Editorial Standards

The Truth Tracking Learning Center explains how to interpret public predictions and prediction records. Its purpose is educational: to clarify concepts, surface limitations, and help readers evaluate evidence. It is not investment, political, or sports-betting advice.

What these articles cover

The articles discuss concepts such as accuracy, calibration, sample size, verification, scoring, and the design of prediction-tracking websites. They are separate from the underlying prediction records and dashboards. A guide may explain a method without claiming that every implementation of that method is equally reliable.

Sources and review

We prefer primary and first-party sources when describing a platform, scoring rule, or methodology. These may include official documentation, published methodology pages, original prediction records, and direct examples that readers can inspect. When a comparison describes a third-party product or service, the article identifies the date on which those claims were reviewed. Features, policies, and interfaces can change after that date.

Articles distinguish, as clearly as practical, between a factual description, an interpretation of evidence, and a limitation or uncertainty. We avoid presenting a single score, leaderboard, or methodology as a complete measure of forecasting quality.

Independence and disclosures

Truth Tracking publishes these guides on its own site, so readers should treat them as first-party editorial material rather than an independent certification of Truth Tracking. Comparisons are intended to describe differences in scope and method, not to declare a universal winner. A platform may be a better fit for one research question and a worse fit for another.

Corrections and updates

If an article contains a factual error, a broken link, or an outdated description, please contact [Truth Tracking](mailto:contact@truthtracking.org) with the relevant page and a proposed correction. We will review substantive reports and update the article when warranted. The published and updated dates identify the current editorial version; comparison pages also show the date through which their product claims were reviewed.

Related methodology

The Learning Center is about evaluation concepts. For details about the site’s prediction records, see the Stock-EN methodology and Politics methodology. Those pages describe the corresponding catalog and verification workflows rather than serving as general advice about forecasting.