How accurate is Saju? Calculation, interpretation, and evidence

Saju does not have one defensible accuracy percentage. A chart can be calculated consistently under stated calendar rules, an interpretation can feel meaningful, and a prediction can still be untested. Ask which of those meanings of “accurate” a reader is claiming.

This distinction lets you explore a tradition without confusing a correct symbol with a guaranteed outcome. It also makes a claim such as “90% accurate” easier to question: accurate at which task, compared with what, and measured on whose results?

Calculation can be checked against a stated method

The Academy of Korean Studies’ Saju entry describes the birth-year, month, day and hour pillars and their calendar conventions. This documents a traditional method; it does not validate its predictions.

To check a calculation, keep the birth inputs and settings together. If two tools disagree, compare their handling of the calendar, local time and day boundary before comparing the prose. Our calendar guide explains why boundary cases matter. Agreement shows that tools may be following the same rules, not that your future has been established.

What one personality study actually found

A 2015 study compared Four Pillars classifications with personality questionnaires in 148 adults. Jung and Yang’s original abstract reports several null findings alongside weak significant correlations, including correlations of 0.18–0.29 between five-element measures and personality scores. The authors interpreted composite results more positively, suggesting potential usefulness for personality classification.

That is a limited personality study, not a prospective test of whether readings predict marriage, income or specific future events. A correlation is not a percentage of readings that “came true.” This paper cannot supply a general Saju success rate.

Recognition is a different kind of result

Suppose a hypothetical reading says you prefer freedom at work. You recognize that you dislike being micromanaged. That may start a useful conversation, but it leaves several questions open: Would you have accepted the same description without your chart? Is the description specific enough to distinguish you from other people? What would count as a mismatch?

Try writing both an example and a counterexample from your life. “I like choosing my working hours, but I prefer clear instructions for unfamiliar tasks” is more informative than a simple yes. You can value the reflection without presenting it as a measurement of predictive power.

What to ask when someone promises accuracy

• What exactly will be predicted, and by what deadline?
• Is the result recorded before the event?
• Are misses counted alongside matches?
• Is there a comparison with ordinary expectations or another method?
• Does the claim apply to calculations, personality descriptions or future events?

If those answers are missing, the number is not something you can independently assess. Testimonials may describe a person’s experience, but they do not answer those measurement questions.

A useful way to approach a reading

Begin with what Saju cannot do, then ask the reader to separate the chart, traditional meaning and practical suggestion. You do not have to decide that every part is either worthless or certain. Keep what helps you think, question what does not fit, and use real-world information for real-world decisions.