Append-only decision log and the running in-game
value trajectory. The substrate for RQ1 (Decision Consistency).
DEPENDENCIES: DATA_Decisions, SM_Game
Pending
Critical Caveat — Placeholder Value Encoding
Researcher review required
STATE_Decisions accepts an externally-supplied
{pa, os, ei} weight triple per decision. The canonical
value-encoding (which choice expresses which value) is the job of
LOGIC_ValueEncoding (Module 12), which has not yet
been built and which requires inter-rater coding (Cohen’s
k > 0.70).
For the purpose of validating this module’s shape, the test
page below uses a placeholder encoding:
HELP \u2192 {pa: 1.0, os: 0.3, ei: 0.0};
REFUSE \u2192 {pa: 0.0, os: 0.7, ei: 1.0}.
This encoding is not the canonical research mapping.
All numbers and trajectory shapes shown here are illustrative
and will change once Module 12 lands. STATE_Decisions itself is
agnostic to the encoding it receives.
Automated Assertions
#
Assertion
Result
Detail
Interactive Workbench — Decision Trajectories
Two charts side-by-side: the per-phase mean of
PA / OS / EI weights, and the cumulative mean drifting
decision-by-decision. Use the controls below to drive a
simulation.
Per-phase mean (RQ1b phase-resolved)
Personal Advancement Organisational Stability Ethical Integrity