Data-driven studies of how prediction markets actually move, built on the same production dataset that runs Crypto Briefing’s prediction-market desk. Every report is open: read it here or download the full PDF, no email required.
Over 56 days ending 25 June 2026, 476,000 story-market pairs from 618 news sources scored on one question: an hour after the story landed, had the price moved a cent or more? The four biggest financial wires sit under the 15.2% all-source rate and three specialist accounts sit above it. Hold the markets constant and about 15% of that gap is left, and the order breaks.
PDF · 757 KB · vera.cryptobriefing.com/research/attention-gap.pdf
Polymarket’s decision ladder and the CME fed funds futures strip, priced side by side over the run-up to the July 28-29 FOMC decision. Across 50 paired sessions the two moved together, rank correlation +0.76, while Polymarket sat about 2.4 basis points below. Neither venue is scored for accuracy.
PDF · 917 KB · vera.cryptobriefing.com/research/two-prices-fed.pdf
The IMF ship count for the Strait of Hormuz is the exact number a Polymarket ladder settles on. Across 43 tests the daily price showed no link to the daily count, at a sample size too small to rule a moderate one out.
PDF · 743 KB · vera.cryptobriefing.com/research/ships-and-the-line.pdf
All 584,210 Polymarket resolution rules read in one pass: 582,754 carry rule text, 1,456 carry none. 96.27% are settled by a machine reading a scoreboard, a feed or a dataset. Human judgement decides 0.92%.
PDF · 809 KB · vera.cryptobriefing.com/research/settlement-registry.pdf
One goal-sized repricing, a whole tree of linked World Cup markets, and a measurement of how far the co-movement actually travels before it dies out.
PDF · 794 KB · vera.cryptobriefing.com/research/propagation-ladder.pdf
1,250 oil and shipping state changes, of which 74 cleared confirmation. A measure of how long a change of state stays true, in counts and durations only.
PDF · 889 KB · vera.cryptobriefing.com/research/oil-window.pdf
The Strait of Hormuz crisis, reconstructed contract by contract: how prediction markets priced a live geopolitical shock, one deadline at a time.
PDF · 600 KB · vera.cryptobriefing.com/research/hormuz.pdf
Does news that clears Vera’s confirmation layer move markets more often than the news it dismisses? A census-scale test of the product’s core claim.
PDF · 644 KB · vera.cryptobriefing.com/research/confirmation-dividend.pdf
Five in six sharp repricings had no headline our pipeline could attach to them. What separated the ones we could source was where we were already looking, not the market itself.
PDF · 632 KB · vera.cryptobriefing.com/research/orphan-moves.pdf
90,633 measured reactions timed from the moment the news broke. The median move took about 80 minutes to reach its peak, and that figure is a floor.
PDF · 743 KB · vera.cryptobriefing.com/research/minutes-myth.pdf
567,033 news-to-market pairings sorted by subject, showing that news about some categories reprices a market three to five times more often than news about others.
PDF · 815 KB · vera.cryptobriefing.com/research/reflex-map.pdf
A taxonomy of how prediction-market moves actually behave, and why a large share fade or reverse instead of holding.
PDF · 813 KB · vera.cryptobriefing.com/research/one-in-three-moves-is-fake.pdf
A machine-learning study of 780,494 news-market pairs finds that a prediction market’s own price geometry, not the content of the news, ranks which markets are about to move.
PDF · 548 KB · vera.cryptobriefing.com/research/words-dont-matter.pdf
New Vera research when it publishes, and nothing else. The reports here are always free to read and download without signing up. This is only if you want a heads-up.
Methodology and provenance. Every Vera report is built on Vera’s production dataset, the same event and price pipeline that runs Crypto Briefing’s prediction-market desk. Figures are computed from the underlying market and news records, and each report states its own sample window and method in its footer.
For research and information only. Not investment advice. Prediction markets and financial markets involve significant risk. Vera is editorial market intelligence, not a trading signal, and nothing here is a recommendation to buy, sell, or hold any asset, contract, or position.