Models

Evidence Quality and Source Selection

Learn how Models weigh source authority, freshness, relevance, duplicates, and disagreement without treating any source as automatically true.

Definition

Evidence quality describes how well research sources match a claim in authority, relevance, freshness, independence, and provenance.

Source types

Evidence may come from official or primary sources, statistical databases, specialist sources, reputable reporting, or secondary aggregation. The appropriate preference depends on the category and claim.

How evidence is selected

Models consider authority, relevance to the specific claim, freshness, primary versus secondary status, publication timing, and whether multiple items repeat the same upstream report. Duplicate reporting should not outweigh independent confirmation.

Authority is not automatic proof

Source authority helps select and organize evidence; it does not make a claim automatically true. A strong source can be incomplete, stale, or addressing a different question, so the evidence still needs to match the resolution criteria.

Research limits

Research depth and Research & compute level keep evidence gathering bounded. More articles or more time can add context, but neither guarantees a better forecast. Models may abstain when the evidence set is too weak or stale.

NEXT STEP

Learn about counterevidence

See how Models handle conflicting evidence.

Learn about counterevidence →
VIDEO SUPPORT READY

A future walkthrough can sit alongside this written guide. No video is currently published.