How GOATYD works
One score. Clear context.
GOATYD combines eligible, reviewed evidence into one canonical GOAT Score. This page explains what the score means, how confidence and publication work, and what remains separate.
- Current implemented version
- placeholder_bundle_capped_v1
- Effective from
- August 3, 2026
One canonical GOAT Score
Every published GOAT uses the same platform-wide 0–100 scale. The score is an evidence-based model output—not a raw vote total, popularity counter or reward balance.
Arena position is different
An Arena asks a specific question and orders only its eligible participants. A GOAT can therefore have a strong canonical score but a different position in a particular Arena. Arenas do not create a second score.
Evidence and protection
Evidence must earn influence
The scoring service accepts only eligible structured signals. Exact weights, caps and abuse thresholds remain protected so they cannot be gamed.
- 01
Structured Arena participation—such as eligible Battle and Bracket outcomes—can enter an evidence bundle under the active methodology.
- 02
Published identity, eligibility and source provenance provide the governed context needed to interpret evidence.
- 03
Repeated or correlated outcomes collapse and are capped by evidence bundle. Suspicious signals can be quarantined, invalidated or restored after review.
Confidence and publication
A score is not published merely because it exists
Collect eligible evidence
Apply integrity controls
Evaluate confidence
Pass publication gate
Confidence describes the strength, diversity, coverage, stability and integrity of the evidence—not how high the score is. Provisional or insufficient outputs remain clearly separated from official publication. New official snapshots add history; they do not overwrite earlier snapshots.
Rewards never multiply influence
Points, XP, Levels, badges and streaks recognize participation. They never multiply voting power. Personal Top 10, Battle and Bracket results express an individual view and never replace an official Arena ranking.
A Canonical GOAT Vote can currently be recorded from a GOAT profile, but it is excluded from production scoring until a later named methodology version explicitly activates bounded influence.
Versioning and governance
Changes leave a visible boundary
Methodology changes require approval, a new version identifier and an effective date. Score history records version transitions. GOATYD may refine eligibility, confidence and integrity protections without exposing operational thresholds that would weaken the system.