Score Integrity
No reputation system can make manipulation impossible. HYFY uses several overlapping safeguards to make rating spam, self-promotion, and coordinated attacks harder while keeping ordinary ratings simple.
Protection Layers
| Protection | What it does |
|---|---|
| Mutual connection for people | Prevents strangers from leaving drive-by ratings on a person |
| No self or same-account rating | Stops direct score inflation across profiles someone controls |
| One current opinion per trait | Prevents repeated versions from stacking as extra votes |
| Update limits | Slows rapid switching and rating spam |
| Bounded reviewer context | Keeps differences in reviewer influence small |
| Time decay | Prevents old recognition from controlling a score forever |
| Independent-rater support | Makes one person's opinion less decisive than broader support |
| Server-side enforcement | Stops the app from bypassing ownership and eligibility checks |
Reviewer Context Without Popularity Voting
HYFY does not silently turn popularity or conformity into extra voting power. Reviewer influence is not increased because someone:
- has more followers,
- already has a high Trust Score,
- submits many ratings, or
- agrees with the majority.
The system may give a small and capped confidence adjustment for signals such as account verification and account age. Every eligible reviewer still counts, and no reviewer receives unlimited influence.
Why Early Scores Are Cautious
A new profile can receive a numeric score from its first eligible rating, but one opinion is not treated as if it represents a whole community. As more different people contribute, the score can move closer to the actual pattern of their ratings.
This helps reduce the effect of a single unusually generous rating and makes it harder for a small cluster of accounts to create an instant perfect score.
Rate Changes Are Controlled
Ratings can be corrected during an update window, but they cannot be changed continuously. A cooldown applies before another rating for the same profile and trait can be submitted.
These controls protect the score without making an honest correction impossible.
Checks Are Enforced on the Server
The app does not get to decide whether a rating is valid. HYFY's authenticated rating service verifies that the reviewer controls the profile they are using, blocks ratings between profiles controlled by the same account, and applies the relationship, profile-type, trait, update, and cooldown rules before accepting the rating.
The ratings table cannot be written to directly by ordinary signed-in users. HYFY also serializes competing submissions for the same reviewer, profile, and trait so that two simultaneous requests cannot become duplicate votes.
These protections make client-side tampering harder, although they cannot prove that every human opinion is honest.
What Still Requires Human Judgment
- A mutual follow shows a two-way connection, not how well two people know each other.
- Account verification raises confidence in account integrity; it does not guarantee good behavior.
- A Trust Score summarizes recognition and reviews; it is not a background check or safety certificate.
Trust Score is a useful signal, not a promise that a profile is safe or that every rating is truthful. Report abuse and use your own judgment, especially for offline meetings, money, or professional decisions.