Forecasters

Scoring Mechanism

• Outcome match (50%): Did the event (fog, rain, wind) actually occur?

• Timing accuracy (30%): Was the event observed during the forecast window?

• Zone specificity (20%): Was the prediction localized enough to matter?

Each forecast receives a composite score, which is tracked as a moving average over time.

Penalty Conditions

• No outcome match: Forecasts that miss entirely reduce rolling average

• Low activity ratio: Forecasts spaced too far apart lose temporal reliability weight

• Overgeneralization: Forecasts tagged as vague or non-local receive scaling penalty

• Forecast scoring does not block participation but reduces multiplier range.

Reward Distribution

• Forecast rewards are delayed until evaluation window closes

• Only forecasts above a minimum score threshold are eligible

• Top-performers (e.g. >85% match over last 10 forecasts) receive bonus multiplier

• Rewards are issued in weekly cycles based on historical accuracy average