Peggy’s Party

Current State

Focus

Peggy’s Party — Multi-Activity Guests and Party Styling

Validation Question

Parked: what new evidence or concrete hypothesis would justify revisiting Peggy’s Party? No active validation work scheduled.

Setup

ready

Primary Reference

Airport Empire

Reference Thesis

Peggy’s Party should reproduce Airport Empire’s first-airport moving-customer idle loop closely enough to test the formula in the party fantasy: visible guests enter, travel to capacity-limited stations, wait and process, produce manually collected rewards, leave, and create throughput pressure that station upgrades visibly relieve.

Open Questions

• Do players understand and repeat invite → style → multiple activities → expand, and does this package improve sustained engagement over pparty05 without harming first-minute comprehension?
• If Peggy is revisited, what new evidence or concrete hypothesis would justify renewed investment?

Next Actions

• No further Peggy implementation, build or submission is scheduled. Return development focus to Fruits Inc; revisit Peggy only on explicit user direction.

Latest Findings

• Latest Poki Player Fit for pparty0061.zip: average session 77.0s, p90 160.0s, median 41.0s, under-3m sessions 91.4%, Poki start rate 68.4%. Snapshot note: Session time reduced again
• Latest stored pparty0061.zip: 77.03s average session, 41s median, 160s p90, 8.6% three-minute reach and 68.4% gameplay start.
• Historical pparty05 comparator: 100.01s average, 50s median, 171s p90, 9.8% three-minute reach and 82.4% gameplay start. Latest results did not show the intended directional improvement.
• Túlio requests parking rather than another iteration; his subjective assessment is part of the decision.
• The current package has not justified further investment for the user. Park the project with its history intact. These bundled, non-randomized results do not establish that customization or multi-activity visits individually caused the decline, nor constitute permanent abandonment.

Updated

2026-10-18T16:10:59.216767Z

Active Hypotheses

A dense, already-busy opening will make Peggy’s Party’s core fantasy legible faster than the previous sparse/onboarding-led starts, improving first impression and early-session engagement without requiring a broad redesign.

Previous evidence showed onboarding changes did not turn around first impression or engagement, and boredom remained dominant. The remaining low-cost question is whether the game failed partly because it did not immediately look and feel like a lively party.

The existing gameplay may find a better audience on CrazyGames than Poki without product redesign.

Making customized guests enjoy several activities, supported by compact upgrades and clearer portrait interactions, will strengthen the party-host fantasy and sustained engagement compared with Peggy's previous busy-opening baseline.

The prior busy opening showed directional improvement but weak sustained engagement. This approved package applies lessons from Fruits Inc while making guest styling and repeated activities central to Peggy.

Latest Experiment

Peggy’s Party — Multi-Activity Guests and Party Styling

decided

Do players understand and repeat invite → style → multiple activities → expand, and does this package improve sustained engagement over pparty05 without harming first-minute comprehension?

Implemented candidate on codex/peggys-party: gameplay/layout checkpoint 6d68d54, analytics 41aa073, and companion build configuration checkpoint cbaf9cd selecting PeggyScene and product version 0.06. Working tree clean at handoff. User accepts layout and will build/upload to Poki Player Fit and playtest. Separate from the prior CrazyGames baseline; no upload or new results claimed.

Expected Result

Directional improvement in median session length and three-minute reach versus pparty05, supported by recordings showing repeat invitations, meaningful styling and continued visits/upgrades. Treat the package as one experiment; it cannot isolate the effect of dress-up from progression/layout changes.

Prototype Scope

Four seeded guests; phone-driven invitations and portrait styling; multi-activity visits with reserved homes and eight-guest cap; five levels per station and three positions; progressive colors at levels 5/7/9; safe expansions/save migration; portrait courtyard and fixed camera; raised signs and phone availability sign; clearer hints/upgrade comparisons; transitions/celebrations and cohort-aware analytics. Existing art placeholders and legacy onboarding retained.

Non-Goals

• Saved looks or guest re-editing
• New bespoke art, automation, additional activities or monetization
• Rewriting prior Poki or CrazyGames evidence
• Claiming causal feature-level uplift from a bundled iteration

Evidence Capture

artifact

player_fit_id · configured_version · playtest_id · branch · commit · gameplay_checkpoint · filename

validation

Verify actual WebGL build and Poki Inspector delivery before interpreting missing analytics. Preserve seeded/invited and fresh/resumed cohorts. Attach future Player Fit and playtest records to this experiment.

qualitative

Can players find the phone, customize, dispatch and collect without confusion? · Do players recognize the same guest enjoying multiple activities and choose to invite again? · Does styling interest persist, and do compact upgrades relieve visible bottlenecks? · Check portrait touch/readability, full party, rotation and legacy onboarding.

comparison

pparty05.zip historical Poki baseline: 100.01s mean, 50s median, 171s p90, 9.8% three-minute reach, 82.4% gameplay start. Record sample, device mix and observation windows; comparison is directional, not randomized.

quantitative

Median/mean/p90 session length, three-minute reach and gameplay start · First invited guest first collection, second activity and visit completion · Repeat invitation thresholds, styling edits versus unchanged confirmations, color use · Upgrade purchase versus applied expansion and blocked dispatch resolution

owner

Túlio builds and uploads; waiting for submission and evidence, with no deadline assigned.

Sources: 4bb93cc9-6160-466e-ad03-294fab0474b0

Latest Analysis

Park for now after pparty0061 results and creator assessment

Túlio explicitly chooses to park Peggy’s Party for now and return focus to Fruits Inc, based on latest results and his own experience of the game.

Findings

• Latest stored pparty0061.zip: 77.03s average session, 41s median, 160s p90, 8.6% three-minute reach and 68.4% gameplay start.
• Historical pparty05 comparator: 100.01s average, 50s median, 171s p90, 9.8% three-minute reach and 82.4% gameplay start. Latest results did not show the intended directional improvement.
• Túlio requests parking rather than another iteration; his subjective assessment is part of the decision.

Interpretation

The current package has not justified further investment for the user. Park the project with its history intact. These bundled, non-randomized results do not establish that customization or multi-activity visits individually caused the decline, nor constitute permanent abandonment.

Open Questions

• If Peggy is revisited, what new evidence or concrete hypothesis would justify renewed investment?

Next Action

No further Peggy implementation, build or submission is scheduled. Return development focus to Fruits Inc; revisit Peggy only on explicit user direction.

Latest Player Fit

pparty0061.zip

Imported: 2026-10-18T16:08:43.967819Z · Captured: 2026-10-18T05:36:01Z

Latest Poki Player Fit for pparty0061.zip: average session 77.0s, p90 160.0s, median 41.0s, under-3m sessions 91.4%, Poki start rate 68.4%. Snapshot note: Session time reduced again

No live performance evidence imported yet.

Latest Decision

None