The User Engagement Plan — Making Dilder Feel Alive¶
What makes a virtual pet worth carrying around? Not the hardware. Not the display. It's the feeling that something in your pocket needs you — and that it notices when you're there. The user engagement plan is the design document for that feeling.
Beyond the Tamagotchi¶
The original Tamagotchi had a simple loop: stats decay, you feed/clean/play, the pet grows. It worked because the obligation was real — neglect it and it dies. But it was also frustrating. Miss an afternoon and your pet is gone.
Dilder takes a different approach: no permanent death, but real consequences. Neglect makes the octopus visibly suffer (sad, hungry, homesick), but recovery is always possible. The goal is the sweet spot between Tamagotchi's harshness and modern apps' toothlessness.
The Stat System¶
Five primary stats decay in real time:
| Stat | Decay Rate | Refill |
|---|---|---|
| Hunger | -1 / 10 min | Feed action |
| Happiness | -1 / 15 min | Pet, play, walk |
| Energy | -1 / 12 min | Sleep |
| Hygiene | -1 / 30 min | Clean action |
| Health | Static unless neglected | Maintain other stats |
Plus secondary stats that accumulate over time: Bond Level, Discipline, Intelligence, Fitness, Exploration. These drive evolution paths and unlock features.
Sensor-Driven Emotions¶
This is where Dilder diverges from every other virtual pet. The 16 emotional states aren't just driven by stats — they respond to the physical world through sensors:
- Bright light wakes the octopus up, darkness triggers sleep
- Loud noise startles it, sustained silence makes it lonely
- Gentle touch (capacitive pads) boosts happiness
- Walking (accelerometer) builds fitness and triggers adventure mode
- Temperature drops make it shiver, humidity makes it happy (it's an octopus)
The emotion engine blends all these inputs: EMOTION = f(stats, sensors, recent_events, personality). Transitions aren't instant — emotions fade between states over 2-3 animation frames.
Life Stages and Evolution¶
Dilder grows through six life stages over 30+ real days:
The adult form depends on how the pet was raised. High intelligence + high bond = Deep-Sea Scholar (glasses, philosophical quotes). High fitness + exploration = Reef Guardian (muscular, encouraging). Low discipline + chaos = Tidal Trickster (jester hat, unpredictable).
Six distinct adult forms, each with unique visuals and dialogue. The evolution is fully deterministic — your care choices create the outcome, not random chance.
Treasure Hunts¶
With a GPS module (Phase 4B), Dilder generates real-world treasure hunts. Virtual "gifts" are placed at GPS coordinates within walking distance. The e-ink display shows a compass bearing and distance — turn the device into a pirate treasure map.
Treasures range from common (60% — snack items) to mythic (<1% — one-of-a-kind cosmetics). They spawn based on step milestones, new locations, and life stage transitions. Hunts expire after 24 hours, creating urgency without punishment.
Step Targets and Streaks¶
Daily step milestones drive both rewards and mood:
| Target | Steps | Effect |
|---|---|---|
| Bronze | 2,000 | +10 XP, happy octopus |
| Silver | 5,000 | +25 XP, bonus snack |
| Gold | 8,000 | +50 XP, cosmetic chance |
| Platinum | 12,000 | +100 XP, guaranteed cosmetic |
| Diamond | 20,000 | Full stat restore, rare cosmetic |
Weekly and monthly targets stack on top. A 30-day streak earns a +750 XP bond boost and an exclusive cosmetic. One free miss per week prevents frustration from sick days or travel.
The Research Behind It¶
The engagement plan is backed by research into 10+ virtual pet and activity games:
- Tamagotchi — deterministic evolution, care mistakes, stat decay rates (reverse-engineered from community documentation)
- Digimon — battle-based evolution (adapted as activity-based)
- Pokemon GO — step tracking, adventure sync, weekly rewards
- Pikmin Bloom — location-type cosmetics, ambient engagement
- Finch — real-world activity as pet nurturing (the closest modern analogue)
- Creatures — neural networks and artificial biochemistry (inspiration for emergent behavior)
- Nintendogs — touch and voice interaction patterns
What's Next¶
The engagement plan is the design document — implementation starts with Phase 3A (core game loop: stats, decay, care actions, emotion resolution). Sensor integration follows in Phase 3B-3C.
The full document is published on the Design docs page — 21 sections, from gameplay loops to hardware sensor specs to a phased rollout plan.