status
|
|
- Successful recovery after an incorrect answer
|
|
- Breadth across related concepts
|
|
|
|
Mastery must not be a reskinned count of questions answered. Each topic should
|
|
show its current state, evidence behind that state, and the next useful action.
|
|
Suggested states are `Unseen`, `Learning`, `Practicing`, `Strong`, and `Mastered`.
|
|
|
|
The dashboard should highlight progress in weak areas and recently mastered
|
|
concepts. A system cannot reach its highest rank by repeatedly farming a narrow
|
|
set of easy questions.
|
|
|
|
## RZ Coin economy
|
|
|
|
Award RZ Coin for learning behaviors:
|
|
|
|
- A first correct attempt
|
|
- Correctly answering a genuinely difficult question
|
|
- Completing a scheduled review
|
|
- Converting a previously missed concept into a correct delayed review
|
|
- Improving a weak topic
|
|
- Completing a timed or simulated-exam block
|
|
- Meeting a balanced daily or weekly study goal
|
|
|
|
Do not primarily reward elapsed time, retries, repeated easy questions, or raw
|
|
volume. Apply diminishing or zero rewards to repeated content where necessary.
|
|
All awards should be server-authoritative, idempotent, and recorded in the coin
|
|
ledger with a reason code.
|
|
|
|
Appropriate purchases include:
|
|
|
|
- Question-bank reset tokens
|
|
- Cosmetic themes, avatars, profile frames, titles, and study environments
|
|
- Optional analytics presentations
|
|
- Limited streak protection
|
|
- Collectible medical artifacts and achievement displays
|
|
|
|
Purchases must disclose their exact price and effect. Randomized paid purchases
|
|
and purchasable power over exam scoring or question correctness are prohibited.
|
|
|
|
## Daily and weekly missions
|
|
|
|
Offer three short daily missions that reflect the learner's needs, for example:
|
|
|
|
- Complete 10 scheduled reviews.
|
|
- Answer five difficulty-4 questions.
|
|
- Improve one weak system.
|
|
- Complete a timed block without skipping.
|
|
|
|
Allow a learner to replace a mission that does not fit the current study plan.
|
|
Missions should not require excessive daily volume or encourage answering after
|
|
fatigue has clearly degraded performance.
|
|
|
|
Weekly goals should emphasize consistency and breadth, such as studying on four
|
|
days, reviewing three systems, or clearing a reasonable number of overdue
|
|
reviews. A flexible weekly target is preferred to an endless daily streak.
|
|
|
|
## Session goals and completion
|
|
|
|
At the beginning of a session, let the learner choose a finish line:
|
|
|
|
- A number of questions
|
|
- A study duration
|
|
- One weak topic
|
|
- One simulated-exam block
|
|
- All currently due reviews, subject to a reasonable cap
|
|
|
|
At the finish line, present a completion screen before offering more work. The
|
|
primary actions should give equal visual weight to `Finish for today` and a
|
|
clearly sized optional next block. Do not create artificial unfinished states to
|
|
pressure continuation.
|
|
|
|
The end-of-session summary should include:
|
|
|
|
- Questions completed and first-attempt accuracy
|
|
- Concepts newly mastered
|
|
- Previously missed concepts recovered
|
|
- Weak areas improved
|
|
- RZ Coin earned, with award reasons
|
|
- Readiness change, with an explanation of how it was calculated
|
|
- A recommended next session
|
|
|
|
## Progression and personal records
|
|
|
|
Introduce skill-based ranks such as `Intern`, `Resident`, `Chief`, and
|
|
`Attending`. Rank advancement should require mastery across several systems,
|
|
successful delayed reviews, and performance on appropriately difficult items.
|
|
It must not depend only on account age or total answers.
|
|
|
|
Personal records should measure meaningful improvement:
|
|
|
|
- Best timed-block accuracy at a defined difficulty mix
|
|
- Longest first-attempt correct streak at difficulty 4 or higher
|
|
- Most concepts mastered in a week
|
|
- Largest reduction in overdue reviews
|
|
- Greatest improvement in a previously weak content area
|
|
|
|
Record definitions must remain stable and visible so the learner understands
|
|
what was achieved.
|
|
|
|
## Social features
|
|
|
|
Potential opt-in features include private study groups, cooperative weekly
|
|
goals, friend challenges, and anonymous percentile bands. Prefer cooperative
|
|
goals over leaderboards based on raw volume.
|
|
|
|
Do not expose failure, study time, predicted score, or inactivity publicly by
|
|
default. Users must be able to leave a group, hide a statistic, and disable
|
|
social notifications.
|
|
|
|
## Notifications and streak safeguards
|
|
|
|
- Let users choose notification categories, quiet hours, and frequency.
|
|
- Use factual reminders rather than guilt, urgency, or threatened loss.
|
|
- Allow recovery days and planned breaks.
|
|
- Cap streak-protection use and describe it transparently.
|
|
- Never send repeated prompts after a learner chooses to finish a session.
|
|
- Detect unusually long sessions and offer a neutral break reminder.
|
|
|
|
## Recommended delivery order
|
|
|
|
### Phase 1: foundation
|
|
|
|
1. Define the mastery model and expose topic-level progress.
|
|
2. Add server-authoritative RZ Coin awards for approved learning events.
|
|
3. Add user-selected session goals and an end-of-session summary.
|
|
4. Add daily missions with one replacement per day.
|
|
|
|
### Phase 2: progression
|
|
|
|
1. Add weekly consistency goals.
|
|
2. Add breadth-based ranks and personal records.
|
|
3. Expand the RZ Coin store with cosmetic and utility items.
|
|
4. Add limited streak protection and recovery days.
|
|
|
|
### Phase 3: social and optimization
|
|
|
|
1. Add private cooperative groups and friend challenges.
|
|
2. Personalize missions using weak-area and spaced-review data.
|
|
3. Experiment with presentation and pacing under the guardrails below.
|
|
|
|
## Measurement
|
|
|
|
Primary metrics should measure learning and healthy retention:
|
|
|
|
- Weekly active learners completing at least one meaningful study goal
|
|
- Scheduled-review completion rate
|
|
- Delayed retention after 7, 21, and 45 days
|
|
- Recovery rate for previously missed concepts
|
|
- Breadth of systems studied
|
|
- Timed-block and simulation improvement
|
|
- Percentage of sessions ending at the learner's selected finish line
|
|
|
|
Monitor guardrail metrics alongside product experiments:
|
|
|
|
- Extremely long sessions and late-night continuation
|
|
- Rapid repetitive answering consistent with coin farming
|
|
- Notification opt-out and mute rates
|
|
- Mission abandonment after oversized goals
|
|
- Increased accuracy volatility associated with fatigue
|
|
- Support reports involving pressure, anxiety, or confusing rewards
|
|
|
|
Do not optimize an experiment solely for session length, questions per session,
|
|
notification opens, or coin spending. A change should improve at least one
|
|
learning or healthy-retention metric without materially degrading the guardrails.
|
|
|
|
## Initial acceptance criteria
|
|
|
|
The first release is complete when:
|
|
|
|
- Mastery is computed from correctness, difficulty, recency, and review outcomes.
|
|
- Coin awards are idempotent and cannot be earned repeatedly from the same event.
|
|
- Every award and purchase has a ledger entry and visible reason.
|
|
- A learner can set and complete a bounded session goal.
|
|
- The completion screen offers both stopping and continuing without deceptive
|
|
emphasis.
|
|
- Daily missions target due reviews or genuine coverage gaps and can be replaced.
|
|
- Core educational content remains available without coin purchases.
|
|
- Product analytics include the primary and guardrail metrics above.
|