Polling: Voting Intention

Every pollster is assigned a score derived from their historic accuracy. Firms with a proven track record of proximity to actual election results exert more influence on the trendline than those with higher historical error margins.

To prevent high-volume pollsters from skewing the average through sheer quantity, weights are adjusted for frequency. If a firm publishes multiple times within a 10-day window, their total weight is distributed across those polls. This ensures weekly trackers don't overpower fortnightly/monthly releases.

A robust regression model handles statistical noise by measuring how far a poll deviates from the trendline. The further a poll is from the trend, the less weight it carries, up to a limit of >10pts receiving zero influence.