ACCURACY POLICY
How source data and analysis are separated
HOI4 War Planner is locked to a bundled vanilla Hearts of Iron IV 1.19.2 baseline. It combines retained game-file data with clearly labeled analytical modeling and does not claim executable parity.
Evidence classes
Game-file exact covers values and relationships directly retained from supplied 1.19.2 files. Executable inferred covers behavior constrained by evidence but not fully provable from data files alone. Planner analytical covers scenario inputs, generated comparisons, grades and other abstractions created by the planner.
Controlled comparisons
The planner is designed to compare two sets of assumptions under the same conditions. Reproducible seeded simulations and uncertainty bands help show whether a conclusion is robust or fragile instead of presenting a single result as certainty.
Known boundaries
Some in-game behavior depends on executable-only logic or interactions that are not fully exposed by the retained data. Those areas remain outside exact-parity claims. When a rule cannot be certified, the interface should label it rather than silently presenting an approximation as game truth.
Gauntlet methodology
Division Gauntlet generates plausible opponent designs to stress-test a template. Quick mode uses roughly 500 opponents. Full mode uses 10,000 deterministic designs across eight terrain types in both attack and defense roles, creating 160,000 matchup checks.
Opponent generation, grades, efficiency scores and percentile rankings are planner analytical. They are useful for finding weak matchups and tradeoffs, not for claiming that the game itself assigns those grades.
Theorycrafting policy
The planner is intentionally theorycraft-first. Prerequisites are generally informational unless source data defines a structural compatibility rule. This keeps the tool focused on comparing designs instead of recreating the full progression system.