Skip to content
WalkRPG Wiki

Progression system — Model A + tuned keystone bonus (~20% inaccessibility gap)

D-016: Progression system — Model A + tuned keystone bonus (~20% inaccessibility gap)

D-006 Pillar 3 ratified the quest↔tree loop with informal ~70% quest grants / ~30% keystone unlocks. The surrounding progression system stayed unratified: no canonical Walker level concept, no XP curve, no per-quest grant table, no target node count vs total point count. Sub-phase 13-2 ran Round 1 of /research progression-system (mechanics-designer + narrative-designer dispatched in parallel by game-director). Round 1 converged on Model A (emergent level) with three CEO options on keystone-bonus tuning. CEO ratified option B: keystone bonus = 1 point, deliberate ~20% inaccessibility gap, ~75/25 split.

Concrete spec:

  • Walker has no level integer field. walker.level is a computed view: level = walker.tree_points_spent. No XP curve. No level-cap independent of point-cap.
  • Per-quest grant: flat 1 tree point on quest completion. Per-step beats grant faction reputation + lore unlocks; no tree points per step. Quest tier / step count / region depth do not affect the grant.
  • Per-keystone grant: 1 bonus point. Keystone allocates atomically (no point cost from bank) AND adds 1 point to the bank. Net: keystone “costs 0 to slot, gives 1 elsewhere.”
  • Target tree size: ~180 nodes (mid-band of D-006’s 150-200).
  • Quest count at full game: ~108. Keystone count: ~36 (6 leak-handler keystones per D-014 + ~30 others authored over the full game).
  • Total achievable points at full game: ~144 (108 quest × 1pt + 36 keystone × 1pt bonus).
  • Reach: ~80% — ~36 nodes mathematically unreachable per playthrough. By design. Forces choice; build identity emerges from path selection across replays; PoE-style sparse allocation rather than Diablo-2-paragon dense reach.
  • Quest-grant / keystone split: ~75% quest / ~25% keystone (108 / 144 = 75%). D-006 §Pillar 3 amended in this decision from “~70%” → “~75%” to match this math.
  • Player-facing surface (Home): Poziom <N> (large single number, reads as strona N rejestru per D-014) + Punkty w skarbcu: <M> (banner when M > 0) + Następny punkt: Quest <id> step <n>/<m> (small progress hint). No XP bar. Tree page shows <N> / 180 allocated.
  • Banner copy on grants (per D-011 chrome-ENG + prose-PL):
    • Quest grant: [ENG] "+1 Tree Point banked." / [PL] „Linijka oddana do składu. Spis przyjął."
    • Keystone unlock: [ENG] "Keystone unlocked. +1 Tree Point banked." / [PL] „Kamień węgielny postawiony. Spis dopisuje rozdział."

Why this shape:

  • Single number resonates with D-014 (Mapa-as-diegetic-register) and D-013 (editorial-commit grammar). The Walker’s progression IS the page count on the Cech Mapa. Two-number models (level + XP bar) import an arcade-game HUD pattern that fights both anchors; narrative-designer rated this dissonant.
  • Smallest Phase 13 engineering surface. No XP curve constants. No second schema field on Walker. No second progression endpoint. Accelerates dispatch of sub-phases 13-3, 13-4, 13-5.
  • Deliberate inaccessibility creates build identity. PoE/Last Epoch achieve build diversity through 8-10% reach across 1100-1300 nodes. WalkRPG’s denser tree (~180 nodes) cannot use that lever; the ~20% gap is the equivalent design move at WalkRPG’s density. Players choose path order AND which nodes to skip, not just which 8% to take.
  • Step-grinder cannot level under Model A is on-pillar, not off-pillar. Walking is energy (D-007), not XP. Tree points come from quests + keystones. The pillar separation is feature: WalkRPG is the walking-RPG where the loop wants you to walk AND quest, not walk alone.
  • D-013 endgame milestone preserved. “Walker level 50 has writing-rights into Spis Wszechrzeczy” (D-013 §Endgame implication, v2.0 four-Spisy arc) remains achievable: 50 tree points spent = ~35% of the ~144 total = comfortably mid-game by design.

Rejected alternatives:

  • Option A (keystone bonus = 2pt, ~60/40 split, ~100% reach): rejected — removes design tension, weakens build identity, drifts further from D-006 phrasing than option B.
  • Option C (Model C: decoupled level integer + XP-from-quests, exact 71/29 split): rejected — two progress numbers on Home dissonant with D-013/D-014, larger Phase 13 engineering surface, level-as-gate primitive has no Phase 13 use case (no level-gated content authored yet).
  • Model B (decoupled level + XP-from-steps): rejected at research Round 1 by narrative-designer — XP-from-steps collapses D-013’s przepisać grammar back into RPG-mobs-give-XP convention. Cannot reconcile with Mapa register.

D-006 §Pillar 3 amendment (light, non-superseding):

D-006 Pillar 3 text is amended to read “Quests are the source of ~75% of passive points” (was “~70%”). The amendment is editorial: D-006’s intent (quests drive the bulk of progression, keystones supplement) is preserved; only the ratio number sharpens to match D-016 math. D-006 remains canonical; D-016 ratifies the math beneath D-006’s phrasing.

Open follow-up work (NOT blockers; W/B-level, dispatched separately):

  1. W-level (mechanics-designer): Author wiki/src/content/docs/tree/progression.md with full D-016 math + worked example + build-identity implications.
  2. W-level (mechanics-designer): Confirm data/src/schemas/walker.ts carries treePointsSpent + treePointsBanked and NO level field. Author migration if drift.
  3. W-level (mechanics-designer): Retrofit Quest 001 + Quest 002 reward fields to grant flat 1 tree point per quest completion.
  4. B-level (mechanics-designer + narrative-designer): Surface differentiation for regional-arc completions — under flat 1pt grant, a 30-min arc final and a 5-min courier quest grant equal points. Should the arc final get a cosmetic / faction badge / region-bound title callout to compensate? Tunable post-13-5.
  5. B-level (mechanics-designer): Are the 6 leak-handler keystones (D-014) a distinct keystone class with potentially different bonus values? Default: same 1pt. Reviewable.
  6. W-level (ui-designer): Does pisarska linijka styling (from research artifact) extend to tree-page allocation indicator? Or strictly Home-bound? Wireframe decision.
  7. B-level (tech-architect + backend-engineer, Phase 13-3+): POST /quest/complete confirms +1 tree point side-effect; POST /tree/allocate confirms +1 bonus point on keystone allocation; document in API surface.
  8. D-level (future): If a future region/quest requires level-gate semantics (e.g. “Mglica access requires Poziom 30”), is level-gate added as additive amendment to D-016 or as new D-decision? Default: level-gate via keystone or faction rep, no level gate primitive. Re-litigate if Phase 14+ regional content demands it.

Reasoning: D-006 Pillar 3’s “70/30” wording was informal — ratifying the math beneath it was always going to require a math decision. The convergence of mechanics-designer (engineering surface + math fit) and narrative-designer (D-013 / D-014 resonance) on Model A made the shape decision trivial. The keystone-bonus tuning between 1pt and 2pt was the only real design call: 1pt yields ~80% reach and forces choice (deeper move, closer to D-006’s wording), 2pt yields ~100% reach and removes choice (cheaper, drifts further from D-006). CEO ratified the deeper move. The amendment to D-006 §Pillar 3 wording (70 → 75) is editorial follow-through, not a supersession.

Research artifact: ops/research/progression-system.md (Round 1, 2026-05-20).