Domains · Friendship

Friendship: how it reads and how it is computed

Friendship reads the thickness of peer support, how cooperation works, and how much social contact costs you.

The weights this domain actually uses (real engine output)

Ten-god vector

Rob Wealthweight 0.7
Companionweight 0.6
Indirect Wealthweight 0.3

Five-phase fingerprint

Earthweight 0.6
Woodweight 0.4

The two sets are combined to compute this domain: ten-god total and five-phase total. The same constants are used by the app, so you can check any result against them.

How the period changes it

Switching period (block, today, week, month, quarter, year, decade) re-runs the calculation with the pillars in force for that window, so the same question can read differently in different periods. The three layers are kept apart as well: base (birth data only), effort (your answers and logs), and fused (the two combined by formula).

Where the weak spot is

These are behaviour-level suggestions: a score describes your position, and action is what changes it. We make no promise that doing X will change your fortune.

FAQ

Does a low score mean few friends?

No. It reads the quality of support and rapport, not the number of people.

Can I raise it?

The base layer cannot be changed; the effort layer can. Reaching out, speaking plainly and joining long-term circles are the levers.

How is the Friendship score computed?

Two routes combined: a five-phase fingerprint (Earth 0.6, Wood 0.4) and a ten-god vector (Rob Wealth 0.7, Companion 0.6, Indirect Wealth 0.3), fused by domain weight. Switching period re-runs it with the pillars in force.

Other domains