What to list next
Every subject the archive holds, scored on the evidence behind it and ranked by the brief's own rule — official flow data above disclosed positions, above explicit trade posts, above chatter. Open any row to read the evidence that produced its score.
How the score works
A subject scores tier weight × its position within each feed, summed across feeds. Positions, never values: NT$166bn of turnover and 240 posts a day share no unit, but "top of its own feed" and "top of its own feed" are comparable.
Tier weights are 8 / 4 / 2 / 1 — it takes a great deal of chatter to outweigh one exchange file. Agreement between feeds adds up to 60%, capped so four chatter feeds nodding at each other cannot beat one official one. Momentum against the previous period tilts the result by at most a third, and never substitutes for evidence.
One feed gets one vote, at its strongest measure — otherwise a source publishing fourteen metrics per subject would outvote three sources publishing one.
Prediction-market contracts are excluded as candidates: a Kalshi contract is evidence about an asset, but its subject is a contract id and resolving that to an underlying is separate work. Participant categories and subreddits are excluded too — they are where demand is observed, not things that can be listed.
The number beside each region is how many feeds cover it. Coverage is uneven — Taiwan is the only market two feeds reach, so its names collect a corroboration bonus nowhere else can earn yet. Read the global list with that in mind, and the per-region lists to compare like with like.