The briefing
The newest chains, each in one line — then the whole argument if you want it.
World Pulse
Who moves the board, where it is contested, and what carries the shock.
Cause, then effect — with the receipts
Not a feed of headlines. Each analysis traces one chain of consequence, and every claim carries its source. The Trust Index is counted from those citations, never assigned.
The method →Balance of Power
Military, economic, diplomatic, technological, energy and soft power — scored, not asserted.
Ask the archive
1226 passages. Every result says whether it is sourced fact, editorial judgment, or a projection.
Historical echo
Seven years, one line each — the moments the current chains are still answering.
Straight answers about the method
Where the numbers come from, what they measure, and what happens when we get one wrong.
Where do these numbers come from?
Every figure is either something a named source published, or a call we made and labelled as ours. Citations are opened and read — the rule is fetch the source, never cite a search result about it.
How we split the two →What does the Trust Index measure?
How far independent sources agree about the facts: supports ÷ (supports + disputes), counted from the citation list. Nobody types it, and no citations means no score. A low number isn't a broken one — it means the record is genuinely divided.
How it's counted →Why do some answers say “likely”?
Three coarse labels: confirmed, likely, uncertain. Most causal claims are “likely” — our reading, which a competent analyst could contest. “71% confident” would imply arithmetic we haven't done.
What each label means →What happens when you get it wrong?
The correction is logged on the analysis itself, not applied silently. A changed number with no notice is indistinguishable from never having been wrong.
The corrections policy →