Where the wire moved first.
Every case below comes straight from our own append-only ledger, added automatically the night it qualifies: the wire flagged the name, and the public consensus — read through one sensor we cite, a prediction market’s price — went on to shift by 20+ points. No hand-picking — the ledger writes the page. As of 2026-07-06.
Attention leading a reprice is an association, not advice. We cite the crowd’s price as a sensor; nothing here is a directional or trading claim.
On our wire from 2026-04-04, spreading across 13 domains (Agriculture Food, Batteries Storage, Energy…). The consensus sensor — “Will the United States send warships through the Strait of Hormuz by June 30, 2026?” — then repriced 63% → 6% between 2026-05-23 and 2026-06-29.
On our wire from 2026-04-04, spreading across 45 domains (Africa, AI Policy, Arctic…). The consensus sensor — “Trump suspends U.S. entry for more countries by by June 30?” — then repriced 83% → 21% between 2026-05-22 and 2026-06-23.
On our wire from 2026-04-04, spreading across 37 domains (Africa, African Economy, Agriculture Food…). The consensus sensor — “Will USD reach 1.9M Iranian rials by May 31?” — then repriced 95% → 3% between 2026-05-03 and 2026-05-30.
On our wire from 2026-04-04, spreading across 18 domains (Africa, African Economy, Australia Economy…). The consensus sensor — “Will OPEC crude oil production be above 18 million barrels per day in May?” — then repriced 73% → 11% between 2026-04-30 and 2026-06-10.
On our wire from 2026-04-04, spreading across 13 domains (Anti Corruption, Autonomous Shipping, Crypto DEFI…). The consensus sensor — “Will Russia enter Ternuvate again by May 31?” — then repriced 70% → 4% between 2026-04-22 and 2026-05-30.
On our wire from 2026-04-04, spreading across 11 domains (AI Safety, Banking Fintech, Cognitive Investor…). The consensus sensor — “Will Anthropic have closed its next funding round by June 30, 2026?” — then repriced 21% → 100% between 2026-04-21 and 2026-05-28.
On our wire from 2026-04-04, spreading across 21 domains (3d Printing, Agriculture Food, China Trade…). The consensus sensor — “Will Israel strike 2 countries in April 2026?” — then repriced 11% → 94% between 2026-04-04 and 2026-04-29.
On our wire from 2026-04-04, spreading across 19 domains (Africa, Banking Fintech, CBDC…). The consensus sensor — “Will the Bitcoin Volatility Index hit 70 by April 30?” — then repriced 83% → 1% between 2026-04-04 and 2026-04-29.
On our wire from 2026-04-04, spreading across 54 domains (Africa, AI Policy, AI Safety…). The consensus sensor — “Will Putin visit China by May 31?” — then repriced 77% → 100% between 2026-04-04 and 2026-05-19.
The regression behind it.
One measured property, stated plainly: on questions our watchers flag, the public consensus goes on to reprice materially more often than on comparable unflagged questions. The comparison is stratified (price band × volume / liquidity) with bootstrap confidence intervals, recomputed nightly against the full market panel — the shock-excluded estimate is the conservative one.
Latest published estimate: on flagged questions the consensus repriced materially at 1.4–1.9× the rate of comparable unflagged ones (95% CI, stratified controls). The nightly recomputation publishes its current numbers here as soon as the next analysis run lands.
It says the consensus moves more often on flagged questions — not which way. Methodology: treatment = first flag date per question; controls matched on price band and volume/liquidity strata; shock-days excluded in the conservative estimate.