Ask Kalicube

Indexing Annotation Confidence Multipliers form Level 4 of Jason Barnard's Indexing Annotation Hierarchy. These seven annotations scale a chunk's confidence score up or down within its assigned competition pool: (1) Verifiability - can the claims be fact-checked? (2) Provenance - first-party claim, third-party validation, or aggregated consensus? (3) Corroboration Count - how many independent sources say the same thing? (4) Specificity - precise and quantified versus vague and general? (5) Evidence Type - research citation, data, expert quote, or anecdote? (6) Controversy Score - widely agreed or disputed? (7) Outlier Flag - matches consensus or contradicts it? Unlike gatekeepers that eliminate or filters that route, Confidence Multipliers SCALE. They determine where within the competition pool a chunk ranks. High multipliers push chunks toward selection; low multipliers push chunks toward obscurity.

Confidence Multipliers are the ranking amplifiers. You've passed the gates, you've been defined, you've been routed to the right pool. Now: where do you RANK in that pool? Third-party validation beats first-party claims. Specific numbers beat vague statements. Research citations beat anecdotes. Multiple sources agreeing beats single sources. These don't eliminate - they amplify or diminish. They're how identical content from different sources gets differentiated.

The shift from "build links" to "build credibility signals" reflects the reality of Confidence Multipliers. Links were a proxy for third-party validation. But modern annotation systems measure validation directly: Who said this? Can it be verified? Do others agree? Is it specific? What evidence supports it? These multipliers explain why authoritative sources consistently outrank equally accurate but less validated content - even within the same competition pool.