Finishing the course does not make someone accurate. Practice does, but only if it is recorded.

A case library turns readings into evidence. It lets you see where your selection, strength, timing, wording, or confidence was right or wrong.

The library is not a portfolio. It is a calibration tool. Its value depends on keeping failures, partial hits, delayed outcomes, and ambiguous charts alongside the clean wins.

Write the prediction before the outcome is known. If the result is already known, the case may still teach mechanics, but it should not be counted as evidence for your forecasting accuracy.

Question Exact wording

Record what was actually asked, not what you later think it meant.

Context Domain and stakes

Career, health, legal, wealth, relationship, travel, documents, weather, or broad redirect.

Chart Cast data

Date, hexagrams, moving lines, target, support, obstacle, and timing state.

Prediction Judgment and timing

What you said, how confident you were, and what date or window you named.

Outcome What happened

Record the actual result, including dates, partial outcomes, and whether the verification window has passed.

Review What changed?

Tag selection errors, strength errors, missed dynamics, timing errors, and wording issues.

Review categories

Selection error Wrong target

The rest of the reading cannot recover if the answer-bearing line was wrong.

Mechanics error Missed state or interaction

Void, break, bind, clash, hidden movement, or transformation was misread.

Timing error Wrong release event

The judgment may be right while the timing candidate is wrong.

Communication error Bad wording

The analysis may be right but delivered with too much certainty or too little use.

Mixed result Right direction, wrong timing

Keep prognosis and timing separate so one success does not hide the other mistake.

Routing error Wrong method or question

The cast may be technically clean but poorly matched to a broad-life, destiny, or unsafe question.

Verification tiers

Tier 1 Direct outcome

You know the event, date, and result from the person or record involved.

Tier 2 Reliable report

The result is reported by someone close enough to the matter to be useful, but not direct.

Tier 3 Unverified note

The case is worth keeping as practice, but should not be used as evidence for a rule.

Tier 3 cases are not bad cases. They are simply not proof. Keep them in the library, but do not use them to promote a rule, confidence claim, or timing shortcut.

Trap

Do not keep only successful cases. A library without misses is not a learning tool.

Continue to Comparing Methods