Practice & cases
Why timing judgments keep their uncertainty, scoping the Useful God by question type, and recording cases to catch errors.
4 articles
Hexagram texts and Najia Liuyao use different evidence
Distinguish canonical line names from casting values, then compare Jie’s First Nine passage with its Water-palace assignments and a local return-generation relation.
Read first: Najia, Six Relatives, and Six Spirits: three chart layers · Moving lines and changed counterparts: auditing Xu to Song
LY-013Define a Liuyao question before selecting its target
Compare document and payment targets in doubled Kan, distinguish proxy subjects and role labels, and keep alternatives within a common observation scope.
Read first: Choosing the useful god: define the subject and retain candidates
LY-014Recording Liuyao cases: separate inputs, claims, and outcomes
Rebuild Jie changing to Kan from a six-value record, detect a hidden transcription error, and keep repeat castings, revisions, and missing follow-up distinct.
Read first: Moving lines and changed counterparts: auditing Xu to Song
LY-012Timing judgments need uncertainty: candidates are not guarantees
Compute still-Rat and moving-Rat candidates within a fixed twelve-day window, then separate calendar arithmetic, event definitions, and observed outcomes.
Read first: Moving lines and changed counterparts: auditing Xu to Song · Cycle void versus month break: two independent calendar tests