Fun Audit - Orbit Snatch (`orbit-snatch`)

Date: 2026-05-28
Auditor: Cursor agent using fun-audit
Evidence: Browser playtest, code/design read, route pre-flight, unit test smoke

Scorecard

🎮 Orbit Snatch — Fun Scorecard
Intent: Hard Fun + Sensory Fun | Session: 30-90s | Mode: Solo + MP
 
FUN STACK     3.0/5  (L1-L8: 4,5,3,2,2,3,2,3)
CORE LOOP     3 PASS / 2 WEAK / 1 BROKEN
FUN TYPES     Hard: 2/5 ⚠️ | Sensory: 3/5 ⚠️ | Social: ?/5
TOP KILLERS   1. Failure does not teach  2. Combo/payoff is underexposed  3. Agency is opaque
PEAK MOMENT   No-input novice run reached Score 17 and set Best 17
DEAD MOMENT   Active flip runs died at Score 1 with only "Orbit Broken"
NEXT          → fun-plan: fair-failure telegraph + death recap + combo/payoff clarity

Phase 0 — Fun Intent

Field	Value
Primary fun type	Hard Fun
Secondary fun type	Sensory Fun
Tertiary / MP target	Social Fun, unscored due no full multiplayer run in this audit
Emotional promise	”One-button mastery panic: flip at the perfect moment, snatch gems, dodge spikes, and chase one more best.”
Session shape	30-90s arcade burst
Mode	Solo practice observed; multiplayer shell/lobby visible but no second-player match observed
Anti-goals	No grind, no unfair deaths, no confusion, no griefing/toxic MP incentives

Phase 1 — Evidence Summary

Run / source	Persona	Notes
Pre-flight	Platform	`funday-frontend.service` active; `/play/orbit-snatch` returned HTTP 200. Hydration produced menu, canvas, HUD, dock, drawer.
Browser load	Novice first 10s	Initial screenshot showed the Funday shell before hydration; after ~4.5s body text included menu, “Solo Practice”, “Tap · Space · Click to flip”, canvas 722x722.
Run 1	Novice / no strategy	Clicked Solo Practice, no active input for ~10s. Run ended with `Orbit Broken`, Score 17, Best 17. Shows game can generate score and replay state without player mastery.
Run 2	Greedy / rapid flips	Fresh tab, clicked Solo Practice, rapid canvas clicks every ~250ms. Run ended quickly with `Orbit Broken`, Score 1, Best 17.
Run 3	Competent rhythm attempt	Fresh tab, clicked Solo Practice, spaced flips around 1.4-2.8s cadence. Run ended quickly with `Orbit Broken`, Score 1, Best 17.
Screenshot	Failure state	Captured `orbit-snatch-fun-audit-end.png`: overlay says `Orbit Broken`, `Score 1 · Best 17`, `Again`; drawer shows `No open matches`.
Code/design	Intent	Manifest defines one-button controls, objectives, maxPlayers 8, leaderboards; design doc promises combo streaks and 30-90s mastery.
Tests	Engine smoke	`games/orbit-snatch/index`: 11/11 unit tests passed.

Phase 2 — Fun Stack

Level	Score	Evidence
1 Usability	4	Route returns 200, frontend service active, menu/canvas/HUD appear, Solo Practice starts. Minor friction: first navigation screenshot showed a blank play shell before hydration completed.
2 Clarity	5	Menu clearly states “Flip direction. Snatch gems. Chain combos. Dodge spikes.” Controls are visible: “Tap · Space · Click to flip” and “TAP TO FLIP”.
3 Responsiveness	3	Canvas clicks trigger game outcomes, score updates, and death state. However flip feedback itself is subtle in text/DOM evidence; no visible “direction changed” cue or death cause is shown.
4 Competence	2	Observed active play attempts both died at Score 1, while no-input play reached Score 17. Failure screen does not teach why death happened or what to adjust.
5 Agency	2	The only clear solo choice is when to flip, but observed evidence makes the strategic value opaque: doing nothing outperformed active flipping.
6 Emotion	3	`Orbit Broken`, score, best score, and `Again` create a light tension/retry loop. No strong peak payoff was visible in observed active runs.
7 Meaning	2	Best score gives minimal persistence/status, but no memorable identity, story, or social result was observed.
8 Memory	3	`Again` + Best 17 supports “one more try”; replay intent is weakened because failure gives little learning and active strategy felt punished.

Average: 3.0/5

Pass bar result: FAIL for shipping feel work. Average is below 3.5, and intent-critical Competence/Agency are at 2.

Phase 3 — Core Loop

Link	Status	Evidence
Goal	PASS	Menu and manifest communicate collect gems, chain combos, dodge spikes.
Action	PASS	One action is clear: tap/space/click to flip direction.
Feedback	WEAK	Score/death feedback exists, but flip direction, hazard threat, combo state, and cause of death are under-communicated in observed runs.
Learning	BROKEN	Failure screen says `Orbit Broken` but not what hit the player, where the hazard was, or why a different flip timing would help.
Reward	WEAK	Score/Best works; combo payoff was not visible in the failed active runs and no-input score gain weakens the sense of earned mastery.
New Choice	PASS	`Again` and `Best` create a retry loop.

