Help & FAQ

What makes a good forecast?

A good forecast is specific, measurable, verifiable, and time-bound. Instead of "AI will change the world," write something like "Write your thoughts here..." You'll see a live quality score as you write.

How is my reputation calculated?

• Quality (Q) rewards specific, measurable, time-bound statements.
• Difficulty (D) raises the stakes for bold, non-obvious calls.
• Calibration (C) measures whether your stated confidence matches your real hit rate.

Reputation grows as your resolved forecasts prove accurate and well-calibrated over time.

What is the difference between Public and Diary forecasts?

Public forecasts appear in the feed and build your public track record. Diary forecasts are private to you and never shown publicly, but still feed your private calibration analytics. Set a default in Settings.

How are forecasts resolved?

Each forecast has resolution criteria and a resolution date. When the outcome is known, the forecast is resolved against those criteria. If you believe a resolution is wrong, you can submit an appeal from the forecast page while it is still provisional.

What is a Forecast Battle?

A Battle compares your confidence against the AI and the crowd on the same question. After resolution, the closest estimate is highlighted — a quick way to see who read the situation best.

Still need help?

Check our About page for an overview, review the Terms and Privacy Policy, or contact support through your account.