Research Standards
Methodology
Air Radar combines live market data, structured signal systems, and editor-reviewed narrative framing. Public coverage is grounded in measurable market inputs before analysis is added.
Market inputs
Articles and public research pages may reference live prices, signal state, breadth measures, asset-class rotation, and historical signal outcomes. Those inputs are refreshed continuously and re-evaluated as new market data arrives.
Signal evaluation
Signal performance reporting is based on stored historical snapshots instead of ad hoc current-price comparisons. Older invalid rows are quarantined or compacted so public accuracy pages reflect a cleaner resolved sample.
Drafting support
Drafting tools can assist with summarization, normalization, and structure. Final visible output is published through Air Radar’s editorial workflow, with timestamps and source references preserved on the article page when available.
What makes a page publishable
- It must explain a real market change, not just restate a headline.
- It must tell the reader what to validate next in price, breadth, or cross-asset confirmation.
- It must preserve visible source context, byline attribution, and update timestamps.
- It should stay useful after the first reading, especially on explainer and research pages.
AI and human review boundary
Air Radar uses AI to accelerate drafting, clustering, and research support, but the public site is not supposed to publish raw model output. Editorial review is where weak framing, generic language, or unsupported claims should be removed before a page becomes public.