yogsoth-ai Futures Calibration
Purpose: Aggregate probabilistic forecasts from multiple perspectives into calibrated predictions. Uses Real-Time Delphi continuous updating or prediction market scoring to produce well-calibrated probability estimates for future events.
When to use:
- Technology timeline estimation
- Market forecasts requiring multiple expert inputs
- Risk probability assessment
- Any question framed as "what is the probability that X by time T?"
Budget
| Parameter | Constraint |
|---|---|
| Rounds | 2–3 (continuous updating preferred) |
| Perspectives | ≥4 independent forecasters |
| Calibration target | Brier score improvement across rounds |
State Ledger
| Key | Type | Description |
|---|---|---|
| question | string | The forecasting question |
| time_horizon | string | When the event would resolve |
| perspectives | array | List of forecaster perspectives |
| rounds | array | History of probability estimates |
| current_aggregate | float | Current aggregated probability |
| calibration_metrics | object | Brier score, log score tracking |
Available Tactics
- iterative-convergence-round — Adapted for probability estimates
- threshold-calibration — Determine confidence bands
Available SOPs
- judgment-collection
- feedback-distribution
- consensus-measurement
- round-decision
- threshold-sweep
- consensus-synthesis
Execution Guidance
- Frame question with clear resolution criteria and time horizon
- Collect initial probability estimates with reasoning
- Share aggregate and reasoning (anonymized) each round
- Track whether estimates are converging or polarizing
- Report final calibrated estimate with confidence interval
Output Format
forecast_question: <question>
time_horizon: <date/period>
calibrated_probability: <float 0-1>
confidence_interval: [lower, upper]
rounds_completed: <int>
convergence_pattern: converging/stable/polarizing
key_considerations: [...]
Available Tactics
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use |
|---|---|
| iterative-convergence-round | Execute one full Delphi round — collect judgments, distribute anonymous feedback, measure consensus, decide whether to continue. |
| threshold-calibration | Systematically sweep consensus thresholds to observe which items achieve consensus at what level, producing a threshold-consensus curve. |
Available SOPs
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use |
|---|---|
| consensus-synthesis | Synthesize all rounds into a final consensus report documenting agreements, dissent, and process. |