Our method

A prediction-market price is evidence about expectations. It is not a fact, a poll, or a guarantee. Our process is designed around that distinction.

1. Select for public relevance

Polyruler looks for active markets with meaningful volume, clear settlement rules, and consequences beyond a narrow trading audience. Volume can draw our attention, but it does not determine coverage. We may skip a highly traded contract if it is repetitive, poorly framed, or lacks a useful public-interest question.

2. Freeze a dated snapshot

Every article records the probability, volume, liquidity, and time at which the data was collected. Market prices move continuously; a static article must never imply that its displayed number is live. We preserve the snapshot so readers can evaluate the commentary in the conditions in which it was written.

3. Read the contract before the story

The settlement terms define the market. We identify the precise trigger, deadline, resolution source, exclusions, fallbacks, and any ambiguity before interpreting price movement. A contract about an official announcement may differ materially from the broader news question suggested by its title.

4. Separate signal from mechanism

We consider whether the number may be shaped by thin liquidity, independent binary markets that do not sum neatly to 100%, a small or concentrated participant base, hedging demand, or stale orders. Large volume strengthens the case for attention but does not eliminate these problems.

5. Add verifiable outside context

Primary sources are preferred: official calendars, statements, election authorities, corporate announcements, and the market’s own resolution rules. When credible reporting is necessary, the article should name and link it. Unverified social posts and unattributed claims are not sufficient.

6. Write without a trading instruction

Polyruler explains what the price appears to say, where that reading can fail, and which future developments would matter. We do not tell readers to buy or sell, frame odds as certain returns, or disguise affiliate promotion as analysis.

Corrections

Material factual errors are corrected in place with a dated note. Ordinary post-publication price movement is not an error and does not silently rewrite the original snapshot. Send correction requests to editor@polyruler.com.