Skip to content

Trust-Tier Badging

Classification badge — instantiates Reputational Signal Governance

Bins subjects into a few coarse trust tiers shown as a badge, and attaches concrete treatment to each tier.

Trust-Tier Badging is the interface where standing meets treatment. Instead of exposing a raw score, it sorts subjects into a handful of named tiers — Unverified, Verified, Trusted, and so on — shown as a badge others read at a glance, and it attaches concrete consequences to each tier: what it unlocks, how much scrutiny it earns, what access it grants. Its defining trade is coarse legibility for precision: it throws away the score's fine gradations in exchange for an instantly readable cue and an explicit "here's what this means for you." That coupling of tier to treatment is what separates it from the index that merely produces a number.

Example

An online community runs the trust-level pattern popularized by forum software such as Discourse. A brand-new member sits at the lowest level: rate-limited, unable to post links, every action lightly checked. As they participate and their flags prove reliable, the system promotes them through tiers that progressively lift those frictions, until the top tier can recategorize and edit others' posts. A moderator triaging a report doesn't reconstruct each member's full history — the tier badge tells them how much latitude to extend. The badge compresses a standing into a decision the reader can make in a second.

How it works

What distinguishes it is bucketing plus attached consequences:

  • Define ordered tiers. A small, ordered set of levels, each with explicit entry and exit criteria.
  • Assign and display. Map a subject's underlying standing onto a tier and show it as a badge — the legible, reader-facing marker.
  • Couple to treatment. Bind each tier to a treatment policy — rate limits, scrutiny level, access, priority — so the badge is actionable rather than decorative.

Tuning parameters

  • Number of tiers — few tiers are legible but blunt; many tiers are finer but confusing and gameable right at the boundaries.
  • Promotion and demotion criteria — how easily a subject rises or falls, and how much hysteresis. Sticky tiers feel stable but lag reality; twitchy tiers thrash on noise.
  • Coupling strength — how sharply treatment differs across tiers. Strong coupling motivates but raises the stakes of a misclassification and the harm of exclusion.
  • Badge salience — how prominently the tier is shown. A salient badge drives behavior and can become a stigma for the unbadged.
  • Reversibility — whether a demoted subject can climb back, and by what path (leave the rehabilitation specifics to the reentry rule).

When it helps, and when it misleads

Its strength is that it compresses a standing into an instantly usable cue and makes the "so what" explicit — letting a busy actor triage trust in a glance and letting a platform scale gatekeeping without hand-reviewing every case.

It misleads because coarseness both hides and cliffs. Two subjects in the same tier can be far apart, and a subject just below a threshold is treated worse than one just above on a hair's difference. Worst, a badge gets read as a guarantee rather than a probability — the halo effect[1] leads people to over-trust a "Verified" mark and drop their own diligence — while low or absent badges harden into a stigma that traps subjects in a low caste. The classic misuse is handing out or withholding tiers to reward insiders and punish disfavored subjects regardless of conduct, coupling real treatment to a politically awarded badge. The discipline is to publish tier criteria, keep the standing-to-tier mapping mechanical and auditable, and pair strong coupling with a working appeal path.

How it implements the components

  • tiered_trust_badge — the small ordered set of trust tiers and the badge that displays a subject's current tier.
  • consequence_coupling_rule — binds each tier to concrete treatment (limits, scrutiny, access, priority), which is what makes the badge bite.

It does NOT compute the underlying standing (that's Reputation Score or Standing Index), capture or verify the traces beneath it (Rating and Review System, Verified Transaction History), or run rehabilitation and reentry for demoted subjects (Moderation Record with Reentry).

Notes

Badging is where reputation actually bites, which makes it the place fairness matters most. The coarser the tiers and the heavier the treatment attached to them, the more the whole thing depends on published criteria and a real appeal path — because a misclassification here is not a wrong number, it is a door that stays shut.

References

[1] The halo effect — a single positive mark (here, a trust badge) bleeds into an unwarranted global judgment, so a narrow "verified identity" signal gets read as broad trustworthiness. It is why a badge should state precisely what it certifies and why over-coupling treatment to it is dangerous. ↩