A deterministic read of the live news corpus: which conflicts are escalating, which stories multiple independent outlets are carrying, and which topics are taking a larger share of coverage than they were. No model wrote these numbers — they are counts, shares and ratios computed from the corpus, reproducible from the same inputs.
Mentions of distress vocabulary across the last 24 hours of the corpus.
0–10 over a trailing 48 hours. Conflict and economic signal, weighted by how many independent outlets carry it — the breadth term is what stops one prolific outlet manufacturing an escalation. Each entry lists the stories behind its score.
113 conflict · 4 economic · 94 mentions across 62 outlets
35 conflict · 4 economic · 66 mentions across 52 outlets
42 conflict · 2 economic · 66 mentions across 47 outlets
75 conflict · 1 economic · 79 mentions across 55 outlets
106 conflict · 1 economic · 99 mentions across 67 outlets
21 conflict · 0 economic · 15 mentions across 14 outlets
20 conflict · 0 economic · 23 mentions across 16 outlets
19 conflict · 0 economic · 33 mentions across 26 outlets
12 conflict · 2 economic · 64 mentions across 43 outlets
15 conflict · 0 economic · 28 mentions across 17 outlets
15 conflict · 0 economic · 14 mentions across 6 outlets
14 conflict · 0 economic · 8 mentions across 8 outlets
12 conflict · 1 economic · 12 mentions across 8 outlets
12 conflict · 1 economic · 12 mentions across 6 outlets
13 conflict · 0 economic · 14 mentions across 13 outlets
Last 24 hours. Ranked by how many independent outlets carry a story, not how many stories exist — five outlets running one wire item is one story with one source, and counting it as five is how a national wire ends up presented as corroboration.
7 stories · leans: CL, C
4 stories · leans: C
3 stories · leans: C
3 stories · leans: C
3 stories · leans: C
3 stories · leans: C
3 stories · leans: C
3 stories · leans: C
2 stories · leans: C
2 stories · leans: C
2 stories · leans: C
2 stories · leans: C
Seven-day spike detection: a term's share of the last 24 hours against its share of the days before. Share rather than raw count, because the corpus is a rolling working set and raw counts make everything look like a spike.
Compared against 7 day(s) of our own full-corpus snapshots (16040 stories).
Computed from the Corvus Intel corpus of 3,500 stories at 2026-11-06T08:33:23.751Z. Deterministic. Computed from the Corvus corpus with no model call — every figure is reproducible from the same inputs. Read how the publication works in the methodology, or open the interactive console with per-desk and per-persona filters and CSV/JSON export.
Open the interactive Analysis Console →