The Operating Picture: One Act of Play, Two Outputs
This page decides nothing. It exists because the company's two halves are documented in two excellent but separate threads — the player side (The Journey, Reward Economy, the quests thread, the spatial-wiki vision) and the data side (the World-Model Data Engine series) — and nothing showed them as the single machine they are. The unifying frame is the vision doc's line: one act of play, two outputs. Where one side has a goal the other side can't answer, this page marks the gap instead of hiding it.
The one loop
Neither side survives alone. The data engine without the player engine is a gig platform that burns out its fleet ("turning play into work" — the exact failure The Journey warns about). The player engine without the data engine is a community with no business model. Each side is the other side's reason to exist — the vision doc's phrase, and the operating principle of this page.
The two tests, applied to everything
The Journey defined the dual test every feature must pass; the data-engine series operationalized the first and inherited the second as a judgment call. Stated together, as the company-wide gate:
- Does it make the data more valuable? — measured (kept-yield, verified coverage units, benchmark deltas).
- Does it make the community stronger? — currently unmeasured. This is the biggest asymmetry in the company's goal structure, and closing it is the point of the Player Engine goals skeleton.
The paired-goals table
Every load-bearing goal on one side, paired with its counterpart on the other. GAP marks a pairing that doesn't exist yet.
| Data side (measured by) | Player side (measured by) | State of the pair |
|---|---|---|
| Naturalness slice — directed data must not drift from organic play (founding experiment step 7) | Quests must not feel like work — candidate metrics: quest NPS, voluntary quest re-uptake, free-play share after quest completion | GAP — player half undefined; proposal in the skeleton |
| Cost per verified coverage unit (experiment output) | Reward-economy solvency — bounty spend as cost-of-goods on signed revenue (Reward Economy) | Paired in principle; needs the coverage-unit basis type in the payments schema (decision #6) |
| Kept-yield / hours-to-Anto (quality of what's captured) | Retention & session depth (Player-Tier KPIs — partially plumbed) | Half-paired — KPI views exist, targets don't |
| Fleet quality — thin input regimes (KBM precision, narrators, full parties) | Recruiting quality funnel — who Kame + Mattia find and how they're vetted | Half-paired — the catalogue's needs have never been handed to recruiting as a spec |
| The founding experiment — directed beats passive at equal hours | GAP — no player-side founding experiment. Candidate: quest-cohort vs free-play-cohort retention/satisfaction A/B, pre-registered the same way | GAP — the skeleton's centerpiece proposal |
| Consent & provenance ledger — every clip traceable | Trust promises kept — "you keep your keys," "no chat, no voice, no personal data" (amended: no incoming voice; opt-in narration), payout reliability | GAP — promises exist, no trust metric (candidates: payout-dispute rate, consent-withdrawal rate, support sentiment) |
| Klindt POC gate — one probe demonstrably fixed | Phase-1 Pioneer gate — the 250-hour Sample Packet, 10 Pioneers | Paired and live — the two gates even share the same operators |
| Publisher partnerships — the rights-cleared premium tier | Spatial wiki — the player-facing product of the same reconstruction stack | Paired by architecture (same data, two surfaces); wiki is R&D |
| Tripwires T1–T6 (competitive/legal/technical) | GAP — no player-side tripwires (quest fatigue, gig-work perception, payout disputes, community trust events) | GAP — drafted in the skeleton for Ale's edit |
Where each side's canon lives
- Player engine: The Journey (narrative + the two tests) → Reward Economy (the spine) → Quests as a Data Product / Quests — Technical Design → EXP System → Player-Tier KPIs → Player Engine — Goals skeleton (proposed, Ale's to shape).
- Data engine: the World-Model Data Engine series (10 pages + roadmap sub-pages + the delegation loop), with the plain-language explainer and one-pager linked from its sidebar.
- The seam: the open-platform / spatial-wiki vision (one act of play, two outputs) and this page.
The two seam risks (what this page exists to watch)
- The engine cannibalizes the play. Bounty-optimized behavior crowds out joy; contributors become gig workers; the naturalness metric degrades on the data side after the community has already soured on the player side. Defense: the paired metrics above, and the rule that quest triage applies both tests, not just the first.
- The play outruns the proof. Community grows on promises the data side hasn't earned yet (the moat-side quiet KPI problem: claims must stay future-tense until proof exists). Defense: the gates — nothing external until the POC and experiment numbers are real.
Compiled 2026-07-08 from existing documents. Maintained alongside the data-engine series; when a GAP row closes, update this table first.