Uses

One question, every field: which forecast should I trust right now?

  1. 01

    Forecast issued

  2. 02

    Memory of past error

  3. 03

    Reliability in context

  4. 04

    Choose · combine · wait · abstain

01

Central banks & macro

in the current corpus

Input
Professional forecaster panels (ECB SPF: inflation, growth, unemployment)
Fragility signal
which survey panellists drift before a turning point
Decision it changes
Weight, discount or wait on a panel before a policy meeting
02

Markets & crypto

in the current corpus

Input
Resolved prediction-market questions and digital-asset return replays
Fragility signal
when the market price is least trustworthy
Decision it changes
Size a position, hedge, or abstain when the crowd is fragile
03

Earth & climate risk

in the current corpus

Input
Weather, air-quality and space-weather forecasts against measured outcomes
Fragility signal
which weather models share the same blind spot
Decision it changes
Trigger early warning or hold, with a stated reliability
04

Critical infrastructure

research direction

Input
Operator demand and capacity forecasts with later measurements
Fragility signal
load and demand forecasts that fail under stress
Decision it changes
Reserve capacity when the forecast mix is least reliable
05

Insurance & reinsurance

research direction

Input
Vendor model outputs, loss experience, exposure changes
Fragility signal
catastrophe models that agree for the wrong reason
Decision it changes
Price or reserve with an audited disagreement record

No vertical carries a proven advantage yet · results are published per task on the evidence page when they exist