Phase 4 — Fun Type Coverage

Type	Target	Actual	Gap	Hypothesis
Hard Fun	Fair challenge + learnable failure	2/5	Death is fast and under-explained; active attempts did not create visible improvement.	Add hazard telegraph/death recap and first-run teaching so each death explains the next better move.
Sensory Fun	”That feels amazing” input/collect/combo payoff	3/5	Canvas, sound toggle, score, and overlay exist, but active runs did not expose a strong peak or combo celebration.	Make combo tiers and near-misses more legible with escalating pulse/pitch/text feedback.
Social Fun	Rivalry/shared ring in MP	?/5	Lobby/drawer visible and manifest supports MP, but no multiplayer match was observed.	Run a separate MP playtest with 2+ sessions before scoring.
System Fun	Risk/reward route optimization	2/5	Gem/hazard systems exist in code, but player-facing strategy is opaque in observed play.	Expose route choice: show upcoming gem/hazard windows and reward intentional flips.

Phase 5 — Fun Killers

Killer	Observed?	Severity	Evidence	Fix direction
Player feels stupid	Yes	🔴 fun-blocker	Active rhythm and rapid flip attempts both ended at Score 1 with no explanation.	Death recap: “spike hit at top-left”, replay ghost, or hazard highlight after death.
Player quits after failing	Risk	🔴 fun-blocker	Failure feels under-taught; `Again` exists but does not explain next action.	Clearer telegraphs + faster/educational retry.
Player uses one dominant strategy	Yes, inverted	🔴 fun-blocker	No-input novice run reached Score 17; active strategies reached Score 1.	Make skillful flips reliably outperform passivity, or teach why/when passivity is correct.
Player wins but feels nothing	Partial	🟡 friction	Score/Best state exists, but combo/peak moments were not visible in active runs.	Escalating combo celebration and summary of best action.
Player feels lost	No	🟢 polish	Menu objective and controls are clear.	Keep current concise onboarding.
Multiplayer gets toxic	Unknown	?	No full MP run observed.	Separate MP audit needed.

Phase 6 — Moment Map

Moment	Expected	Actual	Severity	Evidence
First 10 seconds	Clear goal, obvious action, instant start	Menu copy and controls are clear after hydration; first screenshot before hydration was blank shell.	🟡 friction	Body text after ~4.5s: title, objective, Solo Practice, control hint.
First action	Flip feels meaningful	Game responds, but flip feedback is subtle; no explicit direction-change cue in evidence.	🟡 friction	Active runs changed outcomes but did not visibly teach timing.
First meaningful failure	Fair failure teaches next attempt	`Orbit Broken` with score only; no cause, hazard highlight, or timing lesson.	🔴 blocker	Greedy and rhythm runs ended Score 1.
First peak	Combo/snatch dopamine spike	Novice no-input run reached Score 17, but no observed combo peak or celebratory summary.	🟡 friction	End overlay only reported Score/Best.
First drop	Attention drops when action feels punished	Active input attempts performed worse than no-input run.	🔴 blocker	Score 1 active vs Score 17 no-input.
End/replay	”One more, I know what to do”	`Again` and Best invite retry, but lesson is unclear.	🟡 friction	Failure overlay: `Again`, `Score 1 · Best 17`.

Phase 7 — Prioritized Findings

#	Finding	Tag	I	C	E	Priority	Route to plan
1	Failure does not teach; death screen lacks cause and next-action lesson	🔴	5	5	2	12.5	`fun-plan`: death recap + hazard highlight + first-failure coaching
2	Skill/agency is opaque; observed no-input run beat active flipping	🔴	5	4	3	6.7	`fun-plan`: tune/teach flip timing and make intentional flips outperform passivity
3	Combo/peak payoff is underexposed in observed play	🟡	4	4	2	8.0	`fun-plan`: visible combo tier feedback, end-run “best chain”, stronger collect celebration

Sorted by practical execution: fix #1 first, #3 second (cheap payoff clarity), #2 third (may require balance/engine tuning).

Cross-Skill Notes

Skill	Finding
funday-play-shell	Route/service healthy (`active`, HTTP 200). Initial blank shell before hydration was observed but resolved within ~4.5s. Not a current blocker.
audit	UI/accessibility audit not performed here. Canvas has an aria label and controls are visible, but full a11y/perf pass is separate.
critique	Visual hierarchy is workable: score top, overlay centered, controls concise. Death explanation and reward hierarchy need UX critique in plan.
svelte-code-writer	Svelte code read for audit context only; no edits made.

Backlog (defer)

Full multiplayer Social Fun audit with 2+ clients/sessions.
Telemetry hooks for death reason, time-to-first-score, flip count, combo distribution, retry count.
Hydration/perceived loading polish if first-screen blank persists for real users.
Final audit pass for a11y/performance/responsiveness after feel blockers are addressed.

Sign-off

Audit complete. Review this doc, then invoke fun-plan for the top 3 fixes.

Orbit Snatch Audit