Operators
Forecasts overview
Scoping decisions, skeptical review, and closed loops.
Forecasts are predictions you can defend: scoped decisions, traced evidence, recorded reviews, measured outcomes.
Your job
- Scope the decision, not the method: target, time range, audience. Most bad forecasts are scoping failures, not model failures.
- Dry-run skeptically: drivers plausible? inputs complete? spread bettable or only ranking?
- Walk the evidence chain for every figure you will rely on.
- Review in writing: approve, correct (highest-value feedback), or reject.
- Close the loop: record the decision, watch the outcome, let it teach the next round. Accuracy is measured, not assumed.
First time? Get your first forecast. Review mechanics: the flagship loop.
Image: forecast review queue with approve, correct, and reject actions and recorded outcomes.