Local Mechanics Research: Findings & Evidence

What confirmed means on these pages

An authored value is present in a shipped game table. A native trace follows selected original instructions. An emulation check executes original instructions with explicit synthetic inputs and substituted engine services. A player observation is a separate firsthand gameplay report. We label these distinctions because none should silently become proof of every online gameplay outcome.

The research is tied to Windows build 1.1.1.0. Original files, hashes, instruction ranges, models, validation cases and a continuation journal are retained privately. The public evidence manifest contains reviewed claims and provenance without distributing the executable, raw game packages or decompiled bodies. Follow the evidence links on each section for the exact reviewed record.

The weapon-power model matched 290 selected native comparisons. Threat and loot work has 360 selected checks; balance inheritance has 32. Fixed-trait generation has 112. Equipment-power assignment has 75 scenarios / 91 top-level native invocations. The Health work has 25 health scenarios plus two class-mode checks. These sets have different scopes and can overlap; do not add them into a count of gameplay tests.

Enemy, merchant and talisman additions

Dartback's SpikesFired event is configured to apply a 3-second vulnerability effect with MultiplyAdditive 3.5 on MeleeDamageResistanceMultiplier. This establishes a timed vulnerability in the authored behavior, not a guaranteed 3.5× final hit against every state of the boss.

Local player ability records corroborate the merchant upgrade schedules already listed in the handbook. All 24 talisman progressions store 18,480 XP for the first upgrade and 73,920 for the next. Native service cost escalation remains unresolved.

The site owner directly tested that collecting all eight secret talismans opens the Mosstrosity secret door. This confirms the collection requirement, without establishing whether the items must remain in inventory afterward or how the unlock is shared in multiplayer.

The public calculator and its remaining limits

Research Lab now has a conditional melee damage calculator for 46 of 58 melee weapons. It uses the recovered weapon bases, slot-specific Item Power, threat, difficulty, supported passive effects and explicit target multipliers. It includes automatic fixed traits and checks rarity/Storm Called capacity, duplicate effect families and template power requirements. Unresolved fixed traits and enchantments block the result. Build Lab can copy its melee weapon into this isolated scenario; the other equipment is listed as excluded and weapon Item Power must be entered separately.

The tool reports internal packet arithmetic and the separate display-scale sum, with every multiplier visible. Neither number establishes final Health loss, DPS, a single combined popup, or a proven maximum-hit build. Remaining work includes effect and enchantment lifecycles, empowerment consumption order, ranged and jump assembly, animation hit counts, bangle selection order, Overload, complete application callbacks, final immunity/shielding and display rounding. Selected timer and admission paths are now recovered, but their composed cases still require supplied state and scheduling assumptions. Online balance replacements have not been observed.

Exact named-item drop probabilities additionally need fully resolved encounter inheritance, eligible pools, ownership and recent-drop history. We will not turn partial source weights into authoritative farming odds.

Earlier notes that relied only on a curve ceiling of 8, a player critical baseline of 2, two ordinary Unique random effects, or missing raw status tags have been superseded by later traces. The current pages include those corrections.

Four fixed buffs can now be explicitly selected as already active: Critical Strength, Critical Precision, Strengthened and Grindstone. Duplicate named statuses count once. This does not automatically include equipped artifact activations or solve timing and legal-loadout constraints.

Sources & data coverage

Reviewed local game tables and selected native instructions from Windows build 1.1.1.0. Original-instruction tests use synthetic actors and engine services. They are not gameplay measurements; online overrides have not been observed.

Reviewed evidence manifest and scalar data ↗