Sample data
FED DEC CUT68¢▲ 27SHUTDOWN31¢▲ 12BTC >150K22¢▼ 9SPX >700054¢▼ 6KC TITLE14¢▲ 3RECESSION18¢▼ 2AI MODEL73¢▲ 5CPI <3%47¢▲ 8FED DEC CUT68¢▲ 27SHUTDOWN31¢▲ 12BTC >150K22¢▼ 9SPX >700054¢▼ 6KC TITLE14¢▲ 3RECESSION18¢▼ 2AI MODEL73¢▲ 5CPI <3%47¢▲ 8
PredictionStars Get the daily brief

How it works

Where our numbers come from, how we pick stories and how we keep ourselves honest.

Where the data comes from

Our coverage is built on public market data from major prediction market platforms: prices, 24-hour changes, trading volume and each market’s resolution rules. During our launch period, prices shown on the site are sample data, clearly labeled, so you can see the format before live coverage starts.

How we choose markets

  1. Volume floor. A market must have meaningful trading volume before we feature it. Small markets are easy to push around and can mislead.
  2. Biggest moves. We rank eligible markets by how much the Yes price changed over 24 hours.
  3. Newsworthiness. An editor picks the moves that matter and that we can explain with sourced reporting.

How a story gets made

Software pulls the data and drafts a short explanation of what changed. A separate review step checks every number against the source and flags anything unsupported. Then a human editor approves, edits or kills the story. Stale stories expire instead of publishing late.

Reading our market cards

Each card shows the question, the Yes price (which works as an implied probability), the 24-hour change, a 7-day trend line and volume. Read How to read a prediction market in ten seconds for a quick guide.

The scorecard

Once live coverage begins, we’ll log every market we write about and publish how it resolved, so readers can judge how well market prices predicted outcomes over time.

Corrections

If we get a number or fact wrong, we fix it and note the correction at the bottom of the story.