SAMPLEOne sample bridge. The feed is empty in production until a bridge survives review.
07 / DISCOVERY FEED
Bridges between arenas, after they survive attack.
A bridge is a claim that two graded records imply something neither implies alone. It is a hypothesis, not a forecast — it enters the feed only after a red team has tried to break it and failed, and it never enters the scorecard.
BRIDGE-001 SURVIVED CHALLENGE
PROMOTED 2026-07-22T14:05:00Z
Metro in-migration breaks show up in bank credit stress two quarters before they show up in prices.
Our largest housing misses cluster on markets where the financial-stability arena had already moved its regional credit-stress reading. The bridge claims the ordering is stable, not that either arena is more accurate.
CLAIM A · HOUSING / CANADACalgary prices came in 3.48pp below a locked forecast that carried the in-migration surge forward.hca-2025-06-01/calgary →
CLAIM B · FINANCIAL STABILITYRegional credit stress was called ELEVATED two quarters earlier and graded as a pass on resolution.fst-2025-04-01 →
SHARED EVIDENCE
Co-observed markets3 of 8 · Calgary, Halifax, Seattle
Lead time, median2 quarters · range 1–3
Shared primary sourcenone — independent series
Graded vintages behind it5 · all forward, none backtest
RED TEAM
4 challenges filed · 4 answered · 0 open · last attempt 2026-07-21The strongest surviving objection: the ordering rests on 3 co-observed markets. It is recorded on the bridge and will be re-tested at n=6.
KNOWLEDGE GRAPH38 hypotheses standing. One promoted.Every graded vintage is a node. Bridges are edges that survived attack. The rest of the graph stays off the ledger until it earns a place on it.
NO PROMOTED DISCOVERIES YET
Bridges appear here only after surviving adversarial review. Thirty-eight hypotheses are standing; none has cleared the red team.
The One Lab · Toronto · Every number on this site resolves to a primary source or a graded model output. Where it can't, we say so.DISCOVERY FEED · SAMPLE