Methodology

The index is designed as a live sentiment signal, not a perfect poll. The goal is to raise manipulation cost and make uncertainty visible.

Category scores

Love, disapproval, safety, overrated, underrated, welcoming, and divisive rankings each use category-specific vote types while sharing integrity and confidence controls.

Adjusted score

Votes are weighted by intensity, recency, trust signals, geographic diversity, and anti-brigading dampening. Raw votes are stored separately from score snapshots.

Confidence

Confidence reflects vote volume, geographic diversity, duplicate rate, bot risk, account trust, and abnormal spike detection.

Integrity layer

The MVP uses server-side validation, hashed IP/device signals, rate limiting, Turnstile when configured, duplicate-risk scoring, and admin anomaly review.

About this index

This index measures public disapproval of countries, territories, governments, and state actions. It must not be used to promote hatred, harassment, violence, or discrimination against citizens, ethnic groups, religions, immigrants, or nationalities.