Methodology

PrediBeacon is an independent information service. It does not execute trades, accept deposits, hold customer funds or guarantee outcomes.

Market data

PrediBeacon retrieves public market information from supported venues, currently Kalshi and Polymarket. It normalizes titles, probabilities, volume, closing times and observation timestamps into a common read model.

Freshness

Every view distinguishes fresh, delayed, unavailable and inconsistent timestamps. Prediction-market prices can change after collection and should never be treated as guaranteed or continuously live.

Ranking

Discovery rankings use transparent signals such as probability movement, volume and trend score. Ranking does not mean endorsement, recommendation, equivalence or expected profit.

Contract comparison

Similar titles are not sufficient to establish that two contracts are equivalent. Question wording, deadline, resolution source and complete rules must align. When evidence is incomplete, PrediBeacon fails closed.

Evidence

Primary venue pages, official documents and allowlisted journalism may be attached to a market. Automated matching uses conservative relevance and freshness requirements. A missing match is preferred to an unsupported claim.

AI-assisted content

AI may prepare drafts only from persisted evidence snapshots. Citations are locked, output is reviewed by a human, and approval is separate from publication. Articles can be rolled back without deleting their audit history.

Commercial independence

Editorial ordering must remain separate from any future paid placement or affiliate relationship. Any compensated link or sponsorship will be clearly disclosed and limited by country and eligibility.