Concept demo · not a commercial product
Track record
Every forecast is kept. Once a question's outcome is known it is marked YES or NO, and the forecast is scored here, good or bad.
No resolved questions yet
Scores appear as soon as the first questions are decided. Only forecasts made before the outcome was known are counted.
How scoring works
- Each question is scored on its first forecast, so running a question again can't improve the record.
- Only questions whose outcome was unknown when they were forecast count. Asking about something that has already happened would let the AI use what it already knows.
- The Brier score is the average squared gap between the forecast and what happened (1 for YES, 0 for NO). 0 is perfect; always saying 50% scores 0.25.
- Questions that turn out to be ambiguous are marked "couldn't be decided" and left out of the scores.