Work with the research

Newsrooms

Quote and cite any study with a link to its page. Each has a press note with the citation line filled in, a chart as a file with its source line, and its data as CSV and JSON.

• A figure checked, or a cut we have not published, for a piece with a date: write to the desk with the question and the day it runs.
• A scorecard seen before it is scored: the checkpoints are fixed in advance and the result is published on the stated day.
• An embed for a page or a newsletter: source-timed prices in one frame, no script, no cookie.

For example: The 2026 midterm scorecard, frozen before the vote.

Researchers

Data we collect from public sources is under Creative Commons Attribution 4.0 with its method note. Add .md to a study address for the full text, /data.json for its datasets, or read the research through the MCP server.

• A published test rerun on a venue it has not been run on, with the definitions checked against yours before release.
• Your help named only with your permission; "definitions reviewed by" means the definitions, not the findings.
• Order books and trades for a stated window, with provenance, where the licence allows.

For example: The efficiency series: two PredictIt tests rerun on Polymarket, monthly.

Institutions

Measurement to order: one question answered from what we observe, with the method stated. Tell us the question and the date it is needed.

• A recurring measurement on a schedule, delivered before or with publication.
• A method review: your definitions and figures checked against public records and our observations.
• A data extract: a cut of what we observe, with provenance. Licensed data stays within its licence.
• If paid work becomes a public study, the page says who paid.

For example: The Fed record: every 2026 decision scored against prices locked in advance.

Venues, exchanges and data providers

We welcome work with prediction market venues, crypto exchanges, protocols and data providers under written licences, and publish a summary of any we agree. A licence sets what we may show, never what we conclude.

• A claim you make about your market, measured from the outside and published with your name on it.
• Your data shown within its licence, with the licence named on the page.

For example: How Polymarket settles: 110,576 markets, measured from the outside.

A country brief

One country's prediction-market record, read from the public one. A brief answers a single question about a single jurisdiction and the venues you name: what its authority has done, what each venue's own policy says, and the access observations available in the published record — kept as three separate facts, each with its source and a stated cutoff.

• Scope: one country, the venues you name, one question, and the cutoff the brief is accurate to.
• What it is not: not legal advice, and not a live status — a market we did not test is marked unknown, never read as permitted.
• What you get back: the authority actions, the venue policies, the observed access, the cited sources, and the brief's limitations.

A worked example: the Germany country brief on the Europe record, with its JSON download. To ask for one, say what you need below.

How to cite

Vultax Research, <study title>, <revision date>, <study URL>

Each study’s press note carries the line filled in. The files for every study, the embed and the licence are on the data page. Reuse follows our terms and, where a licence applies, the source provider's rights.