Skip to content
WalkRPG Wiki

Combat resolution canon

D-010: Combat resolution canon

Ratifies three open questions from phase 11b-1 (wiki/src/content/docs/combat/formulas.md, commit 43b8efc) as canon. Math layer (stat dictionary, damage/defense/hit equations, energy cost shape, encounter shape) remains under mechanics-designer B-level authority; only the three philosophical choices below are D-level.

  1. RNG model = seeded PRNG. Combat resolution is deterministic given a seed. Rationale: D-007 anti-cheat requires server-side reconciliation, which requires replayability. PRNG choice (xorshift / PCG / similar) is delegated to phase 11b-2 with a “pick one, freeze it” constraint — the chosen algorithm itself becomes canon at 11b-2 ship.
  2. Turn order = strict Walker-first alternation. One attack per turn. Multi-attacks attach as modifiers, not as turn-order variants. Rejected: speed-based initiative, simultaneous resolution — both expand the stat dictionary and test surface without earning their complexity at phase 11.
  3. Defeat = non-punitive. At hp=0, hp clamps to 1; no XP/point/rep loss, no streak penalty. Rejected: streak-loss (conflicts with D-007 invariant — streak decays only on missed attested walks, not on combat outcomes) and rep penalty (a quest-layer system, not a combat-layer one).

Reasoning: Anchors phase 11b-2 (combat simulator) on a stable canonical foundation, removes the ambiguity that would force the simulator to re-derive philosophy from formulas, and prevents drift across 11e-1 (combat UI) which must surface these mechanics to the player